Ibracon Structures and Materials Journal: February, 2012 ISSN 1983-4195 Volume 5, Number 1

REVISTA IBRACON DE ESTRUTURAS E MATERIAIS
IBRACON STRUCTURES AND MATERIALS JOURNAL

Volume 5, Number 1
February, 2012
ISSN 1983-4195
Contents
Steel fiber reinforced concrete pipes. Part 1: technological analysis of the mechanical behavior
A. D. DE FIGUEIREDO, A. DE LA FUENTE, A. AGUADO, C. MOLINS and P. J. CHAMA NETO
Steel fibre reinforced concrete pipes. Part 2: Numerical model to simulate the crushing test
A. DE LA FUENTE, A. D. DE FIGUEIREDO, A. AGUADO, C. MOLINS and P. J. CHAMA NETO
Simplified damage models applied in the numerical analysis of reinforced concrete structures
J. J. C. PITUBA and M. M. S. LACERDA
Cement-base bearing pads mortar for connections in the precast concrete: study of surface roughness
M. K. EL DEBS and E.K. BELLUCIO
The influence of low temperature on the evolution of concrete strength
V. CECCONELLO and B. TUTIKIAN
Reliability of buildings in service limit state for maximum horizontal displacements
A. G. B. CORELHANO, M. R. S. CORRA and A. T. BECK
A variable limit for the instability parameter of wall-frame or core-frame bracing structures
R. J. ELLWANGER
Editorial Board
Amrico Campos Filho
(Editor, UFRGS, Porto Alegre, RS, Brazil)
Jos Luiz Antunes de Oliveira e Sousa
(Editor, UNICAMP, Campinas, SP, Brazil)
Luiz Carlos Pinto da Silva Filho
(Editor, UFRGS, Porto Alegre, RS, Brazil)
Romildo Dias Toledo Filho
(Editor, UFRJ, Rio de Janeiro, RJ, Brazil)
Antonio Carlos R. Laranjeiras
(ACR Laranjeiras, Salvador, BA, Brazil)
Bernardo Horowitz
(UFPE, Recife, PE, Brazil)
Denise C. C. Dal Molin
(Former Editor, UFRGS,
Porto Alegre, RS, Brazil)
Emil de Souza Snchez Filho
(UFF, Rio de Janeiro, RJ, Brazil)
Geraldo Cechella Isaia
(UFSM, Santa Maria, RS, Brazil)
Gonzalo Ruiz
(UCLM, Ciudade Real, Spain)
Guilherme Sales Melo
(Former Editor, UnB, Brasilia, DF, Brazil)
Ivo Jos Padaratz
(UFSC, Florianpolis, SC, Brazil)
Joaquim Figueiras
(FEUP, Porto, Portugal)
Jos Marcio Fonseca Calixto
(UFMG, Belo Horizonte, MG, Brazil)
Mounir Khalil El Debs
(USP, So Carlos, SP, Brazil)
Nicole Pagan Hasparyk
(Former Editor, FURNAS,
Aparecida de Goinia, GO, Brazil)
Osvaldo Lus Manzoli
(UNESP, Bauru, SP, Brazil)
Paulo Helene
(Former Editor, USP, So Paulo, SP, Brazil)
Paulo Monteiro
(Berkeley, University of California,
Berkeley, CA, USA )
P.K. Mehta
(Berkeley, University of California,
Berkeley, CA, USA )
Pedro Castro Borges
(CINVESTAV, Mxico, D.F , Mxico)
Ronaldo Barros Gomes
(UFG, Goinia, GO, Brazil)
Rubens Machado Bittencourt
(Former Editor, FURNAS,
Aparecida de Goinia, GO, Brazil)
Tlio Nogueira Bittencourt
(Former Editor, USP,
So Paulo, SP, Brazil)
Vladimir Antonio Paulon
(UNICAMP, Campinas, SP, Brazil)
Reviewers
Reviewers are selected by the Editors
among the IBRACON members with
recognized competence in the specific
field of each contribution. They are acknowledged at the end of each volume.
The first issue of the fifth volume of the IBRACON Structures and Materials Journal is now released
(Volume 5, Number 1, February 2012). The seven articles published in this issue cover relevant topics
on concrete structures and materials. The first two address steel fiber reinforced concrete pipes: part
1 discusses technological analysis of the mechanical behavior and part 2 presents a numerical model
to simulate the crushing test. Simplified damage models applied in the numerical analysis of reinforced
concrete structures are presented in the third article. The fourth one reports studies on cement-base
bearing pads for connections in precast concrete, focusing on the surface roughness. The influence
of low temperature on the evolution of concrete strength is discussed in the fifth article. The sixth addresses reliability of buildings in service limit state for maximum horizontal displacements. The closing
article presents a discussion on a variable limit for the instability parameter of wall-frame or core-frame
bracing structures.
We appreciate the interest and dedication of authors and reviewers, responsible for the consolidation
of our journal. Thanks to them we are opening the fifth volume without interruptions or delays.
Amrico Campos Filho, Jos Luiz Antunes de Oliveira e Sousa, Luiz Carlos Pinto da Silva Filho
and Romildo Dias Toledo Filho
Editors
Estamos publicando o primeiro nmero do quinto volume da Revista IBRACON de Estruturas e Materiais (Volume 5, Nmero 1, Fevereiro de 2012). Em sete artigos tpicos relevantes sobre estruturas
e materiais de concreto so discutidos. Os dois primeiros abordam tubos de concreto reforado com
fibras de ao: a parte 1 aborda a anlise tecnolgica do comportamento mecnico e a parte 2 um modelo numrico para simular o ensaio de compresso diametral. Modelos simplificados de dano para
anlise de estruturas em concreto armado so apresentados no terceiro artigo. O quarto apresenta
um estudo da rugosidade superficial em almofadas de apoio feitas de argamassa para ligaes de
concreto pr-moldado. A influncia de baixas temperaturas na evoluo das resistncias do concreto
discutida no quinto artigo. O sexto aborda a confiabilidade de edifcios no estado limite de servio para
deslocamentos horizontais mximos. O nmero fecha com uma discusso sobre um limite varivel para
o parmetro de instabilidade de estruturas de contraventamento formadas por associaes de prticos
com paredes ou ncleos.
Gostaramos de destacar o interesse e a dedicao de autores e avaliadores, responsveis pela consolidao de nossa revista. Graas a eles estamos agora abrindo o quinto volume sem interrupes
ou atrasos.
Amrico Campos Filho, Jos Luiz Antunes de Oliveira e Sousa, Luiz Carlos Pinto da Silva Filho
e Romildo Dias Toledo Filho
Editores
Ibracon Structures
and Materials Journal
is published bimonthly
(February, April, June, August, October
and December) by IBRACON.
IBRACON
Instituto Brasileiro do Concreto
Founded in 1972
Cover Photo:
Burj El Arab, Dubai
Courtesy:

L. C. P. Silva Filho,
Porto Alegre, RS
Volume 5, Number 1
February 2012
ISSN: 1983-4195
R. Julieta do Espirito Santo Pinheiro, 68

Jardim Olmpia, So Paulo SP
Brasil 05542-120
Phone: +55 11 3735-0202
Fax: +55 11 3733-2190
E-mail: arlene@ibracon.org.br
Website: http://www.ibracon.org.br
Editors
Amrico Campos Filho
(Brazil)
Jos Luiz Antunes de O. e Sousa
(Brazil)
Luiz Carlos Pinto da Silva Filho
(Brazil)
Romildo Dias Toledo Filho
(Brazil)
Steel fiber reinforced concrete pipes. Part 1:

technological analysis of the mechanical behavior
A. D. de Figueiredo, A. de la Fuente, A. Aguado, C. Molins
and P. J. Chama Neto
Steel fibre reinforced concrete pipes. Part 2: Numerical

model to simulate the crushing test
A. de la Fuente, A. D. de Figueiredo, A. Aguado, C. Molins
12
Simplified damage models applied in the numerical
analysis of reinforced concrete structures
26
Cement-base bearing pads mortar for connections
in the precast concrete: study of surface roughness
38
The influence of low temperature on the evolution of concrete
strength
68
Reliability of buildings in service limit state for maximum
horizontal displacements
A. G. B. CORELHANO,M. R. S. CORRA and A. T. BECK
84
A variable limit for the instability parameter of wall-frame
or core-frame bracing structures
R. J. ELLWANGER
104
Cover design & Layout:
Ellementto-Arte
www.ellementto-arte.com
Aims and Scope

The IBRACON Structures and Materials Journal is a technical and
scientifical divulgation vehicle of IBRACON (Brazilian Concrete Institute).
Each issue of the periodical has 5 to 8 papers and, possibly, a technical
note and/or a technical discussion regarding a previously published
paper. All contributions are reviewed and approved by reviewers with
recognized scientific competence in the area.
Objectives
The IBRACON Structures and Materials Journals main objectives are:
Present current developments and advances in the area of concrete
structures and materials;
Make possible the better understanding of structural concrete
behavior, supplying subsidies for a continuous interaction among
researchers, producers and users;
Stimulate the development of scientific and technological research
in the areas of concrete structures and materials, through papers
peer-reviewed by a qualified Editorial Board;
Promote the interaction among researchers, constructors and users
of concrete structures and materials and the development of Civil
Construction;
Provide a vehicle of communication of high technical level for
researchers and designers in the areas of concrete structures
and materials.
Submission Procedure
The procedure to submit and revise the contributions, as well as the
formats, are detailed on IBRACONs WebSite (www.ibracon.org.br). The
papers and the technical notes are revised by at least three reviewers
indicated by the editors. The discussions and replies are accepted for
publication after a revision by the editors and at least one member of
the Editorial Board. In case of disagreement between the reviewer and
the authors, the contribution will be sent to a specialist in the area, not
necessarily linked to the Editorial Board.
Contribution Types
The periodical will publish original papers, short technical notes and
paper discussions. Announcements of conferences and meetings,
information about book reviews, events and contributions related to the
area will also be available in the periodicals WebSite. All contributions
will be revised and only published after the Editorial and Reviewers
Boards approve the paper. Restrictions of content and space (size) are
imposed to the papers. The contributions will be accepted for review
in Portuguese, Spanish or English. The abstracts are presented in
Portuguese or Spanish, and in English, independently of the language
in which the paper is written. After the review process, papers originally
written in Portuguese or Spanish should be translated into English,
which is the official language of the IBRACON Structures and Materials
Journal. Optionally, papers are also published in Portuguese or
Spanish.
Original papers will be accepted as long as they are in accordance
with the objectives of the periodical and present quality of information
and presentation. The instructions to submit a paper are detailed in the
template (available on IBRACONs WebSite).
The length of the papers must not exceed 20 pages.
A technical note is a brief manuscript. It may present a new feature
of research, development or technological application in the areas of
Concrete Structures and Materials, and Civil Construction. This is an
opportunity to be used by industries, companies, universities, institutions
of research, researchers and professionals willing to promote their
works and products under development. The instructions to submit a

technical note are detailed on IBRACONs WebSite.
A discussion is received no later than 3 months after the publication
of the paper or technical note. The instructions to submit a discussion
are detailed on IBRACONs WebSite. The discussion must be limited to
the topic addressed in the published paper and must not be offensive.
The right of reply is guaranteed to the Authors. The discussions and the
replies are published in the subsequent issues of the periodical.
Internet Access
IBRACON Structural Journal Page in http://www.ibracon.org.br
Subscription rate
All IBRACON members have free access to the periodical contents
through the Internet. Non-members have limited access to the published
material, but are able to purchase isolated issues through the Internet.
The financial resources for the periodicals support are provided by
IBRACON and by research funding agencies. The periodical will not
receive any type of private advertisement that can jeopardize the
credibility of the publication.
Photocopying
Photocopying in Brazil. Brazilian Copyright Law is applicable to users
in Brazil. IBRACON holds the copyright of contributions in the journal
unless stated otherwise at the bottom of the first page of any contribution.
Where IBRACON holds the copyright, authorization to photocopy items
for internal or personal use, or the internal or personal use of specific
clients, is granted for libraries and other users registered at IBRACON.
Copyright
All rights, including translation, reserved. Under the Brazilian Copyright
Law No. 9610 of 19th February, 1998, apart from any fair dealing for
the purpose of research or private study, or criticism or review, no part
of this publication may be reproduced, stored in a retrieval system,
or transmitted in any form or by any means, electronic, mechanical,
photocopying, recording or otherwise, without the prior written
permission of IBRACON. Requests should be directed to IBRACON:
IBRACON
Rua Julieta do Esprito Santo Pinheiro, n 68 ,Jardim Olmpia,
So Paulo, SP Brasil CEP: 05542-120
Phone: +55 11 3735-0202 Fax: +55 11 3733-2190
E-mail: arlene@ibracon.org.br.
Disclaimer
Papers and other contributions and the statements made or opinions
expressed therein are published on the understanding that the authors
of the contribution are the only responsible for the opinions expressed in
them and that their publication does not necessarily reflect the support
of IBRACON or the journal.
Objetivos e Escopo
A Revista IBRACON de Estruturas e Materiais um veculo de divulgao
tcnica e cientfica do IBRACON (Instituto Brasileiro do Concreto).
Cada nmero do peridico tem 5 a 8 artigos e, possivelmente, uma
nota tcnica e/ou uma discusso tcnica sobre um artigo publicado
anteriormente. Todas as contribuies so revistas e aprovadas por
revisores com competncia cientfica reconhecida na rea.
Objetivos
Os objetivos principais da Revista IBRACON de Estruturas e
Materiais so:
Apresentar desenvolvimentos e avanos atuais na rea de estruturas
e materiais de concreto;
Possibilitar o melhor entendimento do comportamento do concreto
estrutural, fornecendo subsdios para uma interao contribua entre
pesquisadores, produtores e usurios;
Estimular o desenvolvimento de pesquisa cientfica e tecnolgica
nas reas de estruturas de concreto e materiais, atravs de artigos
revisados por um corpo de revisores qualificado;
Promover a interao entre pesquisadores, construtores e usurios de
estruturas e materiais de concreto, e o desenvolvimento da
Construo Civil;
Prover um veculo de comunicao de alto nvel tcnico para
pesquisadores e projetistas nas reas de estruturas de concreto
e materiais.
Submisso de Contribuies
O procedimento para submeter e revisar as contribuies, assim
como os formatos, esto detalhados na pgina Internet do IBRACON
(www.ibracon.org.br). Os artigos e as notas tcnicas so revisadas por,
no mnimo, trs revisores indicados pelos editores. As discusses e
rplicas so aceitas para publicao aps uma reviso pelo editores
e no mnimo um membro do Corpo Editorial. No caso de desacordo
entre revisor e autores, a contribuio ser enviada a um especialista
na rea, no necessariamente do Corpo Editorial.
Tipos de Contribuio
O peridico publicar artigos originais, notas tcnicas curtas e
discusses sobre artigos. Anncios de congressos e reunies,
informao sobre reviso de livros e contribuies relacionadas
rea sero tambm disponibilizadas na pgina Internet da revista.
Todas as contribuies sero revisadas e publicadas apenas aps a
aprovao dos revisores e do Corpo Editorial. Restries de contedo
e espao (tamanho) so impostas aos artigos. As contribuies sero
aceitas para reviso em portugus, espanhol ou ingls. Os resumos
sero apresentados em portugus ou espanhol, e em ingls,
independentemente do idioma em que o artigo for escrito. Aps o
processo de reviso, artigos originalmente escritos em portugus ou
espanhol devero ser traduzidos para ingls, que o idioma oficial
da Revista IBRACON de Estruturas e Materiais. Opcionalmente, os
artigos so tambm publicados em portugus ou espanhol.
Artigos originais sero aceitos desde que estejam de acordo com
os objetivos da revista e apresentam qualidade de informao e
apresentao. As instrues para submeter um artigo esto detalhadas
em um gabarito (disponvel no stio do IBRACON).
A extenso dos artigos no deve exceder 20 pginas.
Um nota tcnica um manuscrito curto. Deve apresentar uma nova
linha de pesquisa, desenvolvimento ou aplicao tecnolgica nas reas
de Estruturas de Concreto e Materiais, e Construo Civil. Esta uma
oportunidade a ser utilizada por indstrias, empresas, universidades,
instituies de pesquisa, pesquisadores e profissionais que desejem

promover seus trabalhos e produtos em desenvolvimento. As instrues
para submisso esto detalhadas na pgina de Internet do IBRACON.
Uma discusso recebida no mais de 3 meses aps a publicao
do artigo ou nota tcnica. As instrues para submeter uma discusso
esto detalhadas na pgina de Internet do IBRACON. A discusso
deve se limitar ao tpico abordado no artigo publicado e no pode ser
ofensivo. O direito de resposta garantido aos autores. As discusses
e rplicas so publicadas nos nmeros subseqentes da revista.
Acesso via Internet

Pgina da Revista IBRACON de Estruturas e Materiais em
http://www.ibracon.org.br
Assinatura
Todos os associados do IBRACON tm livre acesso ao contedo do
peridico atravs da Internet. No associados tm acesso limitado ao
material publicado, mas podem adquirir nmeros isolados pela Internet.
O financiamento para suporte revista provido pelo IBRACON e
por agncias de financiamento pesquisa. A revista no receber
qualquer tipo de anncio privado que possa prejudicar a credibilidade
da publicao.
Fotocpias
Fotocpias no Brasil, A Lei Brasileira de Direitos Autorais aplicada
a usurios no Brasil. O IBRACON detm os direitos autorais das
contribuies na revista a menos que haja informao em contrrio no
rodap da primeira pgina da contribuio. Onde o IBRACON detm os
direitos autorais, autorizao para fotocopiar itens para uso interno ou
pessoal, ou uso interno ou pessoal de clientes especficos, concedida
para bibliotecas e outros usurios registrados no IBRACON.
Direitos autorais
Todos os direitos, inclusive traduo so reservados. Sob a Lei de
Direitos Autorais No. 9610 de 19 de fevereiro de 1998, exceto qualquer
acordo para fins de pesquisa ou estudo privado, crtica ou reviso,
nenhuma parte desta publicao pode ser reproduzida, arquivada em
sistema de busca, ou transmitida em qualquer forma ou por qualquer
meio eletrnico, mecnico, fotocpia, gravao ou outros, sem a
autorizao prvia por escrito do IBRACON. Solicitaes devem ser
encaminhadas ao IBRACON:
IBRACON
Rua Julieta do Esprito Santo Pinheiro, n 68 , Jardim Olmpia,
So Paulo, SP Brasil CEP: 05542-120
Fone: +55 11 3735-0202 Fax: +55 11 3733-2190
E-mail: arlene@ibracon.org.br.
Aviso Legal
Artigos e outras contribuies e declaraes feitas ou opinies
expressas aqui so publicadas com o entendimento que os autores
da contribuio so os nicos responsveis pelas opinies expressas
neles e que sua publicao no necessariamente reflete o apoio do
IBRACON ou da revista.
Diretoria
Binio 2011/2013
Conselho Diretor
Binio 2011/2013
Diretor Presidente
Tlio Nogueira Bittencourt
Conselheiros Individuais
Diretor 1 Vice-Presidente
Jos Marques Filho
Diretor 2 Vice-Presidente
Julio Timerman
Diretor 1 Secretrio
Antonio Domingues de Figueiredo
Diretor 2 Secretrio
Jos Tadeu Balbo
Diretor 1 Tesoureiro
Claudio Sbrighi Neto
Diretor 2 Tesoureiro
Carlos Jos Massucato
Diretor Tcnico
Ins Laranjeiras da Silva Battagin
Diretor de Eventos
Luiz Prado Vieira Jnior
Diretor de Pesquisa e Desenvolvimento
Ana Elisabete Paganelli Guimares A. Jacintho
Diretor de Publicaes e Divulgao Tcnica
Nelson Covas
Diretor de Marketing
Cludia Henrique de Castro
Diretor de Relaes Institucionais
Arcindo Vaquero Y Mayor
Diretor de Cursos
Iria Lcia Oliva Doniak
Diretor de Certificao de Mo-de-obra
Roseni Cezimbra
Tulio Bittencourt
Claudio Sbrighi Neto
Augusto Vasconcelos
Antonio Figueiredo
Waldimir Paulon
Geraldo Isaia
Nelson Covas
Luiz Prado
Ana Elisabete Jacinto
Antonio Carlos Laranjeiras
Conselheiros Mantenedores e Coletivos
ABCIC - Associao Brasileira da Construo Industrializada
de Concreto
ABCP - Associao Brasileira de Cimento Portland
ABESC - Associao Brasileira das Empresas de Servios
de Concretagem
CESP - Companhia Energtica de So Paulo
CNO Norberto Odebrecht
Escola Politcnica da Universidade de So Paulo - POLI
Eletrobrs - FURNAS - Centrais Eltricas S/A
L.A. FALCO BAUER
HOLCIM Brasil S/A
IPT - Instituto de Pesquisas Tecnolgicas do Estado de So Paulo
Conselheiros Permanentes
Eduardo Antonio Serrano
Jos Zamarion Ferreira Diniz
Paulo Roberto do Lago Helene
Ronaldo Tartuce
Rubens Machado Bittencourt
Selmo Chapira Kuperman
Simo Priszkulnik
Volume 5, Number 1 (February, 2012) p. 1-11 ISSN 1983-4195

Tubos de concreto reforado com fibras de ao. Parte 1:
anlise tecnolgica do comportamento mecnico
A. D. de Figueiredo a
antonio.figueiredo@poli.usp.br
A. de la Fuente b
albert.de.la.fuente@upc.edu
A. Aguado c
antonio.aguado@upc.edu
C. Molins d
climent.molins@upc.edu
Abstract
P. J. Chama Neto e
pchama@sabesp.com.br
This paper is the first part of an extensive work focusing the technological development of steel fiber reinforced concrete pipes (FRCP). Here is presented
and discussed the experimental campaign focusing the test procedure and the mechanical behavior obtained for each of the dosages of fiber used. In
the second part (Steel fiber reinforced concrete pipes. Part 2: Numerical model to simulate the crushing test), the aspects of FRCP numerical modeling
are presented and analyzed using the same experimental results in order to be validated. This study was carried out trying to reduce some uncertainties
related to FRCP performance and provide a better condition to the use of these components. In this respect, an experimental study was carried out using sewage concrete pipes in full scale as specimens. The diameter of the specimens was 600 mm, and they had a length of 2500 mm. The pipes were
reinforced with traditional bars and different contents of steel fibers in order to compare their performance through the crushing test. Two test procedures
were used in that sense. In the 1st Series, the diameter displacement was monitored by the use of two LVDTs positioned at both extremities of the pipes. In
the 2nd Series, just one LVDT is positioned at the spigot. The results shown a more rigidity response of the pipe during tests when the displacements were
measured at the enlarged section of the socket. The fiber reinforcement was very effective, especially when low level of displacement was imposed to
the FRCP. At this condition, the steel fibers showed an equivalent performance to superior class pipes made with traditional reinforced. The fiber content
of 40 kg/m3 provided a hardening behavior for the FRCP, and could be considered as equivalent to the critical volume in this condition.
Keywords: concrete pipes, steel fibers, crushing test, mechanical behavior, standards.
Resumo
A primeira parte deste trabalho se concentra no desenvolvimento tecnolgico de tubos de concreto reforado com fibras (TCRF). So apresentados e discutidos os resultados obtidos em um estudo experimental focado no procedimento de ensaio e na resposta mecnica obtida para cada
um dos teores de fibras utilizados nos tubos. Na segunda parte do trabalho (Tubos de concreto reforado com fibra de ao. Parte 2: Modelo
numrico para simular o ensaio de compresso diametral), se apresentam os aspectos associados simulao numrica da resposta mecnica
destes tubos. Este estudo foi levado a cabo com a finalidade de tentar reduzir algumas das incertezas relacionadas com o desempenho dos
TCRF e proporcionar melhores condies para viabilizar o uso das fibras como elemento de reforo em tubos. Neste sentido, se realizou um
trabalho experimental utilizando corpos-de-prova em escala real. O dimetro destes tubos foi de 600 mm e seu comprimento de 2500 mm. Alguns
dos tubos foram reforados com armaduras convencionais e outros apenas com fibras metlicas com a finalidade de comparar as diferentes respostas mecnicas no ensaio de trs cutelos. Empregaram-se dois procedimentos: numa 1 srie, o deslocamento diametral foi medido por meio
de LVDTs posicionados em ambas as extremidades dos tubos, enquanto que, na 2 srie se mediu o deslocamento apenas na ponta dos tubos.
Neste sentido, os resultados obtidos demonstraram que a avaliao da rigidez do sistema depende da forma de medida dos deslocamentos.
Alm disso, foi comprovado que o reforo de fibras metlicas era eficiente, especialmente para nveis de carga prximos s cargas de servio.
Para baixas deformaes, os TCRF apresentaram um melhor desempenho em termos de resistncia em comparao com os tubos de concreto
armado convencional. Por ltimo, se concluiu que o consumo de 40 kg/m3 pode ser considerado o volume crtico de fibras, pois pode obter um
encruamento para a resposta estrutural do tubo.
Palavras-chave: tubos de concreto, fibras de ao, ensaio de compresso diametral, normas.
Department of Construction Engineering, Barcelona Tech (UPC), albert.de.la.fuente@upc.edu, C/Jordi Girona Salgado, 1-3, 08034, Barcelona (Spain)
Department of Civil Construction Engineering, University of So Paulo (USP), antonio.figueiredo@poli.usp.br. Caixa Postal 61548, CEP 05508-900.
So Paulo (Brazil)
c
Department of Construction Engineering, Barcelona Tech (UPC), antonio.aguado@upc.edu, C/Jordi Girona Salgado, 1-3, 08034, Barcelona (Spain)
d
Department of Construction Engineering, Barcelona Tech (UPC), climent.molins@upc.edu, C/Jordi Girona Salgado, 1-3, 08034, Barcelona (Spain)
e
Companhia de Saneamento Bsico do Estado de So Paulo (SABESP), pchama@sabesp.com.br, So Paulo (Brasil).
a
b
Received: 11 Feb 2011 Accepted: 20 Oct 2011 Available Online: 01 Feb 2012
2012 IBRACON
1. Introduction
There is a huge necessity for sewage collecting and treatment systems in underdevelopment countries. The concrete pipes can be considered as a very competitive and sustainable alternative in relation to
competitors in the market (Violas et al. [1]). Brazil has an instrument
to reduce the deficiency in this segment since publication of the standard ABNT NBR 8890 [2] that allow the use of steel fibers for concrete
pipes reinforcement [3]. This standard is equivalent to the European
norm NBN-EM 1916 [4]. Both standards allow the use of conventional
reinforcement with steel bars also. The pipes must to be verified in
terms of mechanical performance through the crushing test or threeedge bearing test, despite the type of reinforcement used (bars of
fibers). However, the steel fibers reinforced concrete pipes (FRCP)
are not regularly used in Brazil nowadays, although the Brazilian standard had been published in 2007. One of the main reasons for this
situation is related to the doubts concerning the FRCP mechanical
performance that still remain in the market segment. These difficulties are also associated with the lack of numerical models capable
of predicting the behavior of the component accurately. In order to
resolve these problems, an experimental study focusing on the behavioral assessment of concrete sewage pipes was carried out. This
study compared the performance between the pipes reinforced with
steel fibers (FRCP) and the pipes reinforced with cold drown steel
bars. The results turn possible the comparative performance evaluation of FRCP and the steel bars reinforced concrete pipes (SBRCP),
and also the validation of a proposed numerical model for prediction
of the pipe behavior during the crushing strength (de la Fuente et al.
[5]). In this study, the influence of changes in testing methodology in
the response of the components was also evaluated.
2. Methodology
The mechanical performance of concrete pipes is regularly measured with the crushing test method as specified by the standards
ABNT NBR 8890 [2] and EN 1916 [4] (Figure [1]). These standards
prescribe two different procedures for the qualification of pipes,
one for SBRCP, and the other for FRCP.
In the case of SBRCP, the pipe is subjected to a continuous
loading up to rupture in order to determine the proof and ultimate loads. The proof load (Fc) is defined as 0.67 Fn, and Fn is
the minimum crushing load. The strength class is defined by the
Fc, and is acceptable to be equal to the crack load (Fcr), corresponding to a crack with an opening of 0.25 mm [2] or less than
0.3 mm [4] and length of 300 mm or more according to Brazilian
[2] and European [4] standards, respectively. The ultimate load
(Fu) is the maximum achieved during testing and shall be equal
or greater than Fn.
In the standard crushing test method for FRCP there is a cyclic
loading [2 and 4]. The first cycle consists of loading the pipe
until an equivalent level of the proof load of the SBRCP. At this
moment, the load is maintained for one minute and the pipe
is checked for evidence of any damage. The pipe is rejected
if any crack or other damage is observed. Thus, it is required
that Fcr shall be greater Fc. Once approved at this stage, the
loading of the pipe is continued until the maximum load (F u).
When the load start to decrease and reach the value of 95% of
Fu it shall be release, finishing the first cycle of the test. On the
second cycle of the test, the pipe is reloaded up to Fc and held
for one minute. The pipe has to withstand this post-peak proof
load in order to be approved. The test procedure established
by the European standard [4] is finished at this point. In the
specific case of the Brazilian standard [2], the second cycle
is continued increasing the loading up to the moment that the
pipe reach the maximum measured post-peak load (F max,pos).
The Fmax,pos shall be greater than 1.05F c, named minimum required post-peak load (F min,pos).
Although there are different test procedures prescribed by the
standards to verify pipes with varied types of reinforcement,
the continuous test was used in this experimental program in
order to evaluate all specimens under uniform conditions of
Figure 1 Front and side views of the configuration adopted

for the crushing test of concrete pipes
IBRACON Structures and Materials Journal 2012 vol. 5 n 1
A. D. de Figueiredo | A. de la Fuente | A. Aguado | C. Molins | P. J. Chama Neto
Figure 2 Configurations adopted for the measurement of displacements

(A) at both extremities in the 1st Series and (B) at the spigot for the 2nd Series
Figure 3 LVDT positioned at the upper part

of the inner surface of the pipe
assessment. This solution was chosen because it was proven that there is no influence of number of cycles used in the
test on the response of FRCP [6 and 7]. Devices were fixed
to specimens providing a continuous acquisition of diametral
displacement in order to improve the verification of pipes mechanical behavior during the tests. A similar test set up was
used in previous studies [8 and 9]. This system consists of
using LVDTs positioned against the upper part of the inner
surface of the pipe and attached at supports fixed at the bottom part of the pipe as shown in Figures [2] and [3]. The set
up used for LVDTs positioning allows the measurement of diametral displacements of the pipes without any interference of
external strains or dislocations in the results.
This paper presents an experimental program broader than previously published by the same authors [10]. Two test configurations
were used to measure the diametrical displacement of the pipes.
At first one, the measurement of displacement occurred simultaneously at the spigot and the socket, corresponding to positions
A and B (Figures [1] and [2a]). So, that situation turns possible to
obtain the curve of load by average diametrical displacement of
the pipes. In the second test configuration, only one LVDT was
placed at position A, corresponding to the spigot of the pipe (Figures [1] and [2b]). Thus, two series of pipes were produced with
fiber and conventional reinforcement to be tested with each of
those test configurations.
Each series was produced in a single day in order to diminish the
influence of intervening variables in the pipes production. In that
sense, the same features and basic materials that were being
regularly used by the company were applied for the pipes fabrication. The demanded amount of fiber was added directly on the
aggregates conveyor belt. The rebars set up used is the same that
had been routinely used by the factory (Figure [4]). Three levels of
steel fibers consumptions (10 kg/m3, 20 kg/m3 and 40 kg/m3) were
produced in each series together with three other SBRCP, totaling
24 pipes per series.
Figure 4. Rebars set up used in SBRCP production that served as

a parameter for comparative performance evaluation with FRCP.
Some problems were observed molding the 1st Series of pipes due to
increased cohesion provided by the steel fibers. These problems were
corrected in the 2nd Series by adjusting the moisture of the mixture to
maintain the fixed level of cohesion of the material. There was also
changing the cement brand from first to 2nd Series due to problems with
suppliers. The materials used in this study and its consumption per cubic
meter of concrete are presented in Table [1]. The steel fiber type used
in the experiment was a glued hooked end drawn wire, 60 mm in length
and diameter of 0.75 mm. So, the aspect ratio of the fiber was 80.
Figure 4 Rebars set up used in SBRCP

production that served as a parameter for
comparative performance evaluation with FRCP
Table 1 Material and its consumption

per cubic meter of concrete used
in the production of pipes used in
this experimental study
Material
Consumption
(kg/m3)
Fine aggregate I (Natural river sand)

Fine aggregate II (crushed sand)
Coarse aggregate (crushed stone)
Cement (CP III 40 RS)*
Water**
629
315
988
329
141
* A different brand of cement was used for each series of pipes.

** This amount of water was changed in the 2nd Series
through visual analysis of the mixture in order to maintain
fixed the concrete consistency.
The main steel reinforcement (transverse reinforcement) was made

with cold drawn steel bars (yield strength of 600 MPa) with a diameter of 6 mm, and spaced each 85mm. The main reinforcement
was positioned at the center of the pipe wall with equal distance for
both inner and outer surface. A support reinforcement (longitudinal)
made with 6 wires of the same type of steel was also used. The main
reinforcement was positioned at the center of the pipe wall. The consumption of steel for this family of pipes was approximately 40 kg/
m3. It was observed that the pipes produced with fibers presented a
very good final finishing of the surface (Figure [5]), which reinforces
the idea of the feasibility of its implementation.
After the production, the pipes remained in the storage yard of the
factory until the moment of the crushing tests. All pipes were tested
over the age of 28 days and at the same time. Other studies [11]
found that there was no significant change in the behavior of reinforced concrete pipe in the hardened state when tested at ages
ranging from 10 to 96 days. So, it was assured that there was no
influence of age of the pipes in the results. The pipes were produced using the vibro-compression method with high speed (5000
rpm) to compact the concrete. The geometric characteristics of the
pipes are presented in Figure [6]. A summary of the differences
between the two series is shown in Table [2].
3. Results and Analysis

The characteristic compressive strength of the concrete was 50
MPa, obtained directly from the regular quality control results
obtained by the factory. For each pipe, it was recorded the load
versus displacement curve during the crushing test. The results
obtained with the 1st Series, where the average displacement was
measured at the spigot and the socket simultaneously, are presented in Figures [7] to [10]. The results obtained with the measurement made only at the spigot, corresponding to the 2nd Series,
are presented in Figures [11] to [14]. In order to facilitate the comparison between the results, they are also presented in Figures
[15] and [16] in terms of average curves obtained for the first and
2nd Series, respectively. Only two results of the 1st Series were lost
Table 2 Differences between the two series of pipes used in the experimental study
Series
LVDT position
Cement
Water content
First
Second
At the spigot and socket

Only at the spigot
Brand 1
Brand 2
Fixed on 141 liters/m

Adjusted for each mix (constant consistency)
due to failures in the test system, where the reading of the strain
was impaired. Because of that, only two curves are shown in Figures [7] and [8].
The fact that pipes have been tested in real scale caused a reduction of the results variability. In ordinary tests to determine the
toughness of fiber reinforced concretes, the coefficient of variation
can exceed 20% [12]. This occurs because the crack area (where
the fibers act as a stress transfer bridge) is much larger in a tube
than in a prismatic specimen. This reduction in variability has been
observed in previous studies [7 and 13].
The first perceptible aspect was the behavior with a well defined
pattern for pipes reinforced with smaller amounts of fiber. This
softening pattern was characterized by a reduction of the strength
of the pipe with increasing vertical displacement. The pipes reinforced with steel bars or 40 kg/m3 of fibers had presented a
Figure 6 Geometric characterization

of the pipe used in the experiment
Figure 5 General appearance of a

new fiber reinforced concrete pipe
Figure 7 Load versus displacement curves

obtained during crushing test of pipes with a
fiber consumption of 10 kg/m3 and with the
diametrical displacement measured
at the spigot and the socket

In order to facilitate the analysis of the influence of the setups for

measuring displacements in the pipes during the crushing test
the results of the average curves obtained with the greater fiber
content are presented in Figure [17]. The results obtained in the
1st Series, where the displacement measurements were made at
the spigot and at the socket simultaneously, had presented higher
stiffness in the beginning of the curve. The end of the initial elastic
line was prolonged by the influence of the socket region where the
concrete matrix contributes more intensely [14]. For the 2nd Series,
where the displacements were measured only at the spigot, they
were more intense for smaller loads, due to lower rigidity of this part

conventional reinforcement and with the


only at the spigot
hardening pattern, which consists of increasing resisted load with

increasing vertical displacement of the pipe. Thus, the resistance
displayed by pipes with regular reinforcement was largest than the
presented by FRCP at higher displacements. However, for lower
strain, the residual strength capacity was greater for FRCP. This
occurs because the bars are positioned at the innermost part of
the pipe wall and they are mobilized more intensely when there
is greater displacement and hence higher level of cracking. On
the other hand, the fibers are distributed throughout the thickness
of the pipe wall. Thus, the fibers reinforcement is mobilized with
small levels of displacement and cracking, which provides better
performance in this situation.

only at the spigot
because the concrete reaches its elastic strain limit in this region
before it occurs in the socket. In the particular case of this study it
occurs around 80 kN of loading, when the stiffness of the component was greatly reduced. As the displacements measured in the
2nd Series were much superior to the 1st Series due to the previously discussed loss of stiffness, it is also possible to conclude for
the equivalence of the two behaviors, because both series reached
close levels to residual strength. It should be emphasized that this
pattern of behavior is totally consistent with the model proposed
by de la Fuente et al. [5] for numerical modeling of pipes behavior
during the crushing test.
An important aspect to have attention is the fact that it is impossible to carry out the standard cyclic test with pipes reinforced with
40 kg/m3 of steel fibers. The difficulty occurs because of the hard-

obtained during crushing test of pipes with
conventional reinforcement and with the
only at the spigot

obtained during crushing test of pipes with
a fiber consumption of 40 kg/m3 and with the
only at the spigot
Figure 15 Average load versus diametrical

displacement curves of pipes tested
in the 1st Series
of the pipe. The difference observed in the pattern of the curves

obtained in the two series of test is only attributed to geometric
factors, and the characteristics of matrix and type of reinforcement
have no influence on the results. So, the displacement measured
at the socket will always be smaller than the one measured at the
spigot. In that sense, the enlargement of the socket is the main
factor influencing the pattern of the curves at the elastic region.
Due to the fact that the spigot will present a higher level of displacement, this region will present the higher level of strain and,
consequently, will reach the elastic limit sooner. So, is possible to
conclude that the cracking of the pipes always begins at the spigot,
Figure 16 Average load versus diametrical

displacement curves of pipes tested
in the 2nd Series
ening behavior of these pipes. In this particular case, the drop in

post-peak strength at low displacements could not possible be observed. Thus there will be a high level of displacement and cracking when the unloading was done during the regular cyclical test.
So, the standard procedure had the implicit concept that the FRCP
should always show a softening behavior to turn possible to perform the cyclical test. Thus, the continuous measurement system
using LVDTs provides greater reliability to the test in these situations. Especially when the serviceability conditions associated to
the level of crack opening is considered.
There were also regions of post-peak instability in the crushing test
response, similar to those observed by Figueiredo [7] when low
amounts of fiber were used. This instability is provoked by the process of slipping and loss of adhesion of the fibers, in addition to
possible rupture of some of them, occurring during the stress transfer from the matrix to the fibers. Furthermore, the increase in fiber
content enhanced the cracking load of the matrix, corresponding to
the end of the initial linear part of the curve. This, however, became
clear only for the 1st Series of pipes. In this series there was an increase of the material cohesion due to the fiber addition in to the matrix. So, there was the correction of water content in the concrete for
the 2nd Series in order to maintain the level of mobility of the matrix.
In that situation, the increase in water/cement ratio gives a reduction
of the matrix strength, which reduced the contribution of fiber to the
elastic behavior of the component. Although this, a reduction in the
load at the end of the elastic part of the curves was not observed. It
occurs due to the increase in the fiber content that had improved the
tensile strength capacity of concrete pipe.
It was observed that the pipes also met the requirements of the
standards [2 and 4] for Class AE2. In the case of pipes with 600
mm nominal diameter the Brazilian [2] and European [4] standards
require that the pipes have to present a load control of 36 kN/m
and an ultimate load of 54 kN/m. As the pipe length was 2.5 m, it
accomplish that a total load control is 90 kN per pipe, and a ultimate load is 135 kN per pipe.
Checking the average results presented in Figures [15] and [16], it is
possible to figure out that all SBRCP exceeded both the control and
ultimate load by a wide margin, especially in the 1st Series. In this
particular case, the limit of the elastic behavior reached about 130
kN (less than the proof load where the pipe may have a certain level
of cracking). The average ultimate load was 230 kN, obtained directly from the test machine record system because it occurs when the
displacement was larger than the measurement limit of the LVDTs.
With these results, these pipes were about to fit the requirement of
the superior class (EA3 for the Brazilian standard), which requires
control and ultimate loads of 135 kN and 202.5 kN, respectively. In
the 2nd Series, the results were slightly lower than the first one for
the ultimate strength, which reached the averaged load of 174.7 kN.
Since the end of the initial elastic segment was close to 90 kN, it is
possible to affirm that the pipes reached the standards requirement
narrowly. Anyway, it is possible to consider them as appropriate to
the requirements of the standards for class EA2.
Observing the curves presented in Figures [15] and [16], it is
possible to affirm that there is proximity between the behavior of
SBRCP and FRCP in the first half of the load versus displacement curves. This behavior occurs in more intense way until a
vertical displacement of about 0.4 mm and 2 mm for the first and
2nd Series respectively. The FRCP with 40 kg/m3 exceeded the
strength loads of SBRCP until displacements of about 2.5 mm and
4.5 mm for the first and 2nd Series respectively. It is possible to
affirm that the FRCP presented a similar behavior to a superior
class of SBRCP in those circumstances. The initial region of the
curves could be associated to the serviceability limit state or, in
other words, corresponding to the elastic or low level of cracking
behavior. So, the fiber contribution will be more effective for the
pipes service life than the contribution presented by the steel bars.
Note that all the pipes of the 1st Series showed the initial elastic
part of the curve slightly exceeding 100 kN. With respect to this
aspect, it was observed that cracking began, in all cases, at the
top of the section A with values of Fcr around 90 kN. On the other
hand, in the pipes from the 2nd Series, it was observed that the appearance of the first cracks coincided with the change in the slope
of the load-displacement curve. This change was not detected in
the pipes from the 1st Series until the springline had also cracked,
Figure 17 Comparison between the average

curves measured at the socket and
spigot (1st Series) and only at the spigot
(2nd Series) for FRCP with 40 kg/m3 steel fibers
Table 3 Maximum load and maximum post-peak load obtained in the two series of tests
Series
st
nd
Cf
3
(kg/m )
F1.2mm y F3mm (kN)
Fu (kN)
Individual values
Individual
values
Average
Average
10
149
120
n/a
135
107
94
n/a
101
20
140
124
n/a
132
118
102
n/a
110
40
156
163
149
156
10
127
132
138
132
97
88
98
94
20
151
145
129
142
121
120
108
116
40
152
140
193
162
a situation which took place when load F reached average values

of 126 kN, 114 kN and 138 kN for the pipes with 10 kg/m3, 20 kg/
m3 and 40 kg/m3, respectively. So, the load Fcr can be deduced
from the load-displacement curve at the point where the first slope
change takes place. On the other hand, if the measurement is performed as in the 1st Series, this procedure would lead to values of
Fcr on the side of insecurity, requiring a visual inspection in order to
detect Fcr., which could indicate if the control load fits the standard
requirements. Since the load displacement curves obtained to the
2nd Series of tests showed that linear region finishing when the load
was about 90 kN. Thus, these pipes would be probably approved
because is possible to associate this level of loading to the Fcr..
The Brazilian standard requirements [2] for FRCP are a little broader than the European [4]. This is because, besides the load control
(Fc) of 90 kN and ultimate load (Fu) of 135 kN, common to both, the
Brazilian standard requires a minimum post-peak load (Fmin,pos) of
94, 5 kN. This fact increases the importance of the post-peak residual strength of the pipes. As the tests performed for this study have
not the cycle of loading, this parameter could not be determinate
Figure 18 Correlations between ultimate

load and fiber content for both series of tests
in order to evaluate if the pipes comply with the requirements. Nevertheless, a displacement level was fixed in order to determining a
residual strength in the post-peak region to assess the influence of
the fiber in this specific behavior of the pipe. The results presented
in Table [3] and also in Figures [15] and [16] were used for the
evaluation of this condition. As tests were performed continuously
without re-loading cycle, it was adopted a reference displacement
a little beyond post-peak instability limit for each series. Consequently, the displacements of 1.2 mm and 3 mm were adopted
for determining the post-cracking load for the first and 2nd Series,
respectively. The different level of displacement was chosen due
to the disparity in geometric rigidity of pipes observed when the
displacement was measured only at the spigot or at the spigot and
socket. That condition provides different level of displacement for
the same level of cracking and post-peak residual strength.
In Figures [18] and [19] are presented the correlations between the
fiber content and ultimate load (Fu) and the post-cracking load at
Figure 19 Correlations between fiber

content and the post-peak load at the
displacements of 1.2 mm and 3 mm for
the first and 2nd Series of tests respectively
the displacements of 1.2 mm and 3 mm, for the first and 2nd Series
respectively. It is possible to observe clearly that the fibers content
interferes in both parameters. The linear regression between fiber
content and ultimate load or post-cracking load provide good level
of correlations, as shown by the coefficients of correlation. Thus,
this type of curves can be used in a mix-design procedure, if the
variability of the test is well known. However, the fiber mix-design
in order to accomplish a required post-peak residual strength has
a greater relevance when the Brazilian standard [2] is used for
pipes control.
5. Acknowledgements
4. Final remarks
6. References
The performance demonstrated by the FRCP was higher than

the presented by the SBRCP when they were submitted to lower levels of displacement and cracking. A similar behavior has
been observed in previous studies [9]. This fact occurs because
the fibers are mobilized early in the process of cracking due to
their position along the wall surface of the component. Since
the steel bars are eventually being placed along the neutral
line, by the requirement of minimum coverage, it will demand
a high level of displacement and cracking in order to mobilize
their resistance. So, the behavior of FRCP with low contents of
fiber is typically softening, while the pipes with rebars present a
typical hardening behavior. Thus, even for low fiber consumptions, the performance of FRCP is comparable or even superior
to SBRCP at low cracking and displacement level. This is particularly interesting since this is the stage of primary interest to
the component application because the pipes are still in good
conditions to meet the durability requirements. Moreover, in the
particular case of this experimental study, the SBRCP had much
better performance than the prescribed class. It also indicates
that the fiber reinforcement is so efficient that approximates
the behavior of FRCP to the one presented by a higher class
SBRCP at the serviceability conditions.
It has been shown that the fiber content affects equally the ultimate
load as the post-peak maximum load presented by the pipe in the
crushing strength test. For this reason, it is important to take into
account the contribution of the fiber in mix design analysis of the
FRCP concerning to meeting the requirements of the ultimate and
maximum post-peak load.
The results also confirmed the critical condition related to the
spigot displacement in relation to the socket area. This fact
shows that the component displacement measured at this position could be associated to the more demanding pipe condition in terms of displacement and crack initiation. However, this
is a conservative approach concerning only to the test method,
since the spigot will be protected by the follow pipe socket during
use. This effect, however, is not applicable to pipes where the
socket has no enlargement [2] [4]. Thus, the crushing test carried
out in pipes which diametrical displacement is measured only at
the spigot is very much in favor of security. Furthermore, this arrangement test proved to be much more suitable to evaluate the
pipes performance. This occurs because, for pipes with higher
fiber consumptions, it would not be possible to verify when the
load drops to 95% of its value at low level of displacement. In
that situation, the technician in charged to run the test could not
observe the right moment to impose the end of the first cycle and
start up the second one.

[01]

[02]

[03]

[04]

[05]

[06]

[07]

[08]

[09]

[10]

[11]

[12]

10
The authors of this paper wish to express their appreciation for

the support received from Fermix Indstria e Comrcio Ltda. and
Belgo Bekaert Arames that turns possible the entire experimental
work.
Likewise, Professor Antonio D. de Figueiredo wishes to thank the
support provided by CAPES -Coordenao de Aperfeioamento
de Pessoal de Nvel Superior for having awarded him the postdoctoral grant that allowed him to participate in this work.
VIOLAS, V., AGUADO, A., JOSA, A. Evaluacin de

la sostenibilidad en tuberas de saneamiento.
II Congreso UPC Sostenible 2015, Barcelona, Spain,
2009.
ASSOCIAO BRASILEIRA DE NORMAS
TCNICAS. Tubo de concreto, de seo circular,
para guas pluviais e esgotos sanitrios. NBR 8890,
ABNT, Rio de Janeiro. 2007.
FIGUEIREDO, A. D., CHAMA NETO, P. J. A nova
especificao brasileira para tubos de concreto
para guas pluviais e esgoto In: 49 Congresso
Brasileiro do Concreto CBC 2007, Instituto Brasileiro
do Concreto (IBRACON), Bento Gonalves. 2007.
NBN EN1916 Concrete pipes and fittings,
unreinforced, steel fibre and reinforced. Belgisch
Instituut voor normalisatie (BIN), Brussel, December 2002.
de la Fuente, et al. Steel fiber reinforced concrete
pipes. Part 2: Numerical model to simulate the
crushing test. Revista RIEM. In press.
FIGUEIREDO, A.D.; CHAMA NETO, P. J.; QUINTA,
M. T.; GIMENEZ, A. B. Avaliao de metodologia de
ensaio de tubos de concreto reforado com fibras para
esgoto. Concreto (So Paulo). Instituto Brasileiro do
Concreto (IBRACON). Vol.XXXIV, p.44 - 51, 2007.
FIGUEIREDO, A.D. de. Evaluation of the test method
for crushing strength of steel fiber reinforced concrete
pipes. 7th International RILEM Symposium on Fibre
Reinforced Concrete, Chennai, India, 2008.
RAMOS, M. F. Anlise experimental de tubos de
concreto reforado com fibras de ao. Campinas,
UNICAMP. Universidade Estadual de Campinas.
Faculdade de Engenharia Civil. Dissertao
(Mestrado em Engenharia Civil). Campinas, 2002.
CHAMA NETO, P.J.; FIGUEIREDO, A.D. de.
Avaliao de desempenho mecnico de tubos.
Revista DAE, 2008, Vol. 178, p. 34-39.
de la FUENTE, A., FIGUEIREDO, A. D., AGUADO,
A., MOLINS, C. CHAMA NETO, P. J. Experimentacin
y simulacin numrica de tubos de hormign com
fibras. Materiales de Construccin, 2011, Vol. 61,
n 302, p. 275-288.
de la FUENTE, A., AGUADO, A., MOLINS, C. Modelo
numrico para el anlisis no lineal de secciones
prefabricadas construidas evolutivamente. Hormign
y Acero, 2008, Vol. 57, n 247, p. 69-87.
MORGAN, D.R.; MINDESS, S.; CHEN, L. Testing and
Specifying Toughness for Fiber Reinforced Concrete
and Shotcrete. In: Second University-Industry Work

[13]

[14]

shop on Fiber Reinforced Concrete and Other

Advanced Materials. Toronto, Canada, 1995.
Proceedings. p.29-50.
LAMBRETCHS, A. Performance clases for steel
fibre reinforced concrete: Be critical, 7th International
RILEM Symposium on Fibre Reinforced Concrete,
Chennai, India, 2008.
SILVA, J. L. da ; El DEBS, M.K. . Influncia da bolsa
no comportamento estrutural de tubos de concreto
armado submetidos compresso diametrical.
In: 51 Congresso Brasileiro do Concreto, 2009,
Curitiba. 2009. p. 1-13.
11

Tubos de concreto reforado com fibras de ao.
Parte 2: Modelo numrico para simular o ensaio
de compresso diametral
A. de la Fuente a
albert.de.la.fuente@upc.edu
A. D. de Figueiredo b
antonio.figueiredo@poli.usp.br
A. Aguado c
antonio.aguado@upc.edu
C. Molins d
climent.molins@upc.edu
Abstract
P. J. Chama Neto e
pchama@sabesp.com.br
This paper is part of an extensive work about the technological development, experimental analysis and numerical modeling of steel fibre reinforced concrete pipes. The first part (Steel fibre reinforced concrete pipes. Part 1: technological analysis of the mechanical behavior) dealt with
the technological development of the experimental campaign, the test procedure and the discussion of the structural behavior obtained for each of
the dosages of fibre used. This second part deals with the aspects of numerical modeling. In this respect, a numerical model called MAP, which
simulates the behavior of fibre reinforced concrete pipes with medium-low range diameters, is introduced. The bases of the numerical model are
also mentioned. Subsequently, the experimental results are contrasted with those produced by the numerical model, obtaining excellent correlations. It was possible to conclude that the numerical model is a useful tool for the design of this type of pipes, which represents an important step
forward to establish the structural fibres as reinforcement for concrete pipes. Finally, the design for the optimal amount of fibres for a pipe with a
diameter of 400 mm is presented as an illustrating example with strategic interest.
Keywords: concrete pipes, fibres, crushing test, numerical model, optimal design
Resumo
Este artigo faz parte de um extenso trabalho relacionado ao desenvolvimento tecnolgico experimental e modelagem numrica de tubos de
concreto reforados com fibra de ao. Na primeira parte (Tubos de concreto reforado com fibras de ao. Parte 1: Anlise tecnolgica do comportamento mecnico), foi apresentado o estudo experimental com enfoque tecnolgico, abordando o procedimiento de ensaio e a discusso do
comportamento estrutural obtido para cada consumo de fibra e tipo de reforo empregado. Nesta segunda parte, so abordados os aspectos de
modelagem numrica. Neste sentido, se apresenta um modelo numrico para a simulao do comportamento de tubos de dimetros inferiores a
1000 mm chamado MAP. So explicitadas as bases do modelo numrico e, posteriormente, seus resultados so confrontados com os obtidos experimentalmente, obtendo-se excelentes nveis de correlaco. Se conclui que a ferramenta numrica til para a otimizao deste tipo de tubos,
o que representa um avano importante para a implantao das fibras estruturais como refuero de tubos de concreto. Alm disso, com o objetivo
de proporcionar um exemplo de interesse estratgico, se apresenta a otimizao do consumo de fibras para un tubo de 400 mm de dimetro.
Palavras-chave: tubos de concreto, ensaio de compresso diametral, modelo numrico, otimizao.
Department of Construction Engineering, Barcelona Tech (UPC), albert.de.la.fuente@upc.edu, C/Jordi Girona Salgado, 1-3, 08034, Barcelona (Spain)
Department of Civil Construction Engineering, University of So Paulo (USP), antonio.figueiredo@poli.usp.br. Caixa Postal 61548, CEP 05508-900.
So Paulo (Brazil)
c
Department of Construction Engineering, Barcelona Tech (UPC), antonio.aguado@upc.edu, C/Jordi Girona Salgado, 1-3, 08034, Barcelona (Spain)
d
Department of Construction Engineering, Barcelona Tech (UPC), climent.molins@upc.edu, C/Jordi Girona Salgado, 1-3, 08034, Barcelona (Spain)
e
Companhia de Saneamento Bsico do Estado de So Paulo (SABESP), pchama@sabesp.com.br, So Paulo (Brasil).
a
b
Received: 11 Feb 2011 Accepted: 20 Oct 2011 Available Online: 01 Feb 2012
2012 IBRACON
1. Introduction
Unreinforced concrete pipes (UCP) and steel bar reinforced concrete pipes (SBRCP) are well-known and accepted solutions for
drainage and sewage pipes (Violas et al. [1]).
On the other hand, fibre reinforced concrete pipes (FRCP) and
those reinforced with steel rebars and fibres (SBFRCP) are other
underdevelopment alternatives (Haktanir et al. [2], de la Fuente et
al. [3 and 4], Figuereido [5], Figueiredo et al. [6] and Lambrechts
[7]). In this respect, the addition of fibres provides advantages from
both the technical and the economic point of view. From the technical point of view, a substantial improvement of several mechanical
properties of concrete is achieved (Asad et al. [8]), especially with
the addition of metallic fibres (Blanco [9]). Likewise, the composite solution leads to a positive structural synergy: the steel rebars
perform the main strength function (Chiaia et al. [10]), whereas
the fibres bridge the cracks, reducing their average spacing and
width. The fibres also contribute to the strength function (Blanco et
al. [11]). The use of fibres also contribute economically, because
allows saving up on the assembling operations related to conventional reinforcement, reducing labor force, equipment use, and associated risks (de la Fuente et al. [12]).
FRCP and SBFRCP have already been considered as alternatives for UCP and SBRCP in several experimental campaigns
both in Brazil (see Figueiredo et al. [6 and 13]) and Spain (see
de la Fuente et al. [3]). However, their introduction in the market is under progress due to several factors such as: (1) the risk
Figure 1 Three edge bearing test

or crushing test
of damage when FRCP are manipulated; (2) the lack of calculation methods for this type of material, and (3) the difficulty to
overcome the inertia towards change (Parrot [14]). Nonetheless,
nowadays there are solutions for such problems: (1) polishing
with emery powder in order to remove imperfections and avoid
possible injuries; (2) constitutive equations to consider the tensile
behavior of the steel fibre reinforced concrete (SFRC) (Hillerborg
et al. [15], Vandewalle et al. [16] and Laranjeira et al. [17]), and
(3) it has been verified that the incorporation of fibres improves
the response of the pipe and leads to a global reduction of costs
(Pedersen 1992 [18]).
Another relevant aspect related to FRCP and SBFRCP technology
is the lack of recommendations and simplified calculation methods.
Because of this, the design of FRCP and SBFRCP is normally carried out by trial and error: trying out several dosages and/or concrete thickness until finding an optimal amount of fibres that meet
the requirements of the desired strength class in the crushing test
(CT) (Figure [1]). This design procedure is hardly operative, uneconomical and inefficient due to the variety of diameters, thickness,
strength classes, types of fibres and the factory limitations. For this
reason, it is necessary to develop analytical and/or numerical tools
that would make possible to carry out the optimal design and the
verification of concrete pipes (CP), especially FRCP and SBFRCP,
in order to avoid the regular procedures traditionally used.
The aim of this paper is, firstly, to introduce a model for the nonlinear analysis of CP of medium-small diameter (less than 1000
mm) called Mechanical Analysis of Pipes (MAP) which is able to
simulate the CT; and, secondly, to contrast the numerical and the
experimental results in order to achieve the model validation.
Initially, a summarized exposition of the normalized CT procedure
is presented. Then, the bases considered in the MAP model are
mentioned, and the model results are contrasted with the results
presented in the first part of this work (Figueiredo et al. [19]). Finally, an example of the application of MAP is presented aiming at
determining the optimal amount of fibres for a pipe with 400 mm
of Di.
2. Crushing Test
The NBR 8890:2007 [20] specifies the procedures and all the details that should be observed during the execution of the CT. Both
the cross and the longitudinal sections of the test configuration are
schematically shown in Figure [2].
The load process and the strength requirements are function of the
type of reinforcement. In the case of steel fibre reinforced concrete
pipes (SFRCP) the requirements are presented below:
1. Withstand the proof load (Fc) during a minute without cracking
or, in other words, without exceeding the first cracking load (Fcr).
Fc is equivalent to the 67% of the minimum failure load (Fn).
2. Reach the ultimate failure load Fu, which must be higher than Fn.
3. When the load has decreased a 5% of Fu, or more, the pipe is
completely downloaded and reloaded until reaching Fc. This
load level must be supported for more than a minute.
4. The loading process must continue until reaching a minimum
post failure load (Fmin,pos) equivalent to, at least, 105% of Fc.
The purpose of this cyclic loading process is to verify if the type
and amount of fibres are the suitable ones to guarantee the Fmin,pos
load and, indirectly, if the fibre-concrete anchorage and the postpeak strength of SFRC are appropriate (Figueiredo [5]).
13
Figure 2 Crushing test: (a) cross section and (b) longitudinal section
l/2
F
Spreading beam
Ridge
DO
Di
Springline
Invert
Supports
F/2
F/2
3. Model for the simulation of the

crushing test
The required subroutine for the simulation of the CT up to high
displacement levels should take in to account paramount aspects
as the cracking the post-failure response of the materials and the
modeling of the SFRC behavior. In that sense, the Analysis of Evolutive Sections (AES) introduced in de la Fuente et al. [21] was
used in order to deal with these aspects.
On the other hand, the MAP routine, which includes the AES
model, was also developed. The bases for the structural model
were already suggested by Pedersen [22] for the analysis of
pipes with a small diameter. However, for this work, several
changes were made as regards the behavior at the sectional
level, the constitutive equations of SFRC and the possibility of
considering the coexistence of steel rebars and structural fibres
as reinforcement.
This section puts forward the main foundations of the AES model,
highlighting the modifications introduced for this work, as well as the
analytical equations of the MAP model and the calculation algorithm.
3.2 Sectional analysis model

3.2.1 Modeling the materials
The AES model discretizes the concrete in 2-D differential elements (dAc), and the steel rebars in elements with concentrated
area As,i in its gravity center ys,i. Then, it assigns the suitable con-
14
stitutive model to each material and integrates the stresses resulting from a given strain plane (see Figure [3a]). The total concrete
strain c(t,t0), assessed at an instant of time t, is considered to
be the sum of the mechanical strains cm(t0), produced instantaneously at to, and the non-mechanical strains cnm(t,t0) (see de la
Fuente et al. [21] and Mar et al. [23]). In this paper, cnm(t,t0) are
not considered since the test only takes a few minutes to be executed, insufficient time to present non-mechanical strains due to
the concrete creep. Likewise, according to Heger [24], shrinkage
hardly influences the stress state of the pipe cross section; hence,
it is also disregarded in this model.
For the simulation of the concrete compressive behavior the
diagram suggested by Thorenfeldt et al. [25] is used, since
it could be adjusted correctly to a wide range of concrete
strengths and suitably simulates the post-failure response. On
the other hand, the tensile behavior and concrete stiffening between cracks is described by means of the equation proposed
in Collins et al. [26].
According to Bencardino et al. [27], the inclusion of metallic fibres modifies the SFRC compressive behavior depending on
the volume of fibres used. In this respect, the expression suggested by Barros et al. [28] fits properly the uniaxial compressive behavior in the post-failure regime of SFRC. On the other
hand, the simulation of its tensile behavior has been dealt by
means of the c-c model (see Figure [3b]) proposed in (Vandewalle et al. [16]), because it has already been used in several
numerical-experimental contrasting tests (see Pujadas [29]),
guaranteeing good results.
Figure 3 (a) Cross section discretization, (b) Compression and tensile SFRC laws
and (c) constitutive diagram adopted to simulate the passive steel behavior
The value of the crack width (w) is calculated considering that the crack
surfaces rotate as a rigid body (see Figure [4]), forming a angle between the crack faces (Eq. 1). This angle is related to the sectional curva-
Figure 4 Rigid body schema adopted

to assess w
Non linear regime
Linear regime
c(c)
Linear regime
ture by means of the length of the hinge lbc through Eq. 2 (see Pedersen
[30]). The value of lbc varies depending on the stress level of the section;
however, some authors (Pedersen [30] and Olesen [31]) establish it as
a constant value of h/2, and still others (Casanova [32]) propose that it
should vary depending on the crack height (sn). For this paper, a constant value of h/2 for lbc has been adopted, following the recommendations proposed by Pedersen [30] for the analysis of FRCP.
w = sn tan
(1)
= lbc
(2)
h
sn
w
lbc = h/2
The steel rebars are modeled with a trilineal diagram, with the
possibility of simulate the hardening response of the material (see
Figure [3c]).
15
3.2.2 Basic hypotheses about sectional behavior

The following classical hypotheses have been adopted for the modeling
the sectional behavior: (1) the sections have a symmetry axis and are subjected to straight flexo - compression; (2) perfect bond between the materials in the section; (3) sections initially planes remain planes after applying
the forces; (4) shear strains are negligible and therefore are not considered;
and (5) the pipe curvature does not affect on stress-strain distribution.
3.2.3 Idealization of the section
The positive signs are for: (1) the bending moments which com-
press the upper fibre; (2) the axial forces which compress the section; (3) compressive stresses and (4) the shortenings.
3.2.4 Equilibrium and compatibility equations
The stress-strain state (see Figure [3a]) resulting from a combination of internal efforts (normal force N and bending moment M) is defined by the stress distribution of the materials
and the plane formed by the strain in the most bottom layer
of the concrete section (taken as the reference layer) and the
sectional curvature (c,inf, ). This state is obtained by applying
the internal balance equations (Eqs. 3 and 4) and establish-
Figure 5 Schema adopted to simulate the crushing test
F
a)
Mc
b)
M
Rm
Rm
Rm
Mc
Ms
Rm
Ms
F
C
Mc
c)
C
d)
vc
M
Rm
Rm
us
S
F
Ms
16
Rm
Rm
ing the hypothesis of perfect bond between concrete and steel

(Eq. 5).
(3)
(4)
(y) = c,inf - y.
(5)
Eqs. 3-5 leads to a nonlinear system of equations which is solved

by using the Newton-Raphson iterative method (see Yang et al.
[33]). After solving the system, the values of the unknown parameters c,inf and , which define the strain plane, are obtained.
3.3 Structural analysis model
The applied force F, and the bending moments at the ridge Mc and
at the spring line Ms (see Figure [5]) are the determinant parameters. They depend on the behavior regime of the pipe and, as a
consequence, the analytical formulation varies depending on the
stress state of the control sections.
In the elastic regime (Stage I), Mc is assessed by means the linear
equation (Ec. 6). Once Mc is known, F and Ms are obtained using
Eq. 7 and Eq. 8, respectively. This regime ends when a crack is
formed at ridge for a rotation ccrk (Eq. 9).
Mc = Ecm I
F=
(7)
( )
crk
=
c
2
FRm
(8)
2lbc 1
hEcm
(9)
At Stage II, Mc (Eq. 10) is numerically obtained with the AES model
since the formation of the first crack in C leads to a non linear
system (Eqs. 3-5). Then, F is calculated by means of Eq. 11 and
Ms is deduced by imposing external bending moment equilibrium
(Eq. 12).
( l )
Mc = Mc (N, c ) = Mc 0,
3.3.2 Behavior equations

The governing equations for the structural problem implemented in
MAP were deduced by Pedersen [22] for the simulation of SFRC
pipes with small diameters. The strategy consists in considering that
the response is represented by the three stages previously described,
and that the pipe behaves elastically throughout the whole test, except the section at the ridge and at the spring line. This response
pattern has also been observed in pipes tested by Figueiredo et al.
[13]. The behavior of both sections is simulated with the AES model.
This paper presents the final form of the governing equations for the
problem. Their analytical deduction can be found in Pedersen [22]. They
are based on the energy theorems by Castigliano (see Timoshenko [34])
and on other classical considerations about the calculation of structures.
(6)
Mc
2 Rm
Ms = 1 -
3.3.1 Basic hypotheses

For the simulation of the CT (Figure [1]), the following hypotheses
have been considered: (1) the structure can be idealized as a medium plane piece with a curved shape and a constant radius Rm;
(2) symmetry with regard to the vertical and the horizontal axes,
so only a quarter of the pipe is simulated; (3) the initial curvature
of the piece does not have an influence over the distribution of the
stresses along the piece, nor over its deformed shape; (4) the axial
and shear effects are disregarded in the assessment of the pipe
displacements; (5) the continuous test is considered to be representative for the simulation of the FRCP behavior up to post-failure
(Figueiredo [5]); and (6) three stages are considered:
n Stage 1: linear elastic stage (Figure [5a]).
n Stage 2: elastic stage with cracking at ridge (point C in Figure [5b]).
n Stage 3: elastic stage with cracking at ridge and springs (point
S in Figure [5c]).
l bc
F=
cb
Ecm I
Mc + 2 ( c - crk
c )
2Rm
2Rm
( l )
Ms = FRm - Mc 0,
cb
(10)
(11)
(12)
Stage III (Figure [5c]) starts when a crack is formed in the section
S. This crack appears when a rotation scrk (Eq. 13) is reached. At
this stage, F is calculated by means of Eq. 14, Mc (Eq. 10) is nu-
17
merically obtained with the AES model and Ms is deduced imposing external bending moment equilibrium (Eq. 15).
2lcb F
+ 1
hEcm h
crk
=
s
(13)
Ecm I
crk
F=
Mc + 2 ( c - crk
c - s + s )
2Rm
2Rm
(14)
Ms F,
(15)
( )
( )
s
c
= FRm - Mc 0,
lcb
lcb
vce =
R2m
1
Mc - FRm
Ecm I
2
(16)
if c = c crk
R2m
1
1
Mc - FRm + Rm c - crk
if c >c crk
c
Ecm I
2
2
(17)
The displacement use is deduced by means of Eqs. 18 and 19.
ues
[ ( )
( )]
R2m
crk
=
Mc - 1 + FRm - 1 if c c
Ecm I
2
4
[ (2 - 1) + FR (4 - 1)] + 12 R ( - ) if >
R2m
e
us = Mc
Ecm I
m c
crk
c
crk
(18)
(19)
The displacements vcp and usp are expressed by means of Eqs.

20-22.
p
p
vc = us = 0 if c c crk
18
1
2
1
p p 1
crk
crk
crk
vc = us = Rm (c - crk
c ) + Rm (s - s ) if c > c y s > s
2
2
(21)
(22)
Relevant results are obtained with application of the MAP model,

including the curves Fvc from the CT. These curves allow understanding the behavior of the structure at each of the stages.
The vertical displacement at the ridge (vc) and the horizontal displacement at the springline (us) (see Figure [5d]) are calculated as
a composition of the elastic strain of the pipe (vce and use) and a
plastic one (vcp and usp) due to the rotation of the rigid body after
the appearance of cracks in the critical sections.
The displacement vce is calculated by means of Eq. 16 and Eq. 17.
vce =
crk
crk
vc = us = Rm (c - crk
c ) if c > c y s s
(20)
3.3.3 Solution procedure

The process is initiated with zero values for the rotations at the
ridge (c) and spring line (s). The control variable is c, which
increases with variable steps depending on the behavior stage.
Establishing N = 0 (simple bending) for each value of c at the
ridge, the value of Mc is obtained by means of the AES model (Eq.6
for elastic regime, State I, and Eq. 10 for cracked regime, States II
and III). After that, the values of Ms and F are calculated with the
expressions previously presented:
n At Stage I: F is obtained with Eq. 7 and Ms, by means of Eq. 8.
At this stage the whole pipe, even the critical sections, works in
a linear regime.
n At Stage II: F is obtained with Eq. 11 and Ms, by means of Eq.
12. At this stage, a degree of non-linearity is introduced due
to the cracking of section C (c > ccrk), whereas section S still
works in a linear regime (s scrk). Therefore, due to the degree of hyperstatism of the system, a redistribution of moments
from C to S takes place.
n At Stage III: F is obtained with Eq. 14 and Ms, by resolving
Eq. 15. At this stage, unlike at the previous ones, the balance
condition (Eq. 15) is non-linear due to the fact that both section C and section S have cracked. For its solution, an iterative
Newton Raphson schema was implemented (see Yang et al.
[33]). In this sense, it has to be noticed that the section S works
under a bending - compression state during the whole load
process (N = F and M = Ms).
The algorithm stops either when the maximum strain is reached
at any of the two critical sections, or when the displacement vc
exceeds the fixed value pre-established by the user vc,max.
Once the F and M values have been obtained, the displacements
in sections C and S can be assessed with Eqs. 16-22.
This procedure guarantees good results in concrete pipes with a
predominantly rigid behavior: pipes with a small-medium diameter (300 - 1000 mm) and with moderate reinforcement densities. With these hypotheses, it can be guaranteed that, in most
cases, the cracks are concentrated in sections C and S, while
the rest of the pipe works with its entire section (de la Fuente
et al. [3]). In the opposite case, cracks appear intermediately
and the model deviates from the experimental results and it is
necessary to resort to other models capable of considering the
distributed cracking, such as the one presented in de la Fuente
et al. [12].
4. Contrasting experimental

and numerical results
4.2 Results obtained
With the aim of verifying the suitability of the MAP model for the
simulation of the mechanical response of FRCP subjected to CT,
the experimental results presented in the first part of this work (see
Figueiredo et al. [19]) were contrasted. So, a comparison of the
curves F-vc captured during the test for the pipes with Di of 600
mm from series 2 (displacement measured in the spigot) is done.
All the pipes in these series of tests were manufactured with the same
concrete composition, although there were some modifications in the
water consumption in order to improve the workability due to the use
of fibres (DRAMIX RC-80/60-BN). The pipes were manufactured
and tested at the same age with the aim of reaching, at least, the
resistance class EA2 established in NBR 8890:2007 [20].
4.1 Modeling the materials

The modeling of the compressive behavior of SFRC was performed
with the equation suggested by Barros et al. [28], considering a characteristic compressive strength (fck) of 50 MPa at 28 days according
to the tests performed during the regular quality control in the factory.
On the other hand, for the simulation of its tensile response, the
trilinear diagram proposed by Vandewalle et al. [16] was used (see
Figure [3b]). However, due to the lack of flexural tests (see Vandewalle et al. [35]), in order to determine the values of concrete tension stress i, the expressions (Eqs. 23-24) calibrated in Barros
et al. [36] have been used to determine the values of the residual
flexural strength fRi as a function of Cf. In this regard, the type of
fibres used both in this campaign and in the one carried out by
Barros et al. [36] are the same: (DRAMIX RC-80/60-BN). Table
[1] shows the values established for i, i and Ecm in order to model
the tensile behavior of SFRC.
fR,1 = 0.0945Cf + 0.702
(23)
fR,4 = 0.926 fR,1
(24)
Figures 6a, 6b and 6c show the curves F-vc obtained both experimentally (individual and average values) and numerically for pipes
with Cf of 10, 20 and 40 kg/m3, respectively.
With reference to what has been previously explained in section 3
and to the requirements of the EA2 class of the NBR 8890:2007,
the load Fcr must be equal or higher than the load Fc (90 kN), and
Fu must be equal or higher than 135 kN for this type of pipes. Finally, the maximum post-failure load Fmax,pos measured in the curve
F-vc must reach, at least, the 105% of Fc (94.5 kN). In this sense,
since the test was carried out in a continuous manner, it was established that the value of Fmax,pos is associated to a vc of 3 mm (see
Figueiredo et al. [19]), and is called F3mm.
Based on the results presented in Fig. 6a, it is deduced that the
MAP model fits properly to the experimental results for the amount
of 10 kg/m3, particularly in the linear elastic regime and in the
post-failure regime. For the latter, the numerical results tend toward the experimental maximum values for displacements higher
than 3 mm. This might indicate that the values fR,i and/or i of the
constitutive equation of tensioned SFRC (see Fig. 3b) are slightly
higher than the real ones. The values for Fcr, Fu and F3mm obtained
numerically are 98 kN, 114 kN and 88 kN, respectively. Therefore,
neither Fu nor F3mm reach the minimum values stipulated by NBR
8890:2007 for the EA2 class. Thus, it can be stated that, according
to the model, the amount of 10 kg/m3 is not enough to guarantee
that level of requirements.
The results gathered in Fig. 6b, concerning the dosage of 20 kg/
m3, highlight that the simulation by means of the numerical model
guarantees values close to the experimental ones. Still, the model
exceeds the experimental results with displacements higher than
5.5 mm, which can be due to the considered excessive values of
fR,i and/or i, as in the case of pipes with 10 kg/m3 of fibres for this
range of displacements. The values for Fcr, Fu and F3mm obtained
numerically are: 98 kN, 123 kN and 108 kN, respectively. Consequently, according to the model, with 20 kg/m3 the EA2 class
would not be reached, since the load Fu (123 kN) is lower than the
required 135 kN.
Finally, for the pipes with 40 kg/m3 of fibres (Fig. 6c), it can be
noticed that the numerical model adjusts suitably to the average
experimental results at all the stages. In this case, Fcr is 98 kN
and Fu is 156 kN, values higher than those specified for a pipe of
600 mm from the EA2 class (NBR 8890:2007). It should be noted
that, due to the hardening behavior, there is no way to assess the
Table 1 Parameters to simulate the tensile behavior of the SFRC for the pipe of Di = 600 mm
Cf
3
(Kg/m )
fR,1
(MPa)
fR,4
(MPa)
Ecm
(MPa)
s1*
(MPa)
e1
(mm/m)
s2
(MPa)
e2
(mm/m)
s3
(MPa)
e3
(mm/m)
10
20
1.647
2.592
1.525
2.400
37000
37000
4.071
4.071
0.104
0.104
0.741
1.166
0.204
0.204
0.564
0.888
25.000
25.000
40
4.482
4.150
37000
4.071
0.104
2.017
0.204
1.536
25.000
* This value is obtained by using the expression 1 = 0.30*(fck)
2/3
gathered in the EHE-08 [37].
19
Figure 6 F-vc curves for the SFRCP with: a) 10 kg/m3, b) 20 kg/m3 and c) 40 kg/m3 of fibres
20
Table 2 Average experimental and numerical values of F obtained for the 2nd series of tests
Fibre dosage
3
(kg/m )
Exp.
(kN)
Fcr
MAP
(kN)
e
(%)
Exp.
(kN)
Fu
MAP
(kN)
e
(%)
Exp.
(kN)
F3mm
MAP
(kN)
e
(%)
10
20
40
94
93
92
98
98
98
-4.3
-5.4
-6.5
132
142
162
114
123
156
13.6
13.4
3.7
94
116
-
88
108
-
6.4
6.9
-
load Fmin,pos for a displacement vc of 3 mm as in the previous cases.

However, it can be asserted that the load Fmin,pos from these pipes
exceeds the value of 94.5 kN, since the pipes with 20 kg/m3 of
fibres showed values 14.3% higher than this (108 kN) according to
the numerical model.
Table [2] gathers the average experimental and numerical values for
Fc, Fu and F3mm. The parameter is the relative error of the numerical
value with regard to the experimental data. Positive values of indicate
that the experimental data exceeds the numerical one, and vice versa.
With regard to the results illustrated in Table [2], it is deduced that:
n The load Fcr obtained numerically is independent from the Cf,
since tension 1 is related exclusively to the concrete matrix
(see Figure [3b] and Table [1]). Fcr depends exclusively on 1,
h and Di (see de la Fuente et al. [3]). The model tends to overestimate Fcr with regard to the average experimental results
between a 4.3% (pipe with 10 kg/m3) and a 6.5% (pipe with 40
kg/m3). This can be due to the incorporation, during the mixing,
of additional water (the bigger the amount of fibres, the bigger
the quantity added) into the concrete used for the contrasted
pipes (see Figueiredo et al. [19]).
n The model underestimates the maximum load Fu with regard to
the experimental values between a 3.7% (pipe with 40 kg/m3)
and a 13.6% (pipe with 10 kg/m3). For this load level, the matrix
of concrete has already cracked and the fibres work with the
maximum efficiency; then, the reason for this difference in the
results can be due to the fact that the values of i used in the
constitutive equation might be too conservative for low levels of
displacement. In this regard, de la Fuente et al. [38] proves that
the fibres in concrete pipes manufactured with traditional systems (for example, turbo-compression) work oriented towards
the stress flow, practically guaranteeing their maximum efficiency. One of the ways to consider this fact is by using an amplifi-
cation coefficient of the parameters i. So, the use constitutive

equations incorporating this effect, in order to take into account
the preferential orientation of the fibres could be considered as a
good alternative, as suggested by Laranjeira [39].
n As regards the load F3mm, the numerical values obtained are
6.4% (10 kg/m3) and a 6.9% (20 kg/m3) lower in comparison
with the experimental data. This could also be due to the underestimation of the parameters i adopted to model the tensile response of SFRC in this regime of vc.
To sum up, the MAP model adjusts satisfactorily to the experimental results, even considering that the input parameters of the constitutive equation used to model the tension response of SFRC are
calibrated from concretes with a fck ranging between 25 and 30
MPa (see Barros et al. [36]), as opposed to the reported 50 MPa.
Similarly, the equation used does not take into account the effect of
the preferential orientation of the fibres within the wall of the pipe.
Therefore, the MAP model tends to underestimate the experimental results in most of the cases, yet these differences do not exceed
the 13.6%. Nevertheless, the results can be considered a success,
taking into account the multitude of variables involved in the problem, their uncertainty and the difficulties for the direct experimental
determination of some of them.
5. Example of the optimal design

of the amount of fibres
The following example of MAP model application to a pipe with Di of
400 mm, h of 67 mm and a total length of 2500 mm is purposed to
illustrate the methodology of the design of the optimal Cf in SFRCP.
This diameter was chosen for two reasons: (1) because it is within
the range of diameters for which the hypotheses from the MAP
model are appropriate, and (2) because it is a commercial diameter
Table 3 Parameters to simulate the tensile behavior of the SFRC for the pipe of Di = 400 mm
Cf
3
(Kg/m )
fR,1
(MPa)
fR,4
(MPa)
Ecm
(MPa)
s1
(MPa)
e1
(mm/m)
s2
(MPa)
e2
(mm/m)
s3
(MPa)
e3
(mm/m)
0
10
0.702
1.647
0.650
1.525
36000
36000
3.795
3.795
0.105
0.105
0.316
0.741
0.205
0.205
0.241
0.564
25.000
25.000
20
2.592
2.400
36000
3.795
0.105
1.166
0.205
0.888
25.000
30
3.537
3.275
36000
3.795
0.105
1.592
0.205
1.212
25.000
21
Table 4 Design loads for each strength class

for pipes with Di = 400 mm
EA2
EA3
EA4
Fc (kN)
60
90
120
Fu (kN)
90
135
180
Fmin,pos (kN)
63
94.5
126
which lately is losing its market share in favor of the plastic pipes.
Therefore, new alternatives and improvements are necessary for
the concrete pipe to be competitive once again (Violas et al. [1]).
For the analysis, a range of Cf between 0 kg/m3 and 30 kg/m3 was
established. The first amount would correspond to an unreinforced
concrete pipe (UCP), and the second one has been established as
a maximum value because of economic criteria.
Likewise, the requirement of a minimum early compressive
strength is necessary for demolding and to manipulate the pipe. In
order to satisfy this requirement an fck of 45 MPa at age of 28 days
has been assumed in order to carry out the analysis.
Table 3 presents the values for the mechanical parameters used
to simulate the tension behavior of SFRC.
The values of Fc and Fu fixed in NBR 8890:2007 for the strength
classes EA2, EA3 and EA4 for pipes with Di = 400 mm are gathered in the Table 4.
Fig. 7 shows the curves F-vc obtained with the MAP model. From
those curves, it is deduced that:
n The load Fcr obtained in the four cases (96 kN) is enough to
reach the load Fc stipulated in classes EA2 (60 kN) and EA3
(90 kN). However, the value of Fc for class EA4 (120 kN) cannot be reached by means of the addition of metallic fibres only
(stress 1 is independent from Cf).
n The 90 kN established for Fu in class EA2 are exceeded if a
fibre dosage of 10 kg/m3 (93 kN) is used. But, in order to reach
the 135 kN stipulated for class EA3, according to the numerical
model, at least 30 kg/m3 of fibres (136 kN) are required. On the
other hand, the Cf necessary to achieve the 180 kN established
for class EA4 is not under the economically competitive values.
n In order to reach the 63 kN specified for the Fmin,pos of class EA2
and the 94.5 kN of class EA3, 10 kg/m3 (64 kN) and 20 kg/m3
(99 kN) are required, respectively.
Therefore, keeping in mind that the three strength requirements
must be fulfilled simultaneously for a fixed Cf, it is possible to conclude that 10 kg/m3 of fibres would be required to achieve the EA2
class requirements, and 30 kg/m3 would be required for class EA3,
according to the numerical model.
Alternatively, in order to achieve class EA4 with an economically
attractive reinforcement configuration, a composite solution could
be proposed (fibres + bars). On one hand, the load Fc (Fw=0.25mm in
case of SBFRCP) would be reached thanks to the use of fibres.
On the other hand, the steel bars would guarantee higher failure
strength due to their higher efficiency and strategic position within
the section.
Figure 7 F-vc curves for the pipes with Di = 400 mm and different reinforcement configurations
22
Fig. 7 also shows a simulation considering a reinforcement consisting of 10 kg/m3 of fibres and 75/ml of CA60 steel bars. This
strategy leads to an Fw=0.25mm = 125 kN (Fc = 120 kN) and also to
an ultimate failure load Fu strictly equal to 180 kN. Consequently,
class EA4 would be reached, with Fu being in this case the critical
parameter in the design process.
When comparing the curve F-vc obtained SFRCP with 30 kg/m3 of
fibres and the one obtained for the SBFRCP (10 kg/m3 + 75/m)
for values of vc up to 1.2 mm (service range), it can be observed
that the behavior of the former is better. This highlights the fact that
the normative prescription which refers to the load Fc is much more
restrictive for FRCP, if compared with the conventional SBRCP.
While FRCP are not allowed any cracking symptoms (Fcr>Fc), the
SBRCP are allowed to reach cracking with a width of up to 0.25
mm (Fw=0.25mm>Fc). In the opinion of the authors, this criterion restricts the range of application of fibres to this product. Besides, it
is not in accordance with the experimental and numerical results,
especially when it is known that the inclusion of the suitable type
and volume of fibres improves considerably the cracking behavior
of the concrete structures.
6. Conclusions
This paper introduces the MAP model for the analysis of concrete
pipes with mid-low diameters (lower than 1000 mm) and reinforced
with traditional steel rebars and/or steel fibres. The bases for this
model were already introduced by Pedersen [22], but this paper
uses the most recent constitutive equations to simulate the behavior
of SFRC. The obtained degree of correlation between experimental and model results can be considered excellent, having obtained
numerical results with an average relative error of 7.0% in the safe
of safety. To improve this aspect, the constitutive equation of tensioned SFRC could be adjusted taking into account that the fibres
are oriented towards a preferential direction within the wall of the
pipe. In short, the MAP model can be considered to be a suitable
tool for designing the optimal configuration of the reinforcement for
this type of pipes. It is intended to be used in precast plants where
the required technology to manufacture and test pipes is available.
It leads to savings as regards both time and economical resources,
since it avoids the extensive test programs required in order to find
the optimal amount of reinforcement. This model is especially interesting when a geometrical condition of the pipe, the type of fibres or
the strength class is modified, or simply if the factory wishes to make
the design tables of the most commercial pipes.
With the aim of illustrating the model capability, an example of
optimal design for a pipe with a diameter of 400 mm has been
included. The conclusions established were that, according to the
model, class EA2 would be obtained with 10 kg/m3 of fibres; class
EA3, with 30 kg/m3; and with 10 kg/m3 + 75/m of CA60 steel,
class EA4 from the NBR 8890:2007 would be also reached.
Nowadays, several experimental campaigns are being carried out
with the purpose of expanding the data bank used to contrast the
model and adjust, if necessary, the convenient bases or parameters.
7. Acknowledgements
The authors of this document wish to express their appreciation for the
financial support received through the Research Project BIA2010-17478:
Procesos constructivos mediante hormigones reforzados con fibras.
Likewise, Professor Antonio D. de Figueiredo wishes to thank the

support provided by CAPES -Coordenao de Aperfeioamento
de Pessoal de Nvel Superior for having awarded him the postdoctoral grant that allowed him to participate in this work.
8. References

[01]

[02]

[03]

[04]

[05]

[06]

[07]

[08]

[09]

[10]

[11]

[12]

[13]
VIOLAS, V., AGUADO, A., JOSA, A. Evaluacin de

la sostenibilidad en tuberas de saneamiento.
II Congreso UPC Sostenible 2015, Barcelona, Spain,
2009.
HAKTANIR, T., ARI, K., ALTUN, F., KARAHAN, O. A
comparative experimental investigation of concrete,
reinforced-concrete and steel-concrete pipes under
three-edge-bearing test. Construction and Building
Materials, 2007, vol. 21, n 8, p. 1702-1708.
DE LA FUENTE, A., ARMENGOU, J. Aplicaciones
estructurales del HRFA: Tubos de saneamiento,
paneles de cerramiento y placas de suelo reforzado.
Aplicaciones estructurales del HRFA, Jornada
Tcnica 2007-JT-02, 9 de Octubre de 2007,
Barcelona: Departamento de Ingeniera de la
Construccin, ETS Ingenieros de Caminos, Canales y
Puertos, UPC, 2007.
DE LA FUENTE, A., LARANJEIRA, F., AGUADO,
A., MOLINS, C. Structural applications of SFRC.
Numerical model for sewer pipes, 2nd National
Congress of precast concrete. Centro de Congresos
do LNEC, Lisboa, Portugal, 2008.
FIGUEIREDO, A.D. de. Evaluation of the test method
for crushing strength of steel fibre reinforced concrete
pipes. 7th International RILEM Symposium on Fibre
Reinforced Concrete, Chennai, India, 2008.
FIGUEIREDO, A.D. de, CHAMA NETO, P.J. Avaliao
de desempenho mecnico de tubos. Revista DAE,
2008, Vol. 178, p. 34-39.
LAMBRETCHS, A. Performance clases for steel
fibre reinforced concrete: Be critical, 7th International
RILEM Symposium on Fibre Reinforced Concrete,
Chennai, India, 2008.
ASAD, S., SAXER, A. Influence of Fibre Geometry
on the Flexural Strength Performance of Steel Fibre
Reinforced Concrete (SFRC), Fibre Concrete 2007,
Prague, Czech Republic, 2007.
BLANCO, A. Durability of steel fibre reinforced
concrete. Minor Thesis, UPC, Barcelona, Espaa,
2008.
CHIAIA, B., FANTILLI, A.P. VALLINI, P. Evaluation of
crack width in FRC structures and application to tunnel
linings. RILEM Materials and Structures, 2009,
Vol. 42, n 3, p. 339-351.
BLANCO, A., PUJADAS, P., de la FUENTE, A.,
AGUADO, A. Comparative analysis of constitutive
models of fibre reinforced concrete. Hormign y Acero,
2010, Vol. 61, n 256, p. 83-100.
DE LA FUENTE, A., AGUADO, A., MOLINS, C.
Integral optimum design of concrete pipes. Hormign
y Acero, 2010, Vol. 61, n 259. [In press].
FIGUEIREDO, A.D. de, de la FUENTE, A., AGUADO,
23

A., MOLINS, C.; VIOLAS, B. Anlise da viabilidade

do uso de fibras metlicas em tubos de concreto.

Parte 1: campanha experimental, 52 Congresso

Brasileiro do Concreto, IBRACON, Fortaleza, 2010.

[14] PARROT, J. Sustainability research in sewage pipes,

Minor Thesis, UPC, Barcelona, Espaa, 2009.

[15] HILLERBORG, A., MODER, M., Petersson, P.E.

Analysis of crack formation and crack growth in

concrete by means of fracture mechanics and finite

elements. Cement and Concrete Research, 1976,

Vol. 6, pp. 773-782.

[16] VANDEWALLE, L. et al. Test and design methods for

steel fibre reinforced concrete. - design method.

RILEM Materials and Structures, 2003, vol. 36,
p. 560-567.

[17] LARANJEIRA, F., AGUADO, A., MOLINS, C. Equao

constitutiva de beto reforado com fibras,

2 Congresso Nacional de Prefabricao em Beto,

Lisboa, Portugal, 2008.

[18] PEDERSEN, C. Fibre reinforced concrete pipes.

UNICON beton I/S. December 1992.

[19] FIGUEIREDO, A.D. de, et al. Steel fibre reinforced

concrete pipes. Part 1: technological analysis of the

mechanical behavior. Revista RIEM, 2011, no prelo.

[20] ASSOCIAO BRASILEIRA DE NORMAS

TCNICAS. Tubo de concreto de seo circular, para

guas puviais e esgotos sanitarios. NBR 8890,

ABNT, Rio de Janeiro. 2007.

[21] DE LA FUENTE, A., AGUADO, A., MOLINS, C.

Numerical model for the non linear analysis of precast

and sequentially constructed sections. Hormign y

Acero, 2008, Vol. 57, n 247, p. 69-87.

[22] PEDERSEN, C. Calculation of FRC pipes based

on the fictius crack model. Department of Structural

Engineering. Technical University of Denmark, 1995.

[23] MAR, A., BAIRN J. Evaluacin de los efectos

estructurales del deterioro, reparacin y refuerzo,

mediante anlisis no lineal evolutivo. Hormign y

Acero, 2009, Vol. 60, n 254, p. 51-63.

[24] HEGER, F.J. A theory for the structural behavior of

reinforced concrete pipes, PhD Thesis, Department

of Civil and Sanitary Engineering, Massachusetts

Institute of Technology, MIT, Massachusetts,

USA, 1962.

[25] THORENFELDT, E., TOMASZEWICZ, A., JENSEN

J.J. Mechanical properties of highstrength concrete

and application in design, Proceedings of the

Symposium Utilization of High Strength Concrete,

Stavanger, Norway, 1987.

[26] COLLINS M.P.; MITCHELL, D. Prestressed concrete

basics. Canadian Prestressed Institute, Ontario

(Canada), 1987.

[27] BENCARDINO, F.; RIZZUTI, L.; SPADEA, G.;

SWAMY, R.N. Stress-strain behavior of steel

fibre-reinforced concrete in compression. ASCE J.

of Materials in Civil Enineering 2008, 20(3):255-63.

[28] BARROS, J.A.O., FIGUEIRAS, J.A. Flexural behavior

of SFRC: Testing and modelling. ASCE Journal of

Materials in Civil Enineering, 1999, Vol. 11, n 4, p. 331-339.
24

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

PUJADAS, P. Durability of polypropilene fibre reinforced

concrete. Minor Thesis, UPC, Barcelona, Espaa, 2008.
PEDERSEN, C. The moment-rotation relationship with
implementation of stress-crack width relationships.
Department of Structural Engineering. Technical
University of Denmark, 1995.
OLESEN, J.F. Fictious crack propagation in
fibre-reinforced concrete beams. J. of Engineering
Mechanics, 2001, Vol. 127, n 3, p. 272-280.
CASANOVA, P., ROSSI, P. Analysis and design of
steel fibre reinforced concrete beams. ACI Structural
J., 1997, Vol. 94, n 5, p. 595-602.
YANG W.Y., WENWU C., CHUNG T.S., MORRIS
J. Applied numerical methods using Matlab. John
Wiley & Sons Inc., Hoboken, New Jersey, 2005.
TIMOSHENKO, S. Strength of Materials: Part 1:
Elementary Theory and Problems. D. Van Nostrand
Inc., New York, N.Y. 1940.
VANDEWALLE, L. et al., Recommendations of RILEM
TC162-TDF: Test and design methods for steel fibre
reinforced concrete: Bending test (final
recommendation). Materials and Structures, 2002,
Vol. 35, p. 579-582.
BARROS, J.A.O., CUNHA, V.M.C.F., RIBEIRO,
A.F., ANTUNES, J.A.B. Post-cracking behavior of
steel fibre reinforced concrete. Materials and
Structures, 2005, Vol. 38, p. 47-56.
COMISIN PERMANENTE DEL HORMIGN
(Ministerio de Fomento). EHE 2008 Instruccin del
Hormign Estructural, 2008.
DE LA FUENTE, A., AGUADO, A., MOLINS, C.
Aplicaciones del HRFA: Tuberas de hormign.
IV Congreso de la asociacin cientfico tcnica
del hormign estrutural, Valencia, Espaa, 2008.
LARANJEIRA, F. Design-oriented constitutive model
for steel fibre reinforced concrete. PhD Thesis, UPC,
Barcelona, Espaa, 2010.
9. Nomenclature
A:
Ac:
As,i:
B:
dAc:
Cf:
Di:
Do:
Ecm:
Es:
F:
Fc:
Fcr:
Fn:
Fmax,pos:
Fmin,pos:
Fu:
F3mm:
Pipe spigot.
Concrete area.
Area of the i-th steel bar.
Pipe socket.
Differential concrete area.
Fibre dosage.
Internal diameter of the pipe.
Outside diameter of the pipe.
Average Young modulus of the concrete.
Young modulus of the steel.
Applied load on the pipe.
Proof load of the pipe.
First cracking load of the pipe.
Minimum failure load (established) of the pipe.
Maximum post-failure load (simulated) of the pipe.
Minimum post-failure load (established) of the pipe.
Ultimate failure load of the pipe.
Post-failure load for a 3.0 mm vertical displacement of the
key (2st series of the tested pipes).
fck:
fcm:
fctm,fl:
fRi:
fs,max:
fy:
h:
I:
kh:
l:
lbc:
M:
Mc:
Ms:
M:
N:
N:
ns:
Rm:
sn:
t:
to:
uS:
use:
usp:
vc:
vce:
vcp:
vc,max:
w:
Characteristic compressive concrete strength.

Average compressive concrete strength.
Average flexural tensile concrete strength.
Residual flexural concrete strength.
Maximum steel strength.
Yielding strength of the steel.
Thickness of the concrete wall.
Moment of inertia around the horizontal axis (I = h3/12).
Size factor.
Length of the pipe.
Length of the hinge.
Applied bending moment.
Bending moment at ridge.
Bending moment at springline.
Bending moment at the angular coordinate .
Applied normal force.
Normal force at the angular coordinate .
Number of steel bars.
Average radius of the pipe.
Height of the crack.
Instant of calculus.
Age of loading.
Horizontal displacement of the springline.
Elastic horizontal displacement of the springline.
Plastic horizontal displacement of the springline.
Vertical displacement of the ridge.
Elastic vertical displacement of the ridge.
Plastic vertical displacement of the ridge.
Maximum vertical displacement of the ridge.
Crack width.
xn:
y:
yc:
yG:
ys,i:
:
:
c:
i:
cm:
cnm:
s:
c,sup:
c,inf:
s,i:
y:
:
:
c:
i:
s:
s,i:
u:
:
c:
ccrk:
s:
scrk:
:
c:
Neutral axis depth.

Height of the analyzed element.
Height of the analyzed concrete layer.
Height of the gravity center of the section.
Height of the analyzed i-th steel bar.
Angle between the supports and the centre of the pipe.
Strain.
Concrete strain.
Tensile strain of the SFRC.
Mechanical concrete strain.
Non-mechanical concrete strain.
Steel strain.
Strain of the concrete section top layer.
Strain of the concrete section bottom layer.
Strain of the i-th steel bar.
Yielding steel strain.
Angular coordinate.
Relative error of the numerical value.
Concrete stress.
Steel fibre concrete tension stress.
Steel stress.
I-th steel bar stress.
Ultimate concrete compressive stress.
Angle between the crack faces.
Rotation of the ridge.
Cracking rotation of the ridge.
Rotation of the springline.
Cracking rotation of the springline.
Sectional curvature.
Curvature of the section of the ridge.
25

Sobre o emprego de modelos simplificados de dano
na anlise de estruturas em concreto armado
J. J. C. PITUBA a
julio.pituba@pq.cnpq.br
M. M. S. LACERDA b
maiza_mz@hotmail.com
Abstract
This work presents one and two-dimensional numerical analyses using isotropic and anisotropic damage models for the concrete in order to discuss the
advantages of these modeling. Initially, it is shortly described the damage model proposed by Mazars. This constitutive model assumes the concrete as
isotropic and elastic material, where locally the damage is due to extensions. On the other hand, the damage model proposed by Pituba, the material
is assumed as initial elastic isotropic medium presenting anisotropy, plastic strains and bimodular response (distinct elastic responses whether tension
or compression stress states prevail) induced by the damage. To take into account for bimodularity two damage tensors governing the rigidity in tension
and compression regimes, respectively, are introduced. Damage activation is expressed by two criteria indicating the initial and further evolution of damage. Soon after, the models are used in numerical analyses of the mechanical behavior of reinforced concrete structures. Accordingly with comparison
of the obtained responses, considerations about the application of the isotropic and anisotropic damage models are presented for 1D and 2D reinforced
concrete structures modeling as well as the potentialities of the simplified versions of damage models applied in situations of structural engineering.
Keywords: damage mechanics, reinforced concrete, constitutive model.
Resumo
Este artigo apresenta anlises uni e bidimensionais utilizando modelos de dano istropo e anistropo para o concreto com o objetivo de discutir
sobre as vantagens deste tipo de modelagem. Inicialmente, o modelo de dano proposto por Mazars brevemente descrito, onde o mesmo considera o concreto como meio istropo e elstico, sendo a danificao decorrente de extenses em nvel local. Por outro lado, o modelo proposto
por Pituba admite o concreto como meio inicialmente istropo e que passa a apresentar deformaes plsticas, bimodularidade e anisotropia induzidas pelo dano. Para levar em conta a bimodularidade, dois tensores de dano governando a rigidez em trao e em compresso so introduzidos. Em seguida, os modelos constitutivos so utilizados em anlises numricas do comportamento mecnico de estruturas em concreto armado.
De acordo com as respostas obtidas, discutem-se algumas consideraes sobre a aplicabilidade de modelos de dano istropos e anistropos na
modelagem de estruturas em concreto armado no mbito de anlises uni e bidimensionais, assim como so apresentadas as potencialidades de
aplicao de verses simplificadas de modelos de dano em situaes da Engenharia Estrutural.
Palavras-chave: mecnica do dano, concreto armado, modelo constitutivo.
Civil Engineering Department, Federal University of Goias, Av. Dr. Lamartine Pinto de Avelar, 1120, 75740-020, Catalo, Gois, Brasil.
E-mail: julio.pituba@pq.cnpq.br
b
Civil Engineering Department, Federal University of Goias, Av. Dr. Lamartine Pinto de Avelar, 1120, 75740-020, Catalo, Gois, Brasil.
E-mail: maiza_mz@hotmail.com)
a
Received: 26 May 2011 Accepted: 21 Oct 2011 Available Online: 01 Feb 2012
2012 IBRACON
J. J. C. PITUBA | M. M. S. LACERDA
1. Introduction
2.
In the Continuum Damage Mechanics (CDM), the damage effects

are evidenced in the stiffness constitutive tensor. The damage
leads to the reduction of several stiffness components, where the
damaged material can either keep its isotropic properties or to become anisotropic.
For isotropic models the damage affects neither the direction nor
the initial number of symmetry planes presented by the material.
Thus, it does not matter if the medium is initially isotropic or anisotropic with some degrees. In this case, those initial characteristics
are preserved during the damage process. Some constitutive damage models have been proposed assuming the concrete as an isotropic medium (Mazars [1], Mazars [2], Comi [3] and Berthaud [4]).
However, in the last decades anisotropic models which can modify
both the direction and the number of the material symmetry planes
have been proposed (Brnig [5], Pituba [6], Pietruszczak [7], Ibrahimbegovic [8] and Dragon [9]).
Besides, another important characteristic presented by many fiberreinforced composite materials is the intrinsic bimodularity, i.e.,
distinct responses in tension and compression prevailing states.
On the other hand, brittle materials, such as concrete, are a kind of
composites that can be initially considered isotropic and unimodular. However when they have been damaged, those materials
would start to present some degree of anisotropy and bimodularity. Assuming small deformations, a formulation of constitutive laws
for either initially isotropic or anisotropic elastic bimodular materials was proposed by Curnier [10]. In order to incorporate damage
effects, the formulation of Curnier has been extended by Pituba
[11]. In particular, a constitutive model for concrete has been derived. Accordingly, the material is initially considered as an isotropic continuous medium with anisotropy and bimodularity induced
by the damage. On one side the class of anisotropy induced and
considered in the model (transversal isotropy) elapses from the
assumption that locally the loaded concrete always presents a diffuse oriented damage distribution as appointed by experimental
observations (Van Mier [12]). On the other hand, the bimodularity
induced by damage is captured by the definition of two damage
tensors: one for dominant tension states and another one for dominant compression states.
In order to use the Damage Mechanics in practical situations of
Structural Engineering, constitutive models presenting a reduced
number of parameters with easy identification and one-dimensional version are desirable. On the other hand, such models must
present reliable numerical results in order to estimate the mechanical behavior of the structure as accurate as possible. In this work,
two examples of this kind of the damage models are used and
here, they are called simplified damage models.
This work intends to discuss the related problems to the numerical applications of the isotropic and anisotropic damage models in
the context of the one and two-dimensional analyses of reinforced
concrete structures. Besides, one intends to show the potentialities
of an anisotropic damage model recently proposed by Pituba [6].
Then, numerical responses supplied by the models are presented
and compared in order to evidence the difficulties and advantages
when one deals with this kind of modeling. Finally, some conclusions about the employment of the simplified versions of these
kinds of damage models are discussed.
This model has been proposed by Mazars [1] and the damage
is represented by the scalar variable D (with 0 D 1) whose
~
evolution occurs when the equivalent extension deformation is
bigger than a reference value. The plastic deformations evidenced
experimentally are not considered. The equivalent extension deformation is given by:
Isotropic Damage Model
e~ =
where
e1
2
+
+ e2
+ e3
(1)
2
+
is a principal deformation component, being
1
[ i + i ].
2
~
The damage activation occurs when = e
positive part, i. e.:
its
, being ed0 the deford0

mation referred to the maximum stress of an uniaxial tension test.
Thus the criterion is given by:
f e , D = e - S D 0 with S(0) = ed0
(2)
Considering the thermodynamics principles, the damage evolution

can be expressed by:
.
D = 0 if
.
f < 0 or f = 0 and f < 0
.
.
D = F (e~ ) e~
if
.
f = 0 and f = 0
(3a)
(3b)
(~ )
where D = dD/dt , i. e., D time derivative; F is written in

~
terms of and defined continuous and positive.
As the concrete behaves differently in tension and compression, the
damage variable D is obtained by combining properly the variables DT
and DC, related to tension and compression, respectively, as follows:
.
D = a T DT + a C DC where a T + a C = 1
(4)
where DT and DC are given by:
DT = 1 -
e d 0 (1 - AT )
AT
~
e
exp[BT (e~ - e d 0 )]
(5a)
27
DC = 1 -
e d 0 (1 - AC )
AC
~
e
exp[BC (e~ - e d 0 )]
(5b)
In Eqs (5) AT and BT are parameters related to uniaxial tension tests

while AC and BC are obtained from uniaxial compression tests. To
compute the aT and aC values defined in Eq.(4), we have to obtain,
initially, the deformations eT and eC associated, respectively, to tension and compression states as follows:
eT =
1 +u *
s
E
eC =
1 +u *
s
E
u
E
s
i
*
i
u
E
*
i
(6a)
(6b)
where I is the identity tensor, E the elastic modulus of a non-dam*

*
aged material,
and
are, respectively, positive and
+
*
negative parts of the stress tensor obtained from the relation
* = D0 , where D0 is the elastic fourth order tensor of the
non-damaged material.
Thus the coefficients T and C are obtained by the following
expression:
aT =
aC =
e
i
Ti
+
V
e
i
Ci
+
V
(7a)
(7b)
where e Ti
and e Ci
are, respectively, positive and nega+
+
tive parts of the deformations eT and eC defined in Eq. (6); V is
+
given by: e V = i e Ti + + e Ci + .
Finally, the constitutive relation can be expressed in terms of the
actual deformation tensor as follows:
s = (1 - D )D0e
(8)
3. Anisotropic Damage Model

In this model, it is assumed that the concrete belongs to a category
of materials that can be considered initially isotropic and unimodular, however they start to present different behaviours in tension
28
and compression when damaged. This model has been proposed

by Pituba [6] and it follows the from the formalism presented in
Pituba [11]. Moreover, the model respects the principle of energy
equivalence between damaged real medium and equivalent continuous medium established in the CDM (Lemaitre [13]).
Now, the damage model is shortly described. Initially, for dominant
tension states, a damage tensor is given by:
DT = f1(D1, D4, D5) ( A A) + 2 f2(D4, D5) [( AI + I A) - ( A A)]
(9)
where f1(D1, D4, D5) = D1 2 f2(D4, D5) and f2(D4, D5) = 1 (1-D4)
(1-D5).
The variable D1 represents the damage in direction orthogonal to
the transverse isotropy local plane of the material, while D4 is representative of the damage due to the sliding movement between
the crack faces. The third damage variable, D5, is only activated
if a previous compression state accompanied by damage has
occurred. In the Eq. (9), the tensor I is the second-order identity
tensor and the tensor A, by definition, Curnier [10], is formed by
dyadic product of the unit vector perpendicular to the transverse
isotropy plane for himself.
On the other hand, for dominant compression states, it is proposed
the other damage tensor:
DC=f1(D2,D4,D5) ( A A) +f2(D3) [( I I ) - ( A A)]

+2f3(D4,D5) [( AI + I A) - ( A A)]
(10)
where f1(D2, D4, D5) = D2 2 f3(D4, D5) ,f2(D3) = D3 and f3(D4, D5)=
1 (1-D4) (1-D5).
Note that the compression damage tensor introduces two additional scalar variables in its composition: D2 and D3. The variable
D2 (damage perpendicular to the transverse isotropy local plane
of the material) reduces the Youngs modulus in that direction and
in conjunction to D3 (that represents the damage in the transverse
isotropy plane) degrades the Poissons ratio throughout the perpendicular planes to the one of transverse isotropy.
Finally, the constitutive tensors are written as follows:
ET = l11 [ I I ] + 2 m1 [ I I ] - l+22 ( D1 , D4 , D5 ) [ A A] - l12+ ( D1 )

[AI+IA] - m 2 ( D4 , D5 ) [ AI + I A]
(11a)
E C = l11 [ I I ] + 2 m1 [ I I ] - l-22 ( D2 , D3 , D4 , D5 ) [ A A]
- l12- ( D2 , D3 ) [AI+IA] - l11- ( D3 ) [II] -
(1 - 2n 0 ) l11 ( D3 ) [I I] - m2 ( D4 , D5 ) [ AI + I A]
n0
(11b)
where l11 = l 0 and 1 = 0 are Lam constants. The remaining parameters will only exist for no-null damage, evidencing
in that way the anisotropy and bimodularity induced by damage.

Those parameters are given by:
l+22 (D1 , D4 , D5 ) = (l0 + 2 m0 )(2 D1 - D12 ) - 2 l12+ ( D1 ) - 2 m 2 ( D4 , D5 )
l12+ ( D1 ) = l0 D1 ; m2 (D4 , D5 ) = 2 m0 [1 - (1 - D4 )2 (1 - D5 )2 ]
l -22 ( D2 , D3 , D4 , D5 ) = (l0 + 2 m 0 )(2 D2 - D22 ) - 2 l12- ( D2 , D3 )
(n - 1 )
+ 0 l11- ( D3 ) - 2 m 2 ( D4 , D5 ) ; l11- (D3 ) = l0 (2 D3 - D32 )
n0
(12)
l12- ( D2 , D3 ) = l0 [(1 - D3 ) 2 - (1 - D2 )(1 - D3 )]

The different dyadic products in Eqs. (8), (10) and (11) have the
function of allocating the material constants in certain positions of
the stiffness constitutive tensors. For more details see Curnier [10]
and Pituba [11].
Observe that the bimodular character is taken into account by the
conditions g (, DT, DC) > 0 or g (, DT, DC) < 0, where g (, DT, DC)
is a hypersurface that contains the origin and divides the strain
space into a compression and tension sub-domains. A particular
form is adopted for the hypersurface in the strain space: a hyperplane g(,D) defined by the unit normal N and characterized by its
dependence of both the strain and damage states. To simplify the
presentation, the hyperplane will be here expressed as the one
obtained by enforcing the direction 1 in the strain space to be perpendicular to the transverse isotropy local plane. Thus, the hyperplane is given by:
g(e,DT,DC) = N(DT,DC) . ee = g1(D1,D2) e Ve + g2(D1,D2) e 11e
(13)
where g1(D1,D2) = {1+H(D2)[H(D1)-1]}h(D1)+{1+H(D1)[H(D2)-1]}

h(D2) and g2(D1,D2) = D1+D2. The Heaveside functions employed
above are given by:
H(Di) = 1 for Di > 0; H(Di) = 0 for Di = 0 (i = 1, 2)
(14)
The h(D1) and h(D2) functions are defined, respectively, for the
tension and compression cases, assuming for the first one that
there was no previous damage of compression affecting the present damage variable D1 and analogously, for the second one that
has not had previous damage of tension affecting variable D2. The
functions can be written as:
h (D1) =
- D1 + 3 - 2 D12
- D + 3 - 2 D22
; h (D2) = 2
3
3
(15)
damage induces anisotropy in the concrete. Therefore, it is convenient to separate the damage criteria into two: the first one is
only used to indicate damage beginning, or that the material is
no longer isotropic and the second one is used for loading and
unloading when the material is already considered as transverse
isotropic. This second criterion identifies if there is or not evolution
of the damage variables. That division is justified by the difference
between the complementary elastic strain energies of isotropic and
transverse isotropic material. For identifying the damage beginning
it is suggested a criterion that compares the complementary elastic
strain energy We , which is computed locally considering the medium as initially virgin, isotropic and purely elastic, with a certain
reference value Y0T, or Y0C, obtained from experimental tests of
uniaxial tension, or compression, respectively. Accordingly, the criterion for initial activation of damage processes in tension or compression is given by:
(16)
fT,C (s) = We* - Y0T,0C < 0
then DT = 0 (i. e., D1 = D4 = 0) for dominant tension states or DC =

0 (i. e., D2 = D3 = D5 = 0) for dominant compression states, where
the material is linear elastic and isotropic. The reference values
Y0T and Y0C are model parameters defined by
02T
2E 0
and
02C
2E 0
respectively, where 0T and 0C are the limit elastic stresses determined in the uniaxial tension and compression regimes.
It is important to note that the damaged medium presents a transverse isotropy plane in correspondence to the current damage
level. Then, the complementary elastic energy of the damaged
medium is expressed in different forms, depending on whether
tension or compression strain states prevail. In the case of dominant tension states (g(e,DT,DC) > 0) assuming that direction 1 in
the strain space be perpendicular to the transverse isotropy local
plane, it can be written:
We* + =
(s 222 + s 332 ) n 0 (s11s 22 + s11s 33 ) n 0 s 22 s33

s112
+
2E 0
E 0 (1 - D1 )
E0
2E 0 (1 - D1 ) 2
(1 + n 0 )
(1 + n 0 ) 2
+
(s 2 + s132 ) +
s 23
2
2 12
E0
E 0 (1 - D 4 ) (1 - D 5 )
(17)
For the damaged medium in dominant compression states, the relationships are similar to the tension case, where the complementary elastic energy is expressed in the following form:
We* - =
(s 222 + s 332 ) n 0 (s11s 22 + s11s 33 ) n 0 s 22 s33

s112
+
2E 0 (1 - D 2 ) 2 2E 0 (1 - D 3 ) 2 E 0 (1 - D 2 )(1 - D 3 ) E 0 (1 - D 3 ) 2
(1 + n 0 )
(1 + n )
+
(s 2 + s132 ) + 0 s 223
2
2 12
E0
E 0 (1 - D 4 ) (1 - D 5 )
(18)
As it has already been pointed out, in the model formulation the
29
Assuming a general situation of damaged medium for dominant tension states, the criterion for the identification of
damage increments is represented by the following relationship:
fT(s) = We* + - Y0*T 0
(19)
where the reference value Y0T is defined by the maximum complementary elastic energy computed throughout the damage process up to the current state. Analogous expressions are valid for
dominant compression states.
.
.
In the loading case, i. e., when DT 0 or DC 0 , it is necessary to update the values of the scalar damage variables that
appear in the DT and DC tensors, considering their evolution laws.
Considering just the case of monotonic loading, the evolution laws
proposed for the scalar damage variables are resulting of fittings
on experimental results and present similar characteristics to those
one described in Mazars [1] and Berthaud [4] works. The general
form proposed is:
Di = 1 -
1 + Ai
with i = 1, 2
Ai + exp[Bi (Yi - Y0 i )]
(20)
where A i, B i and Y 0i are parameters that must be identified. The

parameters Y 0i are understood as initial limits for the damage
activation, the same ones used in Eq. (16). The parametric
identification of the model is accomplished by uniaxial tension
tests in order to obtain A 1, B 1 and Y 01 = Y 0T, by uniaxial compression tests for the identification of the parameters A 2, B2
and Y 02, and finally by biaxial compression tests in order to
obtain A 3, B3 and Y 03 = Y02 = Y0C. On the other hand, the identification of the parameters for the evolution laws corresponding
to the damage variables D 4 and D 5, which influence the shear
concrete behavior, it wont be studied in this version of the

model because the experimental tests are not available yet
to allow the parameter calibration or, even, the proposition of
more realistic evolution laws.
When the damage process is activated, the formulation starts
to involve the tensor A that depends on the normal to the transverse isotropy plane. Therefore, it is necessary to establish
some rules to identify its location for an actual strain state. Initially, it is established a general criterion for the existence of
the transverse isotropy plane. It is proposed that the transverse
isotropy due to damage only arises if positive strain rates exist at least in one of the principal directions, Pituba [6]. After
assuming such proposition as valid, some rules to identify its
location must be defined. First of all, considering a strain state
in which one of the strain rates is no-null or has sign contrary to
the others, the following rule is applied:
In the principal strain space, if two of the three strain rates are
extension, shortening or null, the plane defined by them will be
the transverse isotropy local plane of the material.
The uniaxial tension is an example of this case where the
transverse isotropy plane is perpendicular to the tension stress
direction. Obviously, it can be suggested criteria based on others formulations, such as for instance, the microplanes theory
developed by Bazant [14].
Figure 1 Geometry details of

the plain concrete beam
P
All dimensions in metres
Figure 2 Parametric identification in uniaxial compression and tension tests for plain concrete beam
30
Table 1 Parameter values plain concrete beam

Mazars model
Tension
Compression
AT = 0,7
BT = 8000
ed0 = 0,000067
AC = 1,13
BC = 1250
Tension
4.1 Plain Concrete Beam

In this first numerical application, previously performed by
Guello [15], we have considered the concrete beam without
any reinforcement bar described in Figure [1]. The beam whose
concrete has elasticity modulus Ec = 24700 MPa, is subjected to
two concentrated loads P applied at a distance of 0,225 m from
the symmetrical axes.
Table [1] contains the parameter values of both models employed
in this example. The parametric identification of the damage models has been done using a computational code developed by Pituba [11] based on error minimization procedure. The compression
and tension parameters values of the Pitubas model have been
identified by numerical responses proposed by Guello [15] using
the Mazars model, as described in Figure [2].
Moreover, the parameters values associated to D3 have been
identified by numerical simulation of biaxial stress tests in concrete
specimens, Pituba [11]. It is important to note that the concrete
used for the identification of Y03, A3 and B3 parameters is similar to
the one considered in this numerical application.
For the 1D analysis, bar elements with transversal section
stratified in layers are used, see Figure [3] where layer K can
represent concrete or reinforcement bar. A mesh with 40 elements and 20 layers has been considered. On the other hand,
a mesh with 120 constant strain quadrilateral (4 nodes) elements divided into 5 layers of 24 elements and placed in whole
extension of the concrete beam has been considered for the 2D
-4
Y01 = 0,25x10 MPa

A1 = 50
-1
B1 = 6700 MPa
4. Numerical Applications
Figure 3 1D finite element
Pitubas model
Compression
-3
-3
Y02 = 0,5x10 MPa Y03 = 0,5x10 MPa

A2 = -0,9
A3 = -0,6
-1
-1
B2 = 0,4MPa
B3 = 1305MPa
analysis, as described in Figure [4]. Note that in this numerical

example, the layer with finite elements in black representing the
reinforcement does not exist.
The numerical responses are displayed in Figure [5]. In the
2D analysis context, it can be observed that the difference be-
Figure 4 2D finite element discretization
Z
Y
X
Z
Y
X
Figure 5 1D and 2D numerical responses

for plain concrete beam
31
Figure 6 Damage distribution of Mazars' model, Proena [16]
Figure 7 Damage distribution of Pituba's model for 1D and 2D numerical analyses
32
Figure 8 Geometry of the reinforced concrete bar structure
All dimensions in meters
Figure 9 Parametric identification in uniaxial compression and tension tests

for reinforced concrete bar structure
tween the numerical responses is mainly due to the excessive

stiffness reduction presented by isotropic model, what does not
happen with the anisotropic model. Being an isotropic damage
model, the Mazars model degrades the stiffness exaggeratedly in all directions. Besides, one believes that the quality of
results related to the Pitubas model can be better if the damage process related to the shear behavior of the concrete is
considered, because it can produce an important contribution
to the released energy.
Note that in the one-dimensional analysis the difference between numerical responses of damage models is mainly due
to more aggressive reduction of the stiffness in the anisotropic model. While the Mazars model presents a linear reduction way of the material stiffness (Strain Equivalence - Eq.
(8)), the Pitubas model reduces the material stiffness in a
quadratic way (Energy Equivalence - Eq. (11)). In this analysis, it is possible to reproduce a stiffness break about 45 kN.
Soon after, a subsequent strength recovery of the structural
element is illustrated by the models. Then, the difference
between the numerical responses is increased. Finally, note
that in 1D and 2D numerical analyses, both models present
the same qualitative behaviors.
In order to visualize the damage distribution in the beam,
Figure [6] presents the isodamage curves for Mazarsmodel,
Proena [16]. Figure [7] shows the damage distribution for 1D
and 2D numerical analyses with Pitubas model considering a
loading stage about 50 kN. These curves have been obtained
by interpolation of the damage variables along the adopted

integration points.
In the context of 1D analysis, when the Proenas response
is compared with the Pitubas model, the damage configuration is very similar, mainly due to damage in tension. Note
that Mazars model presents only one damage variable representing a combination of tension and compression damage
processes. For the concrete structures, both models assume
that locally the damage is due to extensions. On the other
hand, the 2D numerical responses show a more intense damage process in compression given by D 2 and D 3 variables. In
fact, the variable D 3 tries to simulate a crushing process of
the concrete near of the supports and in upper middle zone
of the beam.
4.2 Reinforced Concrete Bar Structure

The second numerical simulation deals with a reinforced concrete
bar structure. Figure [8] shows the geometry of the structure. A
reinforcement bar with 10.0 mm is placed in the centre of the section where a force P is applied. This numerical application was performed previously by Mazars [1]. Table [2] presents the parameter
values of each damage model. These parameters were obtained
through procedure similar to the previous example with Ec=30000
MPa, where the parameters suggested for the Mazars model were
taken as reference values, see Figure [9]. For the steel, Ea = 210
GPa has been adopted.
33
Table 2 Parameter values reinforced concrete bar structure

Mazars model
Tension
Compression
AT = 0,8
BT = 20000
ed0 = 0,00001
Pitubas model
Compression
Tension
-4
AC = 1,4
BC = 1850
Y01 = 0,25x10 MPa

A1 = 130
-1
B1 = 1000 MPa
Figure 10 1D and 2D numerical responses

for reinforced concrete bar strucuture
In the 1D analysis, a mesh with 20 elements and 10 layers has

been considered. On the other hand, in the 2D analyses performed
by Guello [15] with Mazars model, a mesh with 38 x 16 triangle
finite elements (6 nodes) has been used in order to represent the
concrete, while a mesh with 30 bar elements represents the steel.
On the other hand, in this work, a mesh with 100 constant strain
quadrilateral (4 nodes) elements divided into 10 layers of 10 elements and placed in whole extension of the structural element
has been considered for the 2D analysis with Pitubas model. Note
that, in this mesh, one layer represents the reinforcement bar, see
Figure [4]. The results are described in the Figure [10].
In the 1D analysis, once again the stiffness degradation is observed
in a more evident way in the anisotropic model by the reasons previ-
-3
-3
Y02 = 0,1x10 MPa Y03 = 0,5x10 MPa

A2 = 3
A3 = -0,6
-1
-1
B2 = 11 MPa
B3 = 1305 MPa
ously explained. However, that difference is more evident after the

adherence loss stage between the reinforcement bar and concrete
presented in the Fig. [10] about 28 kN. Note that the 1D models present a stiffness recovery after to total concrete cracking. That new resistance is just owed to the reinforcement bar that presents an elastic
behavior up to yielding resistance be achieved. Those observations
are in agreement with others works (Mazars [1] and Guello [15]).
On the other hand, in the 2D analysis, the anisotropic model has
presented locking problems related to numerical responses due to
the distortion of the finite element that should be partly responsible
for the excessive stiffness, what does not happen with the Mazars
model (triangular element). Therefore, there was not possibility
of more intense stiffness degradation by the anisotropic model.
Note that in this analysis the elastic behavior for reinforcement
bar has been assumed leading to the increase of difference about
numerical responses given by the models, due mainly to the finite
elements. Besides, in the 2D analysis the strain localization is an
important phenomenon in the behavior of concrete structures. In
particular, this structure is mainly tensioned presenting a localized
cracking configuration[15].
4.3 Reinforced Concrete Beam

The third numerical application is about a reinforced concrete beam
submitted to monotonic loading. This beam was tested by Delalibera [17]. The elastic parameters of the concrete are fck=25MPa
and Ec=32.3MPa. For the reinforcement has been adopted Ea=
205 GPa, yielding stress 590 MPa and ultimate stress 750 MPa.
The geometric characteristics of the beam are given in Figure [11].
The loading is composed by two equal forces applied on the beam.
The uniaxial stress tests that were performed by Delalibera [17]
and they were used in order to identify the models parameters, see
Figure 11 Geometry details of the reinforced concrete beam, dimensions in meters

26,3mm
0,3
0,3
212,5mm
220mm
0,15
34
0,1
2,85
0,1
Figure 12 Parametric identification in uniaxial compression and tension tests

for reinforced concrete beam
Uniaxial Tension Test
Table 3 Parameter values reinforced concrete beam

Mazars model
Tension
Compression
AT = 1,71
BT = 11300
ed0 = 0,0000675
AC = 2280
BC = 12800
Figure [12]. The parameters values as described in the Table [3].

Initially, the longitudinal discretization has been composed by 16
finite elements and cross section was divided into 15 layers, where
there are layers that represent the reinforcement areas located in
their barycenters, see Figure [3].
Figure 13 Experimental and 1D numerical

responses of the reinforced concrete beam
Pitubas model
Tension
Compression
Y01 = 1,137x10-4 MPa
A1 = 5,33
-1
B1 = 5660 MPa
-5
Y02 =5,0x10 MPa

A2 = 0,0086
-1
B2 = 5,71 MPa
The numerical responses obtained with isotropic and anisotropic

damage models are able to simulate the experimental behavior of
the beam, see Figure [13]. Both damage models present a strong
loss of stiffness about 18 kN trying to evidence a possible damage
localization process. In this context, in Delalibera [17] is reported
that the first crack occurs about 25 kN. As it can be seen, the
numerical responses are quite satisfactory when compared to the
experimental results since the beginning of the damage process
up to the complete rupture of the beam.
On the other hand, it is noted that the numerical responses present a difficulty of convergence evidenced by the increase of the
iterations about 115 kN. In this stage, the concrete presents high
values for the damage variables in many layers of finite elements
located in the medium zone of the beam, therefore the beam stiffness is mainly due to the reinforcement bars.
5. Conclusions
The study concerns to the employment of anisotropic and isotropic
damage models to one and two-dimensional concrete structure
analyses. In a general way, the numerical results obtained from
the damage models presented in this work have been quite satisfactory. The potentialities of the Damage Mechanics when deals
with numerical simulation of the non-linear behavior of concrete
structures are shown. In particular, the employment of anisotropic models has shown some advantages in 2D analysis when
35
compared to the isotropic ones, such as, the selective stiffness

deterioration and evolution of cracking configuration supplying a
more realistic numerical response, see Pituba [6]. This feature can
be more evident in three-dimensional analyses. However, it must
be observed that structures with low reinforcement rates can evidence some numerical problems due to plane analysis, see Pituba
[6], Pituba [18] and Comi [19]. In these cases, the cracking process
starts to present a localized distribution limiting the employment of
the damage models. In order to overcome numerical problems a
non-local version of the anisotropic model can be proposed and
implemented in a computational code, for instance, with so-called
Generalized Finite Element Method.
On the other hand, it is important to observe that the proposal
and parametric identification of evolution laws for damage variables D4 and D5 must increase the accuracy of the anisotropic
model. In fact, these cracking processes related to shear behavior of the concrete are significant contributions to the released
energy. This feature has been studied by Pituba [20] and a theoretical analysis has shown that the anisotropic model has advantage upon constitutive models that use the so called shear
retention factor.
The 1D analysis has shown an efficient and practical employment,
without numerical problems and low computational cost. Besides,
the parametric identification is simple. In this case, the anisotropic
or isotropic damage models could be used in estimative analyses
of structures in practical situations, such as: numerical simulation
of displacement in cracking concrete beams in order to propose an
alternative procedure to the Brazilian Technical Code (Pituba [21]),
estimative of ultimate load capacity of frames and beams and collapse configuration of reinforced concrete frames (Pituba [22] and
Pituba [23]), included the numerical analyses of the structures submitted to cyclic loading (Pituba [24]) . Finally, this work has demonstrated that simplified damage models are a good alternative to
estimate the mechanical behavior of reinforced concrete structures.
6. Acknowledgments
The authors wish to thank to CNPq (National Council for Scientific
and Technological Development) for the financial support.
7. References

[02]

[03]

[04]

[05]

36
[01] MAZARS, J. and PIJAUDIER-CABOT,

G. Continuum damage theory application to
concrete. Journal of Engineering Mechanics, 1989,
115, No. 2, 345-365.
MAZARS, J., BERTHAUD, Y. and RAMTANI, S. The
Unilateral Behaviour of Damaged Concrete,
Engineering Fracture Mechanics, 1990, 35, 629-635.
COMI, C. and PEREGO, U. Fracture energy based
bi-dissipative damage model for concrete,
International Journal of Solids and Structures, 2001,
38, 6427-6454.
BERTHAUD, Y., LA BORDERIE, C. and RAMTANI,
S. Damage modeling and cracking closure effect.
Damage Mechanics in Engineering Materials, 1990,
24, 263-273.
BRNIG, M. An anisotropic continuum damage
model: theory and numerical analyses. Latin American

Journal of Solids and Structures, 2004, 1, 185-218.

[06] PITUBA, J. J. C. and FERNANDES, G. R.

Anisotropic damage model for concrete, Journal of

Engineering Mechanics, 2011, No. 9, 137, 610-624.

[07] PIETRUSZCZAK, S. and MROZ, Z. On failure

criteria for anisotropic cohesive-frictional materials.

International Journal for Numerical and Analytical

Methods in Geomechanics, 2001, 25, 509-524.

[08] IBRAHIMBEGOVIC, A., JEHEL, P. and DAVENNE,

L. Coupled damage-plasticity constitutive model and

direct stress interpolation, Computational Mechanics,

2008, 42, 1-11.

[09] DRAGON, A., HALM, D. and DSOYER, Th.

Anisotropic damage in quasi-brittle solids: modelling,

computational issues and applications. Computer

Methods in Applied Mechanics and Engineering, 2000,

183, 331-352.

[10] CURNIER, A., HE, Q. and ZYSSET, P. Conewise

linear elastic materials. Journal of Elasticity, 1995,

37, 1-38.

[11] PITUBA, J. J. C. On the Formulation of Damage Model

for the Concrete (in portuguese), PhD thesis,

University of Sao Paulo, So Carlos-Brazil, 2003.

[12] VAN MIER, G. M. Fracture Process of Concrete.

CRC Press, 1997.

[13] LEMAITRE, J. A Course on Damage Mechanics.

Springer Verlag, 1996.

[14] BAZANT, Z. P. and OZBOLT, J. Nonlocal microplane

model for fracture, damage and size effect in

structures, Journal of Engineering Mechanics, ASCE,

1990, 116, 2485-2505.

[15] GUELLO, G. A. Computational Simulation of Concrete

Structures through Damage Mechanics (in

portuguese), Masters Thesis. University of Sao

Paulo, So Paulo-Brazil, 2002.

[16] PROENA, S. P. B., PAPA, E. and MAIER, G.

Meccanica Del Danneggiamento di Materiali e

Strutture: Aplicazioni al Calcestruzzo ( in italian).

Rapporto sul contratto di ricersa 90-91. Departimento

di Ingegneria Strutturale Politecnico, Milano, 1991.

[17] DELALIBERA, R. G., Theorical and experimental

analysis of reinforced concrete beams with

confinement reinforcement. Masters Thesis,

University of Sao Paulo, So Paulo, Brazil, (2002).
In portuguese.

[18] PITUBA, J. J. C. Formulation of damage models

for bimodular and anisotropic media. Revista

Sul-Americana de Engenharia Estrutural, 2006,

3, 7-29.

[19] COMI, C. A nonlocal damage model with permanent

strains for quasi-brittle materials. Proceedings of the

Continuous Damage and Fracture, Cachan-France,

2000, p. 221-232.

[20] PITUBA, J. J. Evaluation of an anisotropic damage

model taking into account the effects of resistence loss

due to shear. In: XXXI Iberian Latin American

Congress on Computational Methods in Engineering,

2010, Buenos Aires. Mecnica computacional.

Buenos Aires, 2010, v. XXiX. p. 5397-5410.
[21] PITUBA, J. J. C., DELALIBERA, R. G. and SOARES,

F. R. Displacements in Reinforced Concrete Beams

using Damage Mechanics: Contribution to Brazilian

Technical Code. Proceedings of the CMNE 2011

Congress on Numerical Methods in Engineering,
Coimbra-Portugal, 2011.

[22] PITUBA, J. J. C. Local state method applied on

the formulation of damage constitutive models for

concrete. Acta Scientiarum. Technology, Vol. 31, n. 1,

pp. 15-23, 2009.

[23] PITUBA, J. J. C. Validation of a damage model for the

non-linear analysis of reinforced concrete structures.

Acta Scientiarum. Technology, Vol. 32, n. 3,

pp. 251-259, 2010.

[24] PITUBA, J. J. C. An anisotropic model of damage

and unilateral effect for brittle materials. International

Journal of Applied Mathematics and Computer

Sciences, Vol. 4, pp.100-105, 2008.
37

Almofadas de apoio feitas de argamassa para ligaes de
concreto pr-moldado: estudo da rugosidade superficial
M. K. EL DEBS a
mkdebs@sc.usp.br
E.K. BELLUCIO b
ellenkellen@sc.usp.br
Abstract
Bearing pads are used in precast concrete connections to avoid concentrated stresses in the contact area between the precast elements. In the
present research, the bearing pads are Portland cement mortar with styrene-butadiene latex (SB), lightweight aggregate (expanded vermiculiteterm) and short fibers (polypropylene, glass and PVA), in order to obtain a material with low modulus of elasticity and high tenacity, compared
with normal Portland cement mortar. The objective of this paper is to analyze the influence of surface roughness on the pads and test other types
of polypropylene fibers. Tests were carried out to characterize the composite and test on bearing pads. Characterization tests show compressive
strength of 41MPa and modulus of elasticity of 12.8GPa. The bearing pads tests present 30% reduction of stiffness in relation to a reference
mortar. The bearing pads with roughness on both sides present a reduction up to 30% in stiffness and an increase in accumulated deformation of
more than 120%, regarding bearing pads with both sides smooth.
Keywords: bearing pads; connections; precast concrete structures.
Resumo
Almofadas de apoio so usadas nas ligaes de concreto pr-moldado para evitar concentrao de tenses na rea de contato entre os elementos. No presente trabalho, so confeccionadas almofadas em argamassa de cimento Portland e areia, que recebem ainda ltex estirenobutadieno (SB), agregado leve (vermiculita termo-expandida) e fibras curtas (polipropileno, vidro e PVA), buscando tornar o material com baixo
mdulo de elasticidade e elevada tenacidade, comparado com a argamassa de referncia, contendo apenas cimento Portland e areia. O objetivo
deste trabalho analisar a influncia da rugosidade superficial nas almofadas e testar um novo tipo de fibras de polipropileno. Foram realizados
ensaios de caracterizao do compsito e ensaios em almofadas. Os resultados de caracterizao indicam resistncia compresso de 41 MPa
e mdulo de elasticidade de 12,8 GPa. Os ensaios de almofadas revelam uma reduo de 30% da rigidez em relao argamassa de referncia.
Nas almofadas com rugosidade em ambos os lados houve uma reduo de at 30% na rigidez e um aumento do afundamento acumulado de
mais de 120%, em relao s almofadas com ambos os lados lisos.
Palavras-chave: almofadas de apoio. ligaes. estruturas de concreto pr-moldado.
a
b
Department of Structural Engineering, University of So Paulo, mkdebs@sc.usp.br, Av. Trabalhador So Carlense,400, So Carlos, Brazil;
Department of Structural Engineering, University of So Paulo, ellenkellen@sc.usp.br, Av. Trabalhador So Carlense,400, So Carlos, Brazil.
Received: 28 Mar 2011 Accepted: 29 Sep 2011 Available Online: 01 Feb 2012
2012 IBRACON
M. K. EL DEBS | E.K. BELLUCIO
1. Introduction
Precast elementsare characterizedby offering ease ofmanufacture.On the other hand, the need tomake connectionsbetween
these elementsto form structures isoneof the main problemsto
be facedin the use of precast concrete. Thus, the connections are
the most importantparts in the design of precast concrete structures and areof fundamental importance both in relation to their
production (manufacture of part of the adjacent elements to the
connections, mounting structure and complementary servicesin
the buildings)and to the performance of thefinishing structure.
The transfer ofcompressive stresseson the connectionsbetween
components ofprecast concreteis usually madebydirect contact
orby placing boards between the parties.Due to the lowtensile
strengthandbrittle performanceof concrete, the first method(direct contact)is rarely usedand is limited tocaseswhere there
arevery lowcompressive stresses.
The second method consists ofplacing aboard between the elements, which ismade of a material withgood deformation capacityto minimizethe concentration of compressive stresses.
Usually, these boardsaredenominatedbearing pads andare made
ofelastomer, the mostcommonpolychloroprene pads. Being made
of a very soft material, these pads accommodate the irregularities
of thesurface, promoting a more uniformstress distribution and allow certain movements of the structure. The movements allowed
arethe rotation andhorizontal displacement.Therotation causesthe
performance of theconnection to be similar to apinnedconnection.
The horizontal displacement enables the relief of stress caused
by the variation in length of the supported elements, which would
beagreat advantage of thistype of material when remarkable loads
are introduced in the structure. The disadvantages of this type of
materialare cost, less durability compared to concrete, low resistance tofire andrelatively lowcompressive strength.
This paper presents the development of bearing pads made of Portland cement mortar, modified to be more deformable and have a
highertoughness than the usual mortars. The padsmade of this materialare intendedto transfercompressive stress -uniform or not -to
avoid stress concentration and, if appropriate, allow small rotations
between the elements without introducinghigh stressesalongthe
edge of the contact area. These pads are made of Portland cement mortaraddingthe following components:a)lightweight aggregateorair-entraining agents,b)latex, and c) short fibers.
The lightweight aggregate (expanded vermiculite-term) or air-entraining agent significantly increases the deformation capacity of
thematerialin the hardened state. Due to the presence of surfactants used in the production of latex, a significant amount of air
canbe incorporated intothe mix, while increasing the deformation
capacity of thematerial.The addition of fibers reduces the workability of themix and can incorporate air into the hardened material, reducing itsmodulus of elasticity.
Several studies were performed to obtainmixtures withlowmodulus of elasticity but with acceptable compressive strengths:
Barboza et al [1], El Debs et al [2], El Debs et al [3], El Debs
etal[4],Montedor[5] and Siqueira [6].
As the mortar pads are less deformable than the corresponding
polychloroprene pads, they initially promote a more rigid structure.
In the case of beam-column connection with grouted dowel, the
replacement of elastomeric pads with this type of pad makes possible the partial transmission of the bending moment and therefore increasesthe rigidityof the structure.A comparisonbetween
the padsof mortarand the pads of polychloroprenecan be seen
inMontedor [5].
Thus, the performance of the pinned connections becomes semirigid. Compared to the pinned connections, semi-rigid connections present asignificant improvement in the structural behavior
of lateral loads, even for connections with a low degree of rigidity (El Debs et al [7]). On the other hand, the effects of length
variation on structural elements must be carefully considered to
minimizethe loads that may be introducedin the structure,since
themortar padisstifferthan the polychloroprene pad.
Figure 1 Bearing pads manufacture
A Bearing pads casting
Bearing pads in curing process
39
Figure 2 Details of the molds and the texture of pads
A Set molds
C Demolding of pads
D Pads with rough and smooth surface
Continuing previous studies to develop this type of bearing pad,

this paperaims to present anexperimental study of theuse ofsurface roughness on mortar pads with the use ofanew form ofmoldingin order to increasethedeformation capacity of the pads.
2. New bearing pads manufacture

technique
The molds used in previous researches, as in studies of Montedor [5] and Siqueira [6], did not provide control of the surface of
the pad,whichwas exposedand justwent through a finishing processduring the curingof the mortar(Figure1).
A battery mold was developed in Bellucio [8]. In addition to enablingthe introduction and control of roughnesson both sidesof
the pads, the new manufacturing technique also provides large-
40
Molds with glued plastic membrane
scale production. A plastic membrane with roughness of 1.2 mm

of thickness, measured by a digital pachymeter, was glued in the
molds to analyze the effect of surface roughnessof the pads.Thus,
a comparison between the pads with smooth and roughsurfaces
can be made(Figure2).
3. Experimental program
In this paper, the following tests were performed to characterize the material: compressive strength, tensile strength by diametral compression and modulus of elasticity. In addition to these
standardizedtests,tests on the performance of the material were
conducted. Concentrated and bending loads were applied under
four points load on strips of pads. The pads were tested by applyingadistributed load (monotonic and cyclic). In order tostudy
the influence of surface roughness, pads were molded with two

smooth surfaces (LL); a smooth and a rough surface (LR); and
tworough surfaces (RR).
3.1 Utilized Materials

The utilized materials are as follows:a)Portland cementof highinitial resistance (CPV-ARI), b) sand sifted by an ABNT30siftwith
hole diameter of 0.59 mm, c) expanded vermiculite-term with a
maximum dimension of4.8mmd) styrene-butadiene latex; superplasticizeragent based on a modified carboxylic ether chain, polypropylene fibers with nominal length of 10mm, nominal diameter
of 12m and specific weight of 0.90kg/dm.
The mixtures utilized in previous studies which have had better
resultsuse as reference 1:0.3:0.1 (Portland cement, sand and water) and Montedor[5] addedtothis mixture V5L30F3 (5%of sand
replaced by expanded vermiculite-term, added 30% of latex and
3% of PVA fiber). Siqueira[6] utilized V10PP4L30 (10% of sand replaced by expanded vermiculite-term, added 4% of polypropylene
fiber and 30% of latex).Because of the batterymolding processit
was necessary tomodify themixture in order to mold the mortar
and continue to meet the criteria ofstrength and toughness.
The amount of fiber incorporated to the mixture was due to the
workabilityof the mortar becauseitwas necessary to have a fluid mass to mold it into the battery (Fig. 3). Since the goal is to
findamixture withgood strength,but withalow modulusof elasticity, fibers are addedas much as possible.
Based on preliminary studies, the following composition was estab-
Figure 3 Casting of pads utilizing the funnel and the 'comb'
41
Figure 4 Stress versus strain diagram

to obtain the modulus of elasticity
NBR7222/1994[10], withpatternedloading rateat 0.005mm/s.

c) Modulus of Elasticity
Thistest is performedbyapplyinganaxial loadto thecylindrical specimens.The application is performed ina circular areaof
50 mm diameter and two transducers with rods of 50mm are
used, placed on opposite sides of the specimens to be tested, approximately 25mm from the base of the 100mm specimens.Thus, the stress versus straindiagrams are madefrom
the load obtained in theacquisition system of the machine and
the average of the two displacement transducers.From the curve
representing stress versus strain and the equation curve (regression)themodulus of elasticity canbe obtained.The regression ismade between5% and50% of the maximum load of the
specimen.Figure 4 shows a typical diagram.
Figure 5 Monotonic loading test
A Test scheme
lished, in volume,for the mortar (PP12): 1:0.285:0.35:0.02(Portland cement: sand: water, superplasticizer), V15L20F2 added
(1.5% of expanded vermiculite-term, 20% of latex and 2% of
polypropylene fiber). The studies also included a reference mixture, no fibers and expanded vermiculite-term,addingthe amount
of water contained in the latex emulsion, and analyzing the influence of these components. The proportion for this mixture is
1:0.3:0.35:0.01 (Portland cement: sand: water: superplasticizer).
Stress versus deformation diagram to obtain pads stiffness
Figure 6 Diagram of the cyclic

test showing stiffness calculation and
accumulated deformation to stress of 20 MPa
3.2 Standard Tests for Mortar Characterization

a) Simple Compression
The simple compression tests were performed in cylindrical specimens of 100mm x 50mm, according to the recommendations of ABNT NBR5739/2007 [9] with loading rate at
0.005mm/s.
b) Tensile strength by diametral compression
These tests were performed on cylindrical specimens of
100mmx50mm, according to the recommendations of ABNT
42
Figure 7 Results of characterization tests in cylindrical specimens
3.3 Mortar characterization tests in strips of pads

As described, themortar placement in the moldingof the padswas
performedmanually (Figure 3) andalso byvibration table.Thedemoldingprocesswas performed after two daysof the molding,being placed for 24 hours in a humid chamber. Then, the samples
were removed from the humid chamber and placedat room temperature untilthe dayof the testing.The tests were performedafter
14days ofthe molding.
a) Concentrated load
Thisconcentrated load testwas derivedfromtestsperformed
by Montedor [5] and Siqueira [6]in order tomeasure the capacity of accommodation under concentrated loads on strips
ofcompositepad, oraccumulated deformation.Thistestwas
performed to compare the influenceof fiber in theaccommodation capacity.
These testsconsist in cuttingpadsof 150mm x 150mminto strips
of 25mm x 150mm,in which loads are appliedat two pointslocated at 37.5 mm from the edge of the strip. The load is applied by the deformation control machineat a rate of 0.01mm/s.
For each mixture, 10 strip samplesof pads are tested, obtaining
theload versus displacement curve for each sample.
b) Bending tests
Thesetests were performedto verify the performanceof the mortarto
bending. Thebearing capacityafter cracking due to the influence of
the fibers was also verified. Six tests were conducted for eachmixture andeach sample was 150mm x 12mm with 30mmwide.
Although performing specific tests on pads, the requirements of ABNT NBR 12142:1991 [11] were established
as a standard for the test. Thus, the same measures of
support distance andload application according to
ABNT NBR 12142:1991 [11]. The support were placed at
1cm from the edge of the sample and the points of load
application were 1/3 support of the brand. The load was
applied to a deformation where the pad began to lose
load.To compare the results,stress versus displacement
curves were designed and the performance of the specimens was verified.
3.4 Bearing pads tests

a) Monotonic loading
In this case, six pads of 150mm x 150mm were molded for
each mixture and for each roughness. The load was applied to
the limitof the machine(1500kN).
Thepurpose of these tests is to analyzethe performanceof the
pad, which is conventionally named stiffness.The relation can
be made from the diagram stress versus deformation. However, as the displacement isconcentrated on the surfaceof the
pads, a stress versus displacement analysis was performed.
With the diagram stress versus displacementa curveis obtained.Alongthe curve, aline is drawn past many points, starting where the variation of the curve tends to a straight line.
The straight lineis drawnonce the pad is accommodated due
Table 1 Results of characterization tests

Mixture
Compression - MPa
Tensile - MPa
Modulus of Elasticity - GPa
PP12
40,9
3,1
12,8
79,5
3,6
23,2
Reference
43
Figure 8 PP12 samples after tests
A Modulus of Elasticity
Tensile for diametral compression
Figure 9 Load versus deformation diagram
A PP12 Mixture
44
Reference Mixture
Figure 10 Average results of

concentrated load
Stiffness wasanalyzed utilizing the load evolution and the accumulated deformation of cycles.Stiffness is calculated in the
typical manner of the monotonic loadingtests.The accumulateddeformationis the consolidation of thematerial, therefore, the displacement suffered by thepadduring each cycle.
The accumulated deformation is calculated viathe difference
between the initial and final valueof thedisplacementcycle
(Figure6).
4. Results and Discussion

4.1 Mortar characterization tests
Six specimens were tested in three different moldings (two samples per
molding). The average of the results is presented in Figure 7 and Table 1.
Through the analysis of the results it is indicated that adding materials
decreases the resistance of the mortar, but it would still be compatible
with use. The samples of modified mixtures did not present brittle fracture
due to the presence of fibers (Figure 8).
Figure11 (a) Samples test; (b) details of reference sample; (c) details of PP12 sample
to its imperfections. The slope of the straight line represents

thestiffnessof the pad(Figure5).
b) Cyclic Loading
The cyclic loading test is used to determine the evolution
of stiffness when exposed to repeated loading. The basic
schemeof the cyclictestwas also usedforthe monotonic test,
and thepadsusedin this testhave thesame dimensions of
the ones used in previous tests (150mm x 150mmbaseand
thicknessof 12mm).The difference is that the loading was repeated cyclicallyand the load valuesused were similar to the
ones which promote contact pressure of the same magnitude
found in practical situations of use ofthe bearing pads.
The stress usedfor this testwerethe same adopted by Siqueira[6] of 2.5MPa, 5.0MPa, 10.0MPa and 20.0MPa.By adoptingthese valuesstress, loadsin the respective phasesof the
testswere approximately 55kN, 112kN, 225kN and 450kN. For
each valueof stress, cyclesof 50repetitions for each loading
were performed, corresponding to 0.02mm/s. The loadswere
appliedto thelimitsand led toavalue close tozero.
Figure 12 Results of bending tests

of the PP12 mixture
45
Figure 13 (a) Results of bending test of the reference mixture; (b) Details of diagram
of initial deformations
There are no significant differences between the tensile strengths. There

is a tendency of the strength values of the reference mortar being slightly
larger than the others. For the modulus of elasticity test, comparing the
results of modified mortars with the reference mixture, there is a significant decrease in this parameter. This is the main focus of the research: the
material should be as deformable as possible to be used as bearing pad.
The results of this test are consistent with those obtained by Siqueira [6]
(35MPa of compression, 2.7 MPa of tensile strenght and 13.2 GPa of
modulus of elasticity), and even with lower values of compression, they
are sufficient to be used in bearing pads.
Figure 15 PP12 mixture stiffness results
4.2 Mortar characterization tests in strips of pads

a) Test of concentrated load
Ten strips of each mixture were tested and an analysis of the
Figure 14 Picture of the bending test

of the PP12 mixture
Figure 16 Reference mixture stiffness results
46
Table 2 Results of stiffness average of monotonic tests (MPa/mm)
1 Casting
PP12 Mixture
2 Casting
3 Casting
Average
LL
307
461
580
449
LR
286
386
444
372
RR
268
313
340
307
Reference Mixture
1 Casting
2 Casting
3 Casting
Average
LL
636
649
642
LR
536
544
540
RR
473
506
490
load versus displacement diagram (Figure 9) was performed.

Figure 10 shows the comparison between the average curves
of the PP12 and the reference mixture.
As expected, the strip of mortar made of the mixture with fibers presented a better performance, as it resisted a greater
load in the evolution of the loading process. This strip had a
performance similar to the beginning of the loading of the reference mixture to the point of the first crack and the mixture
PP12 absorbed more load. The maximum load achieved by the
mixture PP12 was 53.65 kN to 4.53 mm of displacement. The
maximum load achieved by the reference mixture was 37.60
kN to 2.31 mm of displacement, which corresponds to 43%
increase in load and 96% increase in displacement.
The difference between the performances of the mortars is
clear, as presented in Figure 11. The reference mortar hardly
deformed due to the compression and the sample was completely in pieces. The sample with mortar PP12 deformed well
before cracking, and even with the cracks, the fibers allowed
the sample to remain complete.
b) Bending Test
In the strip bending test, six samples taken from each mixture
were used and an analysis of the load versus displacement
graphs was performed. The results are shown in Figure 12 and
Figure 13.
It was verified that each peak point in the curve corresponded
to a crack in the sample in samples with fibers. Thus, it can
be noted that the volumetric rate of the utilized fibers (2%) ensured the appearance of multiple cracks, which gives an acceptable degree of tenacity for the material. Comparing the
values of ultimate load of the graphics of the samples with and
without fibers, it is clear that the values are very close.
It can be seen in the photo in Figure 14 that the strip has a pronounced curvature due to the good tenacity that fibers promote
to the material.
ence of surface roughness applied on the pads were analyzed through stiffness. The stiffness is the ratio of stress
versus displacement of the pads, excluding the initial part of
the diagram which quantifies the ability of elastic deformation of the material.
To facilitate comprehension, the identification LL was adopted for pads with both sides smooth, LR for pads with
one side smooth and one rough and RR for pads with both
sides rough.
For this test 18 pads were casted in two stages: nine pads
were made in a first casting (being 3LL, and 3LR 3RR) and
other nine pads were made in a second casting. However,
through the results of the mixture PP12 it was possible to observe a significant difference between the two casts. Thus,
we chose to cast nine more pads of this mixture to confirm
these results. Even with a new casting the stiffness values
were quite different, especially for the LL and LR pads. Fig-
Figure 17 Stiffness comparison according

to the applied roughness
4.3 Bearing pads tests

a) Monotonic loading
The difference between the studied mixtures and the influ-
47
Figure 18 PP12 bearing pad details PP12 (a) before and (b) after the test
ure 15 and Table 2 show the average values found.

Based on the results, it was detect that the differences between the results of RR pads were much smaller than the
others. Furthermore, these results have much lower stiffness values compared with the LL and LR pads. Therefore,
as the aim of this work is to reduce the stiffness of the pads,
the results were satisfactory compared to those found by
previous studies. It is possible to conclude that, even with
different results between the castings, the RR pads have

good characteristics for their application. The results for the
reference mixture are presented in Figure 16 and Table 2.
In this case, there was no significant difference between the
two casts. Thus, we conclude that the differences found in the
mixtures with fibers are related to the addition of new materials.
Figure 17 shows a comparison of the stiffnesses. In this diagram, the significant reduction in stiffness of the pads with the
Figure 19 Stiffness results comparing 10MPa and 20MPa stresses
48
Figure 20 Stiffness results comparing PP12 mixture and reference mixture
addition of fibers, latex and expanded vermiculite-term can be

noted. There was a reduction of more than 30% in stiffness.
In all mixtures, the application of roughness to the surface reduced
the stiffness of the pad; therefore, the RR pads present a better
result. The difference between the LL and RR pads reaches 30%
of reduction in stiffness. Analyzing once more the stiffness values
of the mixture PP12 in Figure 16, it is clear that, even though it has
been a difference between the values in the casting, there is not a
considerable difference between the stiffnesses for RR pads. The
after testing pads remained intact. (Figure 18).
b) Cyclic Loading
For this test, 18 pads of each mixture were tested: six pads for
each type of roughness, as in the monotonic test. When analyzing the stiffnesses, there is a difference between the stress
of the mixture with fibers and the reference mixture in Figure
19, for the cycle subjected to stress of 10MPa and 20MPa.
Considering the average of the values found, it is possible to
state that the mixture PP12 obtained less stiffness than the
reference mixture. Figure 20 presents the difference between
the stiffness of 10MPa and 20MPa.
Figure 21 Stiffness results comparing applied roughness
49
Figure 22 Accumulated deformation results comparing stiffness
The stiffness for the loading of 20MPa are higher than those of
10MPa. This is due to the fact that this sample was submitted
to the cyclic loads of 2.5MPa and 5MPa, as the 10MPa sample.
It also received all the cyclic loading of 10MPa, therefore, the
sample is more compressed.
Analyzing the values of the results and comparing the roughness, it is observed in Figure 21 that there were not significant
differences between the results. It is believed that the pads
have suffered adaptations due to the previous loadings. Therefore, the roughness did not affect the results.
Besides stiffness, the accumulated deformations were also

analyzed. The results for the studied mixtures are shown in
Figure 22. Through the results it is clear that the pad suffers
a higher plasticization for the loading of 10MPa. This is due
to the fact that this loading was applied before the loading of
20MPa; hence the sample is already deformed. The performance of the mixtures can be observed in Figure 23.
It is possible to see a larger deformation in the pads of mixture
PP12 when analyzing the diagram. This deformation occurs as
expected, since the purpose of this pad is to be more deform-
Figure 23 Accumulated deformation results comparing mortars
50
Figure 24 Accumulated deformation results comparing applied roughness
able. A comparison between the roughness in Figure 24 was

made utilizing the deformation results found.
Considering the two loadings, there was a considerably higher
displacement on the pads with applied roughness. This result
is considered positive, confirming the importance of the use of
these pads.
5. Conclusions
From the analysis of the characterization tests, it is observed that:
a) the values of compressive strength of the material samples are
around 41MPa and of the reference mixture are 79.5 MPa. Thus,
there was a reduction of approximately 48% of the value of the modified mixture compared to the reference mixture. The value around
41MPa can be considered sufficient to be utilized in bearing pads.
b) for the tensile tests, the values found for the mixture with fibers
and the reference mixture are very close, on the order of 3.0
MPa, and this difference is not significant;
c) the modulus of elasticity of the material reached a 45% reduction in relation to the reference mixture, reaching 12.8 GPa;
d) in the concentrated load tests there was a better performance
of the mortar with the use of fibers. It was able to absorb 43%
more load than the reference mortar, and it allowed a displacement of 96% more than the reference mortar.
e) the volumetric rate of the utilized fibers (2%) ensured the appearance of multiple cracks, which provides an acceptable
level of tenacity for the material.
From the bearing pads tests, it can be concluded that:
a) There was a significant difference between the mortar with the
use of fibers and roughness and the reference mortar;
b) In the monotonic loading tests it was observed a clear reduction of
the stiffness of the pad with the application of roughness on both
sides (RR). Comparing the LL and RR pads, the difference was
more than 30%, which indicates that the application of roughness
on both sides enables a significant reduction of this parameter;
B
c) Although the mixture with fibers presented a variation in the
results between the casts, these differences were not significant for the RR roughness, that is, for the results of the mixture PP12 and RR roughness, the values found were satisfactory. For this mixture it was obtained 307MPa/mm, which is
37% less stiffness than that found for the reference mixture,
RR (490MPa/mm). Through the accumulated deformation it
was observed that the mixture PP12 has good deformation
capacity compared to the reference mixture, reaching 37%
more displacement than the reference mixture. It was also
observed that, with the increase of the loading cycles, the material tends to have little elastic deformation. It was also concluded that there was a better performance of the RR roughness pads, and the displacement was 120% higher compared
to LL pads.
Based on all the results achieved, it can be conclude that the
PP12 mixture has a good performance with a reduction in stiffness
compared to previous results, and that the best performance occurred with the application of roughness on both sides of the pad
as shown in the results of monotonic loading.
6. Acknowledgements
To FAPESP, for the financing of tests through the thematic project,
and CNPq for the scholarship of the author.
7. References

[01]

BARBOZA, A.S.R.; SOARES, A.M.M. ; EL DEBS,

M.K. A new material to be used as bearing pad in
precast concrete connections. In Proc.
1st International Conference on Innovation in
Architecture, Engineering and Construction. eds.
Anumba, C.J. Egbu, C.; Thorpe, A., Loughborough,
UK, 2001, p. 81-91.
51
[02]

[03]

[04]

[05]

[06]

[07]

[08]

[09]

[10]

[11]

52
El DEBS, M.K.; BARBOZA, A.S.R.; SOARES, A. M.

M. . An alternative material for bearing pad in precast
concrete connections. In: 17th INTERNATIONAL
CONGRESS OF THE PRECAST CONCRETE
INDUSTRY, Istanbul. Proceedings of 17th
INTERNATIONAL CONGRESS OF THE PRECAST
CONCRETE INDUSTRY, 2002.
EL DEBS, M.K.; BARBOZA, A.S.R.; MIOTTO, A.M.M.
Development of material to be used as bearing pad in
precast concrete connections. Structural Concrete.
n 4 v.4 p. 185-193, 2003.
EL DEBS, M.K.; MONTEDOR, L.C.; HANAI, J.B.
Compression tests of cement-composite bearing
pads for precast concrete connections. Cement and
Concrete Composites, n 28, p. 621-629, 2006.
MONTEDOR, L.C. Desenvolvimento de compsito a
ser utilizado como almofada de apoio nas ligaes
entre elementos pr-moldados, So Carlos, 2004,
Dissertao (Mestrado) Escola de Engenharia de
So Carlos, Universidade de So Paulo, 144p.
SIQUEIRA, G.H. Almofada de apoio de compsito
de cimento para ligaes em concreto pr-moldado,
So Carlos, 2007, Dissertao (Mestrado) Escola
de Engenharia de So Carlos, Universidade de
So Paulo, 169p.
EL DEBS M.K, MIOTTO A.M., EL DEBS, A.L.H.C.
Analysis of a semi-rigid connection for precast
concrete. Proceedings of the Institution of Civil
Engineers: Structures and Buildings, 2010;
163: 41-51.
BELLUCIO, E.K. Influncia da rugosidade superficial
e o uso de novos tipos de fibras em almofadas de
argamassa para ligaes de concreto pr-moldado,
de Engenharia de So Carlos, Universidade de So
Paulo, 106p.
TCNICAS. Concreto Ensaio de compresso em
corpos-de-prova cilndricos, ABNT NBR 5739, Rio de
Janeiro, 2007.
TCNICAS. Argamassas e Concretos Determinao da Resistncia Trao na
Compresso Diametral de Corpos de Prova
Cilndricos: Mtodo de Ensaio, ABNT NBR 7222,
Rio de Janeiro, 1994.
TCNICAS. Concreto Determinao da Resistncia
Trao na Flexo em Corpos de Prova Prismtico,
ABNT NBR 12142, Rio de Janeiro, 1991.
Volume 5, Number 1 (February 2012) p. 38-67 ISSN 1983-4195

Almofadas de apoio feitas de argamassa para ligaes de
concreto pr-moldado: estudo da rugosidade superficial
M. K. EL DEBS a
mkdebs@sc.usp.br
E.K. BELLUCIO b
ellenkellen@sc.usp.br
Abstract
Bearing pads are used in precast concrete connections to avoid concentrated stresses in the contact area between the precast elements. In the
present research, the bearing pads are Portland cement mortar with styrene-butadiene latex (SB), lightweight aggregate (expanded vermiculiteterm) and short fibers (polypropylene, glass and PVA), in order to obtain a material with low modulus of elasticity and high tenacity, compared
with normal Portland cement mortar. The objective of this paper is to analyze the influence of surface roughness on the pads and test other types
of polypropylene fibers. Tests were carried out to characterize the composite and test on bearing pads. Characterization tests show compressive
strength of 41MPa and modulus of elasticity of 12.8GPa. The bearing pads tests present 30% reduction of stiffness in relation to a reference
mortar. The bearing pads with roughness on both sides present a reduction up to 30% in stiffness and an increase in accumulated deformation of
more than 120%, regarding bearing pads with both sides smooth.
Keywords: bearing pads; connections; precast concrete structures.
Resumo
Almofadas de apoio so usadas nas ligaes de concreto pr-moldado para evitar concentrao de tenses na rea de contato entre os elementos. No presente trabalho, so confeccionadas almofadas em argamassa de cimento Portland e areia, que recebem ainda ltex estirenobutadieno (SB), agregado leve (vermiculita termo-expandida) e fibras curtas (polipropileno, vidro e PVA), buscando tornar o material com baixo
mdulo de elasticidade e elevada tenacidade, comparado com a argamassa de referncia, contendo apenas cimento Portland e areia. O objetivo
deste trabalho analisar a influncia da rugosidade superficial nas almofadas e testar um novo tipo de fibras de polipropileno. Foram realizados
ensaios de caracterizao do compsito e ensaios em almofadas. Os resultados de caracterizao indicam resistncia compresso de 41 MPa
e mdulo de elasticidade de 12,8 GPa. Os ensaios de almofadas revelam uma reduo de 30% da rigidez em relao argamassa de referncia.
Nas almofadas com rugosidade em ambos os lados houve uma reduo de at 30% na rigidez e um aumento do afundamento acumulado de
mais de 120%, em relao s almofadas com ambos os lados lisos.
Palavras-chave: almofadas de apoio. ligaes. estruturas de concreto pr-moldado.
a
b
Department of Structural Engineering, University of So Paulo, mkdebs@sc.usp.br, Av. Trabalhador So Carlense,400, So Carlos, Brazil;
Department of Structural Engineering, University of So Paulo, ellenkellen@sc.usp.br, Av. Trabalhador So Carlense,400, So Carlos, Brazil.
Received: 28 Mar 2011 Accepted: 29 Sep 2011 Available Online: 01 Feb 2012
2012 IBRACON
1. Introduo
Os elementos pr-moldados se caracterizam por oferecer facilidades na fabricao. Por outro lado, a necessidade de realizar
as ligaes entre esses elementos, para formar as estruturas,
constitui um dos principais problemas a serem enfrentados no
emprego da pr-moldagem. Assim, as ligaes so as partes
mais importantes no projeto das estruturas de concreto pr-moldado, sendo de fundamental importncia tanto no que se refere
sua produo (fabricao de parte dos elementos adjacentes s
ligaes, montagem da estrutura e servios complementares na
obra) como para o comportamento da estrutura definitiva.
A transferncia de tenses de compresso nas ligaes entre
componentes de concreto pr-moldado normalmente feita mediante contato direto ou mediante a colocao de placa entre as
partes. Devido baixa resistncia trao e o comportamento
frgil do concreto, a primeira maneira (contato direto) pouco
empregada e limitada a casos onde ocorrem tenses de compresso muito baixas.
Na segunda forma, uma placa colocada entre os elementos,
sendo esta feita de material com uma boa capacidade de deformao para minimizar a concentrao das tenses de compresso.
Normalmente, estas placas so denominadas almofadas de
apoio e so feitas de elastmero, sendo as mais comuns as
almofadas de policloropreno. Por ser um material bastante deformvel, estas almofadas acomodam as irregularidades da superfcie, promovendo uma distribuio de tenses mais uniforme
e permite certos movimentos da estrutura. Os movimentos permitidos so a rotao e o deslocamento horizontal. A rotao
faz com que o comportamento da ligao seja prximo de uma
ligao articulada. O deslocamento horizontal possibilita aliviar
as tenses provenientes da variao de comprimento dos elementos apoiados, o que seria uma grande vantagem deste tipo
de material quando esta variao de comprimento introduz esfor-
os de grande magnitude na estrutura. As desvantagens deste

tipo de material so o custo, a durabilidade inferior a do concreto,
baixa resistncia a situaes de incndios e resistncia compresso relativamente baixa.
Este trabalho apresenta o desenvolvimento de almofadas a base
de argamassa de cimento Portland, modificadas para serem mais
deformveis e apresentar maior tenacidade que as argamassas
usuais. As almofadas deste material tm a finalidade de transferir tenses de compresso, uniformes ou no, de forma a evitar
a concentrao de tenses e, se for o caso, permitir pequenas
rotaes entre os elementos, sem introduzir elevadas tenses
junto a borda da rea de contato. Estas almofadas de argamassa
so feitas de cimento Portland e areia, adicionados os seguintes
componentes: a) agregado leve ou um aditivo incorporador de ar,
b) latex, e c) fibras curtas.
O agregado leve (vermiculita termo-expandida) ou agente incorporador de ar aumenta significantemente a capacidade de deformao do material no estado endurecido. Devido presena
de tensoativos usados na produo do ltex, uma quantidade
significativa de ar pode ser incorporada mistura, aumentando
tambm a capacidade de deformao do material. A adio das
fibras reduz a trabalhabilidade da mistura e pode incorporar ar
no material endurecido, reduzindo o seu mdulo de elasticidade.
Vrios estudos foram realizados para obter misturas com o mdulo de elasticidade reduzido porm com foras compressivas
aceitveis: Barboza et al [1], El Debs et al [2], El Debs et al [3], El
Debs et al [4], Montedor [5] e Siqueira [6].
Como as almofadas de argamassa so menos deformveis que
as correspondentes de policloropreno, estas promovem, a princpio, uma maior rigidez da estrutura. No caso da ligao viga-pilar
com chumbador grauteado, a substituio das almofadas elastomricas por este tipo de almofada torna possvel a transmisso
parcial do momento fletor e, consequentemente, aumenta a rigidez da estrutura. Uma comparao entre as almofadas de argamassa e as de policloropreno pode ser vista em Montedor [5].
Figura 1 Confeco das Almofadas
A Moldagem da almofada
54
B Almofada em processo de cura
Figura 2 Detalhes da frma e das placas com superfcie lisa e rugosa
A Frma montada
C Desmoldagem das almofadas
D Almofada com superfcie lisa e outra rugosa
Assim, o comportamento da ligao passa de articulado para semi-rgido. Em comparao com as ligaes articuladas, as semi-rgidas apresentam uma melhora significativa no comportamento
estrutural sobre cargas laterais, mesmo para conexes com um
baixo grau de rigidez (El Debs et al [7]). Por outro lado, os efeitos da variao do comprimento sobre os componentes pr-fabricados deve ser cuidadosamente considerado para minimizar as
foras que podem ser introduzidas na estrutura, uma vez que a
almofada de argamassa mais rgida.
Em continuidade aos estudos anteriores para desenvolver este
tipo de almofada de apoio, este trabalho tem como objetivo apresentar um estudo experimental do emprego de rugosidade superficial nas almofadas de argamassa, com a utilizao de uma nova
forma de moldagem, visando o aumento da capacidade de deformao das almofadas.
Frma com a manta colada
2. Nova forma de fabricao das

almofadas
As frmas utilizadas em trabalhos anteriores, como nos estudos de
Montedor [5] e Siqueira [6], no proporcionavam controle de uma das
superfcies da almofada, a qual ficava exposta e apenas passava por
um processo de acabamento durante a cura da argamassa (Figura 1).
Em Bellucio [8] foi desenvolvida uma frma em bateria. Alm de
possibilitar a introduo e controle de rugosidades nas duas faces
das almofadas, a nova forma de fabricao tambm proporciona
a produo em grande escala. Para a avaliao do efeito da rugosidade superficial das almofadas foram coladas nas frmas uma
manta com rugosidade de 1,2 mm de espessura, medido atravs
de paqumetro digital. Assim, pode-se fazer uma comparao entre as almofadas de superfcies lisas e rugosas (Figura 2).
55
3. Programa Experimental
3.1 Materiais Utilizados
Neste trabalho foram realizados os seguintes ensaios de caracterizao do material argamassa: compresso simples,
trao por compresso diametral, mdulo de elasticidade.
Alm destes ensaios padronizados, foram realizados ensaios
relativos ao comportamento do material, com aplicao de
fora concentrada e de flexo sob quatro pontos de carga
em faixas de almofadas. As almofadas foram ensaiadas com
aplicao de uma fora distribuda, monotnica e cclica.
Para o estudo da influncia da rugosidade superficial, foram
realizadas moldagens de almofadas com as duas superfcies
lisas (LL); uma superfcie lisa e outra rugosa (LR); e as duas
superfcies rugosas (RR).
Os materiais utilizados esto apresentados a seguir: a) cimento

Portland de alta resistncia inicial (CP V-ARI); b) areia passada
na peneira ABNT 30 com dimetro de furos de 0,59mm; c) vermiculita termo-expandida com dimenso mxima de 4,8mm; d) ltex
estireno-butadieno; aditivo superplastificante com base em cadeia
de ter carboxlico modificado; fibras de polipropileno com comprimento nominal de 10 mm, dimetro nominal de 12 mm e peso
especfico de 0,90 kg/dm.
Os traos utilizados em trabalhos anteriores que obtiveram melhores resultados, tomaram por referncia 1:0,3:0,1 (cimento, areia e
gua) e Montedor [5] acrescentou a esse trao V5L30F3 (substituiu 5% de areia por vermiculita termo-expandida, acrescentou
Figura 3 Moldagem das almofadas utilizando o funil e o 'pente'
56
Figura 4 Grfico tenso versus deformao

para obteno do mdulo de elasticidade
componentes. A proporo para esse trao de 1:0,3:0,35:0,01

(cimento:areia:gua:superplastificante).
3.2 Ensaios padronizados de Caracterizao

da argamassa
a) Compresso Simples
Os ensaios de compresso simples foram realizados em corpos-de-prova de 100mmx50mm, conforme as recomendaes
da NBR5739/2007 [9] com taxa de 0,005 mm/s.
b) Trao por Compresso Diametral
Esses ensaios foram realizados em corpos-de-prova de
100mmx50mm, seguindo as recomendaes da NBR7222/1994
[10], com taxa de 0,005 mm/s.
Figura 5 Ensaio de Carregamento Monotnico
A Esquema de ensaio
30% de ltex e 3% de fibra de PVA). Siqueira [6] utilizou V10PP4L30 (substituiu 10% de areia por vermiculita termo-expandida,
acrescentou 4% de fibra de polipropileno e 30% de ltex). Com o
processo de moldagem por bateria, foi necessrio modificar o trao para que a argamassa pudesse ser moldada e continuar atendendo aos critrios de resistncia e tenacidade.
Para a dosagem, a quantidade de fibra foi incorporada conforme
a trabalhabilidade da argamassa, uma vez que fez-se necessrio
uma massa fluida o suficiente para que fosse possvel mold-la na
bateria (Figuras 3). Como o objetivo encontrar um trao com boa
resistncia, porm com um baixo mdulo de elasticidade, adiciona-se a maior quantidade de fibra possvel.
Com base em estudos preliminares foi fixado a seguinte composio, em volume, para a argamassa (PP12): 1:0,285:0,35:0,02 (cim
ento:areia:gua:superplastificante), adicionados V1,5L20F2 (1,5%
de vermiculita termo-expandida, 20% de ltex e 2% de fibra de
polipropileno). O estudo incluiu ainda um trao de referncia, sem
fibras e vermiculita termo-expandida, adicionando a quantidade de
gua referente emulso de ltex e analisando a influncia destes
Curva tenso x deslocamento para obteno da rigidez da almofada
Figura 6 Grfico do ensaio cclico

demonstrando o clculo da rigidez e
afundamento acumulado para tenso de 20 MPa
57
Figura 7 Resultados dos ensaios de caracterizao em corpos-de-prova cilndricos
c) Mdulo de Elasticidade
Este ensaio realizado por meio de aplicao de uma carga
axial ao corpo-de-prova. A aplicao se d numa rea circular
de 50mm de dimetro e utilizam-se dois transdutores com haste de 50mm, colocados em lados opostos do corpo-de-prova
a ser ensaiado, aproximadamente a 25mm da base do corpo-de-prova de 100mm de altura. Assim, os grficos tenso versus deformao so feitos a partir da carga obtida no sistema
de aquisio da mquina e a mdia dos dois transdutores de
deslocamento. A partir da curva representativa de tenso versus deformao e da curva da equao (regresso) pode-se
obter o mdulo de elasticidade. A regresso feita entre 5% e
50% do valor de carregamento mximo do corpo de prova. Na
Figura 4, pode-se observar um grfico no qual mostrada uma
das curvas.
3.3 Ensaios de Caracterizao da argamassa

em faixas de almofada
Conforme descrito, o adensamento na moldagem das almofadas foi realizado manualmente (Figura 3) e tambm atravs
de mesa vibratria. O processo de desmoldagem foi realizado
aps dois dias da moldagem, sendo colocados por mais 24 horas em cmara mida. Em seguida, as amostras foram retiradas da cmara mida e colocadas em temperatura ambiente
at o dia dos ensaios. Os ensaios foram realizados aps 14
dias da moldagem.
C
a) Fora Concentrada
Este ensaio de fora concentrada derivou-se dos ensaios de
fora concentrada realizados por Montedor [5] e Siqueira [6] visando medir a capacidade de acomodao de cargas pontuais
em tiras de almofada do compsito, ou afundamento. Optou-se por realizar esse ensaio para comparar a influncia da fibra
na capacidade de acomodao.
Este ensaio consiste em cortar as almofadas de 150mm x
150mm em fatias de 25mm x 150mm, nas quais so aplicadas
duas cargas pontuais em dois pontos localizados a 37,5mm
da borda da fatia. Aplica-se a carga com controle de deformao da mquina a uma taxa de 0,01mm/s. Para cada trao, 10
amostras de tiras de almofadas so ensaiadas obtendo assim
a curva fora x deslocamento para cada amostra.
b) Ensaio de Flexo
Este ensaio foi realizado para verificar o comportamento da argamassa flexo, em particular, a capacidade portante aps
a fissurao devido a influncia das fibras. Foram realizados
seis ensaios de cada trao e cada amostra tinha 150mm x
12mm com 30mm de largura.
Mesmo se tratando de um ensaio especfico nas almofadas,
verificou-se as prescries da ABNT NBR 12142:1991[11]
para que se pudesse ter um padro para o ensaio. Assim, optou-se por utilizar as mesmas medidas de distncia de apoio
e aplicao da carga conforme a ABNT NBR 12142:1991 [11].
Os apoios foram colocados a 1cm da borda da amostra e os
pontos de aplicao da carga estavam a 1/3 da marca dos
Tabela 1 Resultados dos ensaios de caracterizao

Traos
Compresso - MPa
Trao - MPa
Mdulo de Elasticidade - GPa
PP12
40,9
3,1
12,8
Referncia
79,5
3,6
23,2
58
Figura 8 Amostras PP12 aps os ensaios
A Mdulo de Elasticidade
Trao por Compresso Diametral
Figura 9 Curvas fora x deslocamento
A Trao PP12
Trao de referncia
59
Figura 10 Curvas mdias dos resultados

de fora concentrada
grande quantidade de pontos, a partir do momento em que a

variao da curva se aproxima de uma reta, ou seja, a reta
traada a partir do momento em que j ocorreu a acomodao
da almofada em razo de suas imperfeies. Traada a reta,
avalia-se seu coeficiente angular de forma que este represente a rigidez da almofada (Figura 5).
b) Carregamento Cclico
No ensaio de carregamento cclico procura-se determinar a
evoluo da rigidez para cargas repetitivas. O esquema bsico do ensaio cclico foi o mesmo utilizado para o ensaio
monotnico, sendo que as almofadas utilizadas neste ensaio
possuem as mesmas dimenses do ensaio anterior, ou seja,
150mm x 150mm de base e espessura de 12mm. A diferena
est nos carregamentos que foram repetidos ciclicamente,
alcanando valores de cargas que promovem presses de
contato da mesma ordem de grandeza do que se poderia
verificar em situaes prticas de utilizao das almofadas
de apoio.
As tenses adotadas para este ensaio foram as mesmas de
Figura 11 (a) Amostra no ensaio; (b) detalhe do ensaio na amostra de referncia; (c) detalhe da amostra PP12
apoios. A carga foi aplicada at uma deformao onde a almofada comeava a perder carga. Para comparar os resultados
encontrados, traou-se as curvas tenso-deslocamento onde
verificou-se o comportamento da amostra.
Figura 12 Resultados do ensaio de flexo

para o trao com fibra
3.4 Ensaios de Almofadas

a) Carregamento Monotnico
Neste caso, foram moldadas seis almofadas de 150mm x
150mm para cada trao e para cada rugosidade. O carregamento foi aplicado at o limite da mquina (1500 kN).
O objetivo deste ensaio analisar o comportamento da almofada, o que convencionou-se de rigidez. A relao pode
ser feita a partir do grfico tenso versus deformao, porm, como os deslocamento se concentram na superfcie
das placas, optou-se por fazer uma anlise tenso por
deslocamento.
Com o grfico de tenso versus deslocamento obtm-se uma
curva. Junto curva, traa-se uma reta que passe por uma
60
Figura 13 (a) Resultados de flexo para o trao de referncia; (b) Detalhe do grfico
nos primeiros deslocamentos
Siqueira [6] de 2,5MPa, 5,0MPa, 10,0MPa e 20,0MPa. Com

a adoo destes valores de tenses, as foras nas respectivas fases dos ensaios foram de aproximadamente 55kN,
112kN, 225kN e 450kN. Para cada valor de tenso foram
realizados ciclos de 50 repeties para cada carregamento correspondente com velocidade de 0,02mm/s. As cargas
foram aplicadas at os valores limites e levados a um valor
prximo de zero.
Neste ensaio foram analisadas as rigidezes, utilizando a evoluo do carregamento e o afundamento acumulado dos ciclos. A
rigidez calculada como nos ensaios de carregamento monotnico. O afundamento acumulado a consolidao do material,
ou seja, o deslocamento sofrido pela almofada durante cada
ciclo. O afundamento calculado atravs da diferena entre o
valor final e o inicial de deslocamento do ciclo (Figura 6).
Figura 15 Resultados de rigidez do trao PP12
Figura 14 Foto do ensaio de flexo

do trao PP12
Figura 16 Resultados de rigidez do trao
de referncia
61
Tabela 2 Resultados das mdias das rigidezes do ensaio monotnico (MPa/mm)
1 Moldagem
Trao PP12
2 Moldagem
3 Moldagem
Mdia
LL
307
461
580
449
LR
286
386
444
372
RR
268
313
340
307
Trao de Referncia
1 Moldagem
2 Moldagem
3 Moldagem
Mdia
LL
636
649
642
LR
536
544
540
RR
473
506
490
4. Resultados e Discusses
Foram ensaiados seis corpos de prova em trs moldagens diferentes (duas amostras por moldagem). As mdias dos resultados
esto apresentados na Figura 7 e na Tabela 1.
Analisando os resultados verifica-se que a adio de materiais
diminui a resistncia da argamassa, mas esta ainda seria compatvel com a utilizao. Os corpos de prova dos traos modificados
no apresentaram ruptura frgil devido a presena de fibras (Figura 8).
No ensaio de trao, verifica-se que no h diferenas significativas
entre as resistncias de trao. H uma tendncia dos valores de
resistncia da argamassa de referncia serem um pouco maiores
que os demais. Para o ensaio de mdulo de elasticidade, comparando os resultados das argamassas modificadas com o trao de
referncia, observa-se uma significativa reduo deste parmetro.
Este o foco principal da pesquisa: que o material seja o mais deformvel possvel para ser utilizado como almofada de apoio.
Comparando os resultados encontrados com os valores obtidos
por Siqueira [6], observa-se que os valores esto consistentes
(35MPa de compresso; 2,7MPa de trao e 13,2 GPa de mdulo
de elasticidade), e que mesmo com valores reduzidos de compresso, estes so suficientes para serem utilizados nas almofadas.
gamento do trao de referncia at o ponto da primeira fissura

e depois o trao PP12 seguiu absorvendo mais fora. A fora
mxima alcanada pelo trao PP12 foi de 53,65 kN para 4,53
mm de deslocamento e a do trao de Referncia foi de 37,60
kN para 2,31 mm de deslocamento, o que corresponde a 43%
no aumento da fora e 96% no aumento de deslocamento.
Na Figura 11 percebe-se a diferena entre o comportamento das argamassas. A argamassa de referncia praticamente
no deformou com a compresso e a amostra ficou totalmente
espedaada. J a amostra com a argamassa PP12 deformou
bem antes de fissurar, e mesmo com as fissuras, as fibras permitiram a amostra continuar ntegra.
b) Ensaio de Flexo
No ensaio de flexo de faixas foram utilizadas seis amostras
tiradas de cada trao e realizou-se uma anlise dos grficos
fora versus deslocamento. Os resultados so apresentados
na Figura 12 e Figura 13.
Figura 17 Comparao da rigidez conforme

a rugosidade aplicada

em faixas de almofada
a) Ensaio de Fora Concentrada
Realizou-se uma anlise em grficos fora versus deslocamento de 10 tiras de cada trao (Figura 9). A Figura 10 apresenta a comparao entre as curvas mdias os traos PP12 e
de referncia.
Com era de esperar, a faixa de argamassa feita do trao com
fibras apresentou um comportamento melhor, j que esta resistiu a uma carga maior na evoluo do carregamento. Esta
teve um comportamento bem parecido com o incio do carre-
62
Figura 18 Detalhe da placa de argamassa PP12 (a) antes e (b) depois do ensaio
Verificou-se neste ensaio que, nas amostras com fibras, cada

pico na curva correspondia a uma fissura na amostra. Assim,
pode-se notar que a taxa volumtrica das fibras utilizada (2%)
garantiu o aparecimento de mltiplas fissuras, o que confere
um grau de tenacidade aceitvel para o material. Comparando os valores de carga ltima dos grficos das amostras com
fibras e sem, percebe-se que os valores so bem prximos.
Observa-se na foto da Figura 14 que a tira apresenta uma
curvatura acentuada devido a boa tenacidade que as fibras

promovem ao material.
4.3 Ensaios de Almofadas

a) Carregamento Monotnico
Por meio da rigidez, foi analisada a diferena entre os traos
estudados e a influncia da rugosidade superficial aplicada nas
Figura 19 Resultados de rigidez com carregamento de 10MPa e 20MPa
63
Figura 20 Resultados de rigidez para o trao de referncia e PP12
almofadas. A rigidez a relao tenso versus deslocamento

das almofadas, excluindo a parte inicial do diagrama, onde se
quantifica a capacidade de deformao elstica do material.
Para facilitar o entendimento, adotou-se a identificao LL para
almofadas com ambos os lados lisos, LR para almofadas com
um lado liso e outro rugoso e RR para ambos os lados rugosos.
Para este ensaio foram moldadas 18 almofadas, sendo confeccionadas em duas etapas: nove almofadas em uma primeira moldagem (3LL, 3LR e 3RR) e mais nove almofadas numa
segunda moldagem. Porm, ao verificar os resultados do trao
PP12, observou-se uma significativa diferena entre as duas

moldagens. Assim, optou-se por moldar mais nove almofadas
deste trao para confirmar estes resultados. Mesmo com uma
nova moldagem os valores de rigidez foram bem diferentes,
principalmente para as almofadas LL e LR. A Figura 15 e na
Tabela 2 mostram as mdias de valores encontrados.
Com base nos resultados, verifica-se que as diferenas entre
os resultados das almofadas RR eram bem menores que as
demais. Alm disso, esses resultados apresentam valores de
rigidez bem menores se comparados com as almofadas LL e
Figura 21 Resultados de rigidez comparando as rugosidades aplicadas
64
Figura 22 Resultados de afundamento acumulado comparando as rigidezes
LR. Portanto, como o objetivo do trabalho reduzir a rigidez das

almofadas, os resultados foram satisfatrios se comparados
aos encontrados por trabalhos anteriores. possvel concluir
que, mesmo com resultados diferentes entre as moldagens,
as almofadas RR apresentam boas caractersticas para a sua
aplicao. Para o trao de referncia foram encontrados os resultados expostos na Figura 16 e na Tabela 2. Neste caso, no
houve diferena significativa entre as duas moldagens. Com
isso, podemos concluir que as diferenas encontradas no trao
com fibras so referentes adio dos novos materiais.
Na Figura 17 tem-se a comparao das rigidezes. Percebe-se

neste grfico a reduo significativa da rigidez das almofadas
com a adio das fibras, ltex e vermiculita termo-expandida.
Houve uma reduo de mais de 30% na rigidez.
Em todos os traos a aplicao da rugosidade na superfcie
reduziu a rigidez da almofada, logo, as almofadas RR apresentam melhor resultado. A diferena entre as almofadas LL e
RR chega a 30% de reduo de rigidez. Analisando novamente os valores de rigidez do trao PP12 na Figura 16, por mais
que se tenha uma diferena entre os valores nas moldagens,
Figura 23 Resultados de afundamento acumulado comparando as argamassas
65
Figura 24 Resultados do afundamento acumulado comparando as rugosidades aplicadas
A
verifica-se que para as almofadas RR no h uma diferena
to grande entre as rigidezes. As almofadas mantiveram-se
ntegras aps os ensaios. (Figura 18).
b) Carregamento Cclico
Para este teste foram ensaiadas 18 almofadas de cada trao, sendo seis almofadas de cada rugosidade como no ensaio
monotnico. Analisando primeiramente as rigidezes, tem-se a
diferena entre as tenses do trao com fibras e o de referncia na Figura 19 para o ciclo submetido a tenso de 10MPa e
de 20MPa.
Observando as mdias dos valores encontrados verifica-se
que o trao PP12 obteve rigidez menor que o trao de referncia. Analisando a Figura 20 tem-se a diferena entre as rigidezes de 10MPa e 20MPa. As rigidezes encontradas para o carregamento de 20MPa so maiores que as de 10MPa. Isso se
deve porque a amostra, alm de ter sofrido os carregamentos
cclicos de 2,5MPa e 5MPa do mesmo modo que a de 10MPa,
tambm recebeu todo o carregamento cclico de 10MPa, portanto, a amostra est mais compactada.
Analisando os valores encontrados e comparando as rugosidades, observa-se na Figura 21 que praticamente no houve
diferena entre os resultados. Acredita-se que isso se deve
ao fato das almofadas j terem sofrido acomodaes dos carregamentos anteriores, portanto a rugosidade no influenciou
nos resultados.
Alm da rigidez, tambm foram analisados os afundamentos
acumulados. Na Figura 22 tm-se os resultados para os traos
estudados. Atravs dos resultados percebe-se que, para o
carregamento de 10MPa, a almofada sofre uma maior plastificao. Isso se deve porque esse carregamento foi aplicado
antes do carregamento de 20MPa e portanto a amostra j est
deformada. Pode-se observar o comportamento dos traos na
Figura 23.
Analisando os grficos percebe-se uma deformao maior nas
almofadas de trao PP12, como se esperava, j que seu pro-
66
psito ser mais deformvel. Com os resultados encontrados

de afundamento, tambm se fez uma comparao entre as
rugosidades na Figura 24.
Considerando os dois carregamentos, houve um deslocamento consideravelmente maior nas almofadas com rugosidade
aplicada. Este resultado considerado positivo, corroborando
a importncia do uso dessas almofadas.
5. Concluses
Das anlises dos ensaios de caracterizao, observa-se que:
a) os valores de resistncia compresso das amostras dos
compsitos esto em torno de 41MPa e para o trao de referncia de 79,5MPa. Assim, houve uma reduo de aproximadamente 48% do valor de resistncia do trao modificado em relao ao trao de referncia. Para a utilizao como
almofada de apoio, este valor em torno de 41MPa pode ser
considerado suficiente;
b) para os ensaios de trao, os valores encontrados para o trao com fibra e o trao de referncia esto bem prximos, da
ordem de 3,0MPa, sendo sua diferena pouco significativa;
c) os mdulos de elasticidade dos compsitos chegaram a uma
reduo de 45% em relao ao trao de referncia, chegando
a 12,8GPa;
d) nos ensaios de fora concentrada houve um melhor comportamento da argamassa com a utilizao das fibras, conseguindo
absorver 43% a mais de fora que a argamassa de referncia,
alm de permitir um deslocamento de 96% a mais que a argamassa de referncia.
e) a taxa volumtrica das fibras utilizada (2%) garantiu o aparecimento de mltiplas fissuras, o que confere um grau de tenacidade aceitvel para o material.
Dos ensaios de almofadas, pode-se concluir que:
a) houve uma diferena significativa entre a argamassa com a
utilizao de fibras e rugosidade e a argamassa de referncia;
b) para os ensaios de carregamento monotnico observou-se

claramente a reduo da rigidez da almofada com a aplicao da rugosidade em ambos os lados (RR). Comparando as
almofadas LL e as RR, a diferena chegou a mais de 30%, o
que permiti concluir que a aplicao da rugosidade em ambos
os lados possibilita uma reduo significativa deste parmetro;
c) apesar do trao com fibra apresentar uma variao nos resultados entre as moldagens, para a rugosidade RR essas
diferenas no foram muito grandes, ou seja, para os resultados do trao PP12 e rugosidade RR, os valores encontrados foram satisfatrios. Para este trao, obteve-se 307MPa/
mm, que 37% menos rigidez que a encontrada para o trao
de referncia, RR (490MPa/mm); Atravs dos afundamentos
acumulados observou-se que o trao PP12 apresentou boa
deformao comparada com o trao de referncia, chegando
a 37% mais deslocamento que o trao de referncia. Observa-se tambm que com o aumento dos ciclos de carregamento
o material tende a deformar menos elasticamente. Para este
ensaio tambm se conclui que houve melhor comportamento
das almofadas com rugosidade RR, sendo 120% maior o deslocamento comparado com as almofadas LL.
Assim, com base de todos os resultados, conclui-se que o trao
PP12 obteve um bom comportamento com uma reduo na rigidez comparado aos resultados anteriores, e que o melhor comportamento se deu com a aplicao de rugosidade em ambos os
lados da almofada como demonstrado nos resultados de carregamento monotnicos.
6. Agradecimentos
FAPESP, pelo financiamento dos ensaios atravs do projeto temtico, e ao CNPq, pela bolsa da autora.
7. Referncias Bibliogrficas

[01]

[02]

[03]

[04]

[05]

[06]

[07]

[08]

[09]

[10]

[11]

entre elementos pr-moldados, So Carlos, 2004,

Dissertao (Mestrado) Escola de Engenharia de
So Carlos, Universidade de So Paulo, 144p.
SIQUEIRA, G.H. Almofada de apoio de compsito
de cimento para ligaes em concreto pr-moldado,
Paulo, 169p.
EL DEBS M.K, MIOTTO A.M., EL DEBS, A.L.H.C.
Analysis of a semi-rigid connection for precast
concrete. Proceedings of the Institution of Civil
Engineers: Structures and Buildings, 2010;
163: 41-51.
BELLUCIO, E.K. Influncia da rugosidade superficial
e o uso de novos tipos de fibras em almofadas de
argamassa para ligaes de concreto pr-moldado,
Paulo, 106p.
TCNICAS. Concreto Ensaio de compresso em
corpos-de-prova cilndricos, ABNT NBR 5739, Rio de
Janeiro, 2007.
TCNICAS. Argamassas e Concretos Determinao da Resistncia Trao na
Compresso Diametral de Corpos de Prova
Cilndricos: Mtodo de Ensaio, ABNT NBR 7222,
Rio de Janeiro, 1994.
TCNICAS. Concreto Determinao da Resistncia
Trao na Flexo em Corpos de Prova Prismtico,
ABNT NBR 12142, Rio de Janeiro, 1991.
BARBOZA, A.S.R.; SOARES, A.M.M. ; EL DEBS,

M.K. A new material to be used as bearing pad
in precast concrete connections. In Proc.
1st International Conference on Innovation in
Architecture, Engineering and Construction. eds.
Anumba, C.J. Egbu, C.; Thorpe, A., Loughborough,
UK, 2001, p. 81-91.
El DEBS, M.K.; BARBOZA, A.S.R.; SOARES, A. M.
M. . An alternative material for bearing pad in precast
concrete connections. In: 17th INTERNATIONAL
CONGRESS OF THE PRECAST CONCRETE
INDUSTRY, Istanbul. Proceedings of 17th
INTERNATIONAL CONGRESS OF THE PRECAST
CONCRETE INDUSTRY, 2002.
EL DEBS, M.K.; BARBOZA, A.S.R.; MIOTTO, A.M.M.
Development of material to be used as bearing pad in
precast concrete connections. Structural Concrete.
n 4 v.4 p. 185-193, 2003.
EL DEBS, M.K.; MONTEDOR, L.C.; HANAI, J.B.
Compression tests of cement-composite bearing pads
for precast concrete connections. Cement and
Concrete Composites, n 28, p. 621-629, 2006.
MONTEDOR, L.C. Desenvolvimento de compsito a
ser utilizado como almofada de apoio nas ligaes
67
The influence of low temperature on the evolution

of concrete strength
A influncia das baixas temperaturas na evoluo das
resistncias do concreto
V. CECCONELLO a
vcecconello@gmail.com
B. TUTIKIAN b
bftutikian@unisinos.br
Abstract
The mechanical strength of concrete can be affected by many conditions, among them the proportion of the mixture and the influence of external
agents, such as external temperatures. Thus, the objective of the paper is to analyze the influence of low temperature on the evolution of concrete strength, from the molding of tests specimens cured at various temperatures. The specimens were submitted in the first 7 days, as curing
temperatures ranging from 0C to 25C, simulating the lowest possible value for the start of the cement hydrations reactions to the considered
ideal for development resistances. The results of the experimental program demonstrated at lower cure temperatures the strength of the mixture
develops more slowly, as might be expected, but also found after 14 days, the strength of the mixtures was reversed, i.e., concretes cured at higher
temperatures lower values. Thus, it was concluded the concrete in cold weather can be beneficial to its mechanical performance to over time, at
the expense constructive steps in normal temperature.
Keywords: external agents, low temperatures cure, evolution of resistance, hydration heat of Portland cement paste.
Resumo
As resistncias mecnicas do concreto podem ser afetadas por inmeros condicionantes, como a proporo entre os seus constituintes e a
influncia de agentes externos, como a temperatura ambiente. Com isso, o objetivo deste artigo analisar a influncia das baixas temperaturas na evoluo das resistncias do concreto a base de cimento Portland, a partir da moldagem de corpos-de-prova curados em temperaturas
variadas. Os corpos-de-prova foram submetidos, nos primeiros 7 dias, a temperaturas de cura que variaram de 0C at 25C, simulando as temperaturas mais baixas possveis para o incio das reaes de hidratao do cimento at os valores considerados ideais para o desenvolvimento
das resistncias. Os resultados do programa experimental demonstraram que em temperaturas de cura inferiores a resistncia compresso
da mistura se desenvolve mais lentamente, como era de se esperar, porm tambm verificou-se que a partir dos 14 dias, as resistncias das
misturas se inverteram, ou seja, os concretos curados a temperaturas superiores apresentaram valores menores. Com isso, concluiu-se que a
concretagem em dias frios pode ser vantajosa para o seu desempenho mecnico ao longo do tempo, apesar de prejudicar etapas construtivas
em situaes normais de temperatura.
Palavras-chave: agentes externos, curas a baixas temperaturas, evoluo de resistncia, calor de hidratao da pasta de cimento Portland.
Programa de Ps Graduao em Engenharia Civil, Ppgec, Unisinos, vcecconello@gmail.com Av. Unisinos, 950, centro 6 Exatas,
So Leopoldo / RS, Brasil;
b
Programa de Ps Graduao em Engenharia Civil, Ppgec, Unisinos, bftutikian@unisinos.br Av. Unisinos, 950, centro 6 Exatas,
So Leopoldo / RS, Brasil.
a
Received: 21 Aug 2011 Accepted: 01 Dec 2011 Available Online: 01 Feb 2012
2012 IBRACON
V. CECCONELLO | B. TUTIKIAN
1. Introduction
The reinforced concrete has been established over the years as
one of the most used building systems. According to Helene and
Levy (2003) [1], the concrete will have a more promising future in
the coming decades, once its architectural versatility, combined
with increased durability, will enable the implementation and the
development of any project.
For the concrete to achieve the forecasted resistances for the good
functioning, it is necessary that hydration reactions are completed
between cement and water. Therefore, the external temperature
behaves as a catalyst of these reactions, the hotter it is, and the
faster these reactions occur.
According to Shoukry et al. (2010) [2], the temperature, along with
the relative humidity of the air plays a big role in the reactions of
the hydration of the cement paste by affecting the properties of
fresh and hardened concrete.
However, Brazil is comprised of a large territory, encompassing
many kinds of weather conditions, including low temperatures in
certain periods of the year. In the South region, for example, the
weather can be compared to that of other Latin American countries and the world, where the incidence of low temperatures over
long periods of the year is characterized by a temperate climate.
Yet, the Brazilian norm of project of concrete structures, the ABNT
NBR 6118:2007 [3] does not distinguish between the regions of
the country or external temperatures. Structural designs, and consequently, the execution have been achieved without caution, in
respect to temperature control in the early days of curing, which
may lead to irreversible problems in these structures.
According to Mehta and Monteiro (2008) [4]:
In this respect, the concrete equates to a child: to become
a healthy adult, the newborn needs special attention in the
first times of growth. However, in both instances, there is
no clear definition of how long this early age period lasts.
The Portland cement, by reacting with water, forms a paste with
more or less fluidity, depending on the percentage of added water.
In the course of time, this mixture hardens by irreversible reaction
of water with cement, acquiring mechanical strength able to turn it
in an excellent material for structural performance, under the most
diverse exposure environments (Isaia, 2005 [5]).
According to Anka et al. (2009) [6], the entire process of hydration
of the cement in concrete is very complex, and can be influenced
by many factors. Among the properties most impact the hydration
process is the initial temperature of the concrete and also the temperature of air.
The tricalcium aluminate (C3A), a component of the cement hydrates at a rate much quicker than the silicates. Neville (1997) [7]
quotes the reactions of pure C3A with water occur in a violent manner and results in an immediate hardening of the cement paste,
also known as instant concrete setting time. In order to prevent
this, gypsum (CaSO.2H2O) is added to the clinker manufacturing
process of Portland cement. According to Kirchheim et al. (2010)
[8], an example of the action of the C3A can be seen in the execution of Iber Camargo museum in Porto Alegre, where it was used
white Portland cement with C3A contents above 9%.
The calcium silicates (C3S e C2S) are responsible for about 75%
of the constitution of ordinary Portland cement, exerting a dominating role in determining the characteristics of hardness, also known
as the rate of development of resistance. The hydration of C3S and

C2S in the Portland cement produces C2S in a family of the
hydrated calcium silicates, which have similar structures that vary
widely in the relation between calcium and silica in addition to the
chemically combined water content (Mehta and Monteiro, 2008)
[4]. However, the tetracalcium ferroaluminate (C4AF) results in hydration products structurally similar to the products of C3A, though
the reactivity of C4AF is slower (Neville, 1997) [7].
In parallel to the reactions, the heat of hydration of the Portland
cement is generated which can be observed as a thermal energy
resulting from the contact with water. This development of heat occurs quickly from the mixture of Portland cement clinker milled with
water. The hydration of cement is characterized as an exothermic
reaction, that is, a reaction that generates heat, thus the hydration
process is directly related to the amount of heat generated.
Hence, what is expected of the Portland cement is a good correlation between the rate of the cement hydration and the heat
of hydration. Cements with high hydration heat, hydrate faster
compared to cements that generate less heat of hydration (Taylor,
1967) [9]. Kirchheim et al. (2010) [8] add that in certain works use
large amounts of concrete, its thermal conductivity can be low and
with a high heat release due to the hydration reaction, which can
generate cracks through thermal retraction.
According to Mehta and Monteiro (2008) [4] the heat of hydration
can sometimes be unfavorable, for example, in the mass-concrete
structures. And sometimes be favorable, for example, in concrete
work in the winter or in regions of low temperatures when the environment temperature may be too low to supply the energy required
for the activation of the initial reactions of hydration. According to
Zhang et al. (2008) [10], there is not a single value of activation
energy for all concrete and their different formulations.
According to Pinto Barbosa et al. (2006) [11], the whole process
of hydration of the cement can be compared starting from an evolutionary rate that depends on the concentration and temperature
of all reagents of the solution. The hydration is accelerated at high
temperatures and reduced at lower temperatures. Kim et al. (2002)
[12] emphasize that concrete subjected to high temperatures at
early ages reach higher early resistance, but reduce the growth
rate of resistance over time. Husem and Gztok (2004) [13] add
that cures at high temperatures reduce the strength of concrete including being applied to conventional molded parts in conventional
concrete and of high performance, which can cause problems in
the future.
When the concrete is in a period of solidification or in the beginning
of hardening, and has interference by low temperatures, these actions have a tendency to be decelerated or even reversed owing
to the reduction in the rate of hydration of the active components
of the cement (Cnovas, 1988) [14]. As for the heating, the mechanical properties of concrete decrease remarkably, resulting in a
decrease in the quality of the concrete structure (Kelestemur and
Demirel, 2010) [15].
According to Shoukry et al. (2010) [2], one must consider in structural design the variability of the evolution of concrete properties
due to temperature, as this may affect their structural behavior.
The authors demonstrated for a thermal variation of 80 C (-25 to
55 C) at 28 days of curing, there is a reduction of 38% and 26% of
the resistance to compression and tension, respectively. However,
the work was conducted in extreme temperature conditions of cur-
69
Table 1 Chemical composition of cement

Mechanical Properties
Testing of cement
Finesse #200 (75mm)

Finesse #325 (45mm)
2
Finesse Blaine (g/cm )
Initiation take time (min)
End take time (min)
Compressive strength, 3 days (MPa)
Compressive strength, 7 days (MPa)
Compressive strength, 28days (MPa)
2,49
10,81
3472
190
310
26,1
31,3
40,4
Chemical Properties
Testing of cement
Limit Standard
1,24
4,10
1,81
2,00
0,11
0,71
Insoluble residue (%)

Loss on ignition (%)
MgO (%)
So3 (%)
Na2O (%)
K2O (%)
ing, a condition rarely observed. There is no agreement on the
behavior of concrete under low temperature, but positive, even for
the lack of results with these characteristics.
Limit Standard
122600
60
600
321020
2,5
6,5
4,0
6,5
-
In this research, in the middle of this, it was intendded to analyze the

influence of external low temperature of the concrete at early ages in
order to determine their mechanical properties under this condition.
Table 2 Particle size analysis of fine and coarse aggregate
Opening of sieves (mm)
Fine aggregate
Retained (%)
Coarse aggregate
Accumulated
retained (%)
Retained (%)
Accumulated
retained (%)
# 19
# 12,5
31,6
31,6
# 9,5
45,2
76,8
# 6,3
22,6
99,4
# 4,8
0,6
100
# 2,4
0,3
0,3
100
# 1,2
1,1
1,4
100
# 0,6
10,4
11,8
100
# 0,3
45,5
57,3
100
# 0,15
33,2
90,5
100
background (< 0,075)
9,5
100
100
Fineness modulus
1,61
6,77
Maximum size
1,2 mm
19 mm
Specific gravity
2618 kg/m
2590 kg/m
Unit mass released in the state
1419 kg/m
1197 kg/m
70
2. Objective
The overall goal of this study was to compare the resistances to
compression at 1, 7, 28 and 91 days and 28 days to traction for the
cured concrete at temperatures ranging from zero to 25 C, with
intervals of 5 C, thus analyzing the influence external temperature
on the development of the mechanical properties of concrete.
3. Experimental Programme
For the accomplishment of the trial program it was aimed the reproduction of the effect of the external temperature on the concrete,
by controlling the curing temperature of the specimens. Thus, the
samples were conditioned to various cures in the first seven days,
for testing the resistance to compression and traction.
It was given a trace to obtain the resistance to compression
common to project, around 25 MPa at 28 days. For this purpose, it was defined the trace, in mass of 1: 2: 2.5 (cement: fine
aggregate: coarse aggregate) with consistency, performed by
the procedure specified in ABNT NBR NM 67:2003 [16], set at
90 +/- 10 mm.
3.1 Physical and chemical characterization

of the materials
The characteristics of the materials used in the concrete are directly related to the performance and properties of the mixture. For
the characterization of materials, tests were performed according
the ABNT norms.
3.1.1 Portland Cement
The Portland cement which was used was CP II E 32. It was specified because it is a cement with a lower rate of additions, among
the commercially ones available, and for not being a cement with
high initial resistance. The specific mass used for the Portland cement was
= 3000 kg / m, determined according to the ABNT
norms NBR NM 23:1998 [17], while the mechanical and chemical
properties are listed in Table 1.
3.1.2 Aggregates
The fine aggregate used was extracted from the river sand that
presented a unitary mass
= 1419 kg/m, determined according to recommendation by the ABNT NBR NM 45:2006 [18], and
density of
= 2618 kg/m, determined according to ABNT NBR
9776:1987 [19]. The coarse aggregate consists of crushed basaltic rock, with unit mass
= 1197 kg/m, performed according to
ABNT NBR NM 45:2006 [18]. The density was determined at
= 2590 kg/m, by the method of hydrostatic balance.
The test for sieve analysis of aggregates was performed according
to ABNT NBR NM 248:2003 [20], which foresees the determination of particle size distribution of coarse and fine aggregates for
concrete. The results are shown in Table 2.
3.1.3 Water
Water coming from the public water supply was used for the execution of concrete.
Table 3 Water/cement ratio

and consistency of molding
Moldings
a/c ratio
Consistency (mm)
80
25 C
0,58
20 C
0,56
90
15 C
0,57
100
10 C
0,56
100
5 C
0,55
90
0 C
0,57
100
3.2 Mixture of concrete and molding

of the specimens
At this stage, 62 cylindrical specimens were molded for trials to
compression and 10 traction trials samples for a total of 72 samples metal molded in shapes with dimensions of 10 cm of diameter
and 20 cm in height, two for each age, for six curing temperatures.
The mixture was conducted in a concrete mixer with vertical shaft,
with maximum capacity of 56 liters and rotational speed of 29 revolutions per minute (RPM). The time for adding all materials and
homogenization of the mixture for all moldings was 18 2 min. The
order of the material in the mixer was the following:
a) Course aggregate;
b) 80% of water;
c) Cement
d) Fine Aggregates;
e) 20% of water;
After the stage of mixing the materials, it was measured the consistency of concrete through testing the concrete Abrams cone,
according to ABNT NBR NM 67:1996 [16]. It was established the
a/c ratio, the curing time and the consistency of the concrete, only
for evaluating the influence of the curing temperatures on the mechanical properties of materials. The time for conducting the test
was 8 2 minutes. The results of the slumps and a/c ratio are
presented in Table 3. The concretes analyzed are similar, with all
properties fixed, in order to analyze only the influence of the curing
temperature in the final properties of the material.
Afterwards, the samples were molded in a period of 25 2 minutes. Thus, the total time between the contact of the cement with
water and the storage of specimens in the cure environment with
controlled temperature at 51 6 minutes.
3.3 Conditions of curing
The condition of cure of the samples was carried out in an environment of low controlled temperatures, simulating the behavior of the
concrete when subjected to this incidence, as shows Table 4, and
relative humidity of 50%. The equipment used was a sealed box
with digital temperature control, which allows setting the temperature without affecting the relative humidity of the air for a predetermined time, in the case study, for 7 days.
In this condition, by keeping the samples at various curing temperatures in the first seven days, the behavior of the first reactions
which occur at early ages can be analyzed.
71
Figure 1 Evolution of compressive strength

in the first 7 days in controlled healing
Table 4 Weather conditions and

temperature of the molding
Moldings
Time and temperature of cure

0 a 7 days
8 a 91 days
25 C
20 C
20 C
20 C
3
4
15 C
10 C
20 C
20 C
5 C
20 C
0 C
20 C
4. Presentation and Result Analysis

After the rupture of the specimens in the mechanical tests, it was
compared the results to reach conclusions regarding the subject.
The results of the resistance to compression were divided into two

stages, because it presents different behaviors. The first stage
ranged from 1 to 7 days, while the second stage was conducted
from 8 to 91 days.
The Table 5 shows the evolution of the resistance compression in
the first 7 days, when the samples were conditioned to a controlled
cure at low temperatures, ranging from 0 to 25 C at 5 C.
For better viewing, it was presented the results in Figure 1, relating the resistance compression in MPa, in the y axis at the age of
rupture of the samples of the trial in x axis, in days.
In this first period of curing it can be observed a greater variation of
resistance in the first day of measurement, which was 11.4 MPa for a
temperature of 25 C and at 1.2 MPa for a temperature of 0 C. The
higher curing temperatures presented superior outcomes of resistance
to compression and the resistance was gradually decreased to the lowest temperature with the worst performance, as might be expected.
On the seventh day this variation was 24.4 MPa for a temperature of 15 C and at 18 MPa for a temperature of 5 C. The difference between the highest and lowest value decreased 37.25%
compared to the first period. That is, in the 7 days of curing, the
influence of initial curing temperatures have been lower, reaching even to overlap different curves. Although the resistance gain
was more significant for cured concretes at lower temperatures,
approaching the cured concrete resistance at 0 C increased about
1,600% between the 1st and 7th day, compared to a gain of 88%
of the cured at 25 C.
Table 6 shows the evolution of resistance from 14 to 91 days of
curing. It is also illustrated in Figure 2, the results in chart form, for
better viewing.
In the second period of cure it is observed the approximation of
the results from day 7, the inversion in resistance values occurring
Table 5 Evolution of compressive

strength in healing controlled
Table 6 Evolution of resistance to

compression in air-conditioned room (20 C)
Compressive strength
1 day
(MPa)
Compressive strength
7 days
(MPa)
14 days
(MPa)
28 days
(MPa)
91 days
(MPa)
25
11,4
21,5
25
24,3
26,9
29,2
20
11
23,4
20
27,5
31
31,4
15
8,3
24,4
15
28,7
21,7
27,8
10
6,5
21,7
10
27,3
31,8
28,7
3,5
18
20,9
35,6
37,6
1,2
18,8
30,3
38,6
38,8
4.1 Trial of compression of the cylindrical specimens
72
Figure 2 Evolution of resistance to

compression in air-conditioned room (20 C)
Figure 4 Evolution of compressive

strength as a function of curing temperature
Evolution of compressive strength as a function of
curing temperature
50
fc (MPa)
from day 14, with the concrete cured at lower temperatures and
showing better results than those cured at higher temperatures.
At 28 days the ratio of resistance with the previous age followed the
trend of higher growth for lower temperatures. The variation of resistance at this age was 21.7 MPa for a temperature of 15 C, 38.6 MPa
for a temperature of 0 C, considering a 43% reduction in the range
of 15 C, for the total range of 25 C the reduction was 30%, similar
to the results in the referenced bibliography. For the age of 91 days
the ratio between the resistance of the concretes maintained the tendency, with the cured at zero and 5 C with higher rates of resistance
to compression. It is presented in Table 7 with all values grouped.
Figure 3 shows the behavior in graph form, where it was observed
the highest resistance to compression for concretes cured under
high temperatures at early ages, with the inversion occurring at 14
days, showing that the slow hydration is an advantage over time
for the concrete structure.
Such behavior is evidenced in Figure 4, which correlates the
compressive strength in MPa, in the y axis, with the curing
Figure 3 Evolution of compressive strength
1d
40
7d
30
28d
14d
91d
20
Linear (1d)
Linear (7d)
10
Linear (14d)
Linear (28d)
0
0
10
15
20
25
30
Linear (91d)
t (C)
temperature, in C, in the axis x to the ages of the tests of

the samples. It is observed the behavior of the outcomes at 1
Table 7 Evolution of compressive strength at 91 days
1 day
MPa
25
20
15
10
5
0
11,4
11
8,3
6,5
3,5
1,2
Resistncia compresso
7 days
14 days
MPa
MPa
21,5
23,4
24,4
21,7
18
18,8
24,3
27,5
28,7
27,3
20,9
30,3
28 days
MPa
91 days
MPa
26,9
31
21,7
31,8
35,6
38,6
29,2
31,4
27,8
28,7
37,6
38,8
73
Table 8 Behavioral equations relating the

compressive strength and curing temperature
Property
Equation behavior
1 day
fc = 0,4303*t + 1,6048
7 days
fc = 1,1851*t + 18,986
14 days
fc = -0,0503*t + 27,129
28 days
fc = -0,4709*t + 36,819
91 days
fc = -0,3857*t + 37,071
day is increasing, that is, the higher the cure temperature, the
higher the compression resistance at this age. This trend is
again observed at 7 days, but more smoothly. At 14 days the
equation is almost linear, thus independent of the curing temperature on the compression resistance is similar. While at 28
to 91 days the behavior of the results is decreasing, that is, the
concrete compression resistance is affected by the increase of
the curing temperature.
For all ages there were determined behavioral equations, shown
in Table 8. The described behavior can be analyzed through equations, for the first number, which multiplies the t (cure temperature) is positive for increasing results at 1 and 7 days, close to zero
at 14 days, showing the linearity and negative when the values
decrease to 28 and 91 days. Yet, the isolated portion of the equation represents the value that the line intersects the axis y, that is,
the compressive strength when cured at 0 C. This value is higher
for curing temperatures that provided the greatest resistance to
compression and lower for those who reached the lowest values, it
means, higher for higher temperatures.
4.2 Trial to determine the resistance to traction

by diametrical compression of the cylindrical

specimens
The presentation of the resistance to traction results at 28 days of
curing are shown in Table 9. Figure 5 illustrates the behavior of the
Table 9 Tensile strength at 28 days

Traction resistance
Temperature (C)
28 days (MPa)
74
Figure 5 Traction resistance at 28 days
equation, relating the resistance to traction, in the y axis, with the

curing temperature in C, in the x axis.
It is observed that, again, the results at 28 days were higher than
the cured concretes at lower temperatures, confirming the behavior seen for the resistance to compression. The resistance variation for age was 9.0 MPa to 25 C, 13.2 MPa for a temperature of 0
C, considering a 31% reduction in the range of 25 C, approaching
the results presented in referenced bibliography. And the correlation coefficient was 0.85, considered to be acceptable for this type
of trial.
5. Conclusions
After carrying out this study, it can be observed in the early
ages, between 1 and 7 days, the resistance to compression
of cured concrete at higher temperatures was higher compared to the resistance to lower curing temperatures, as was
expected. At these ages the increase in resistance was due to
the high degree of hydration of these parts, explained by the
high value of the activation energy. It is noteworthy the curing
temperatures of 20 C and 25 C for 1 day, which showed the
highest values. However, on day 7, when the activation energy
does not exercise much influence in the parts, the resistance
had increased closeness
From the day 14 there was an inversion in the samples of greatest
resistance, that is, those who were initially cured at low temperatures
achieved the best results, but with lower differences between them.
These differences increased at 28 days, with the best performance
to the parts initially cured at 0 C. Finally, it was proven that the more
slowly it is the hydration of Portland cement, the better the formation
of its crystalline structure, thus justifying superior performance for
the parts cured in its early ages at low temperatures, that is, temperatures which slow down and / or slow the hydration process.
25
20
9,0
10,5
15
10,0
10
10,8
6. References
11,3
13,2
[01] HELENE, P. R. L.; LEVY S. M. Estado da arte do

concreto como material de construo. So Paulo:

Exacta. 2003. 8p.
[02] Shoukry, S. N.; William, G. W.; Downie, B.;

Riad, M. Y.; Effect of moisture and temperature on the

mechanical properties of concrete. USAL: Elsevier,

2010.

[03] ASSOCIAO BRASILEIRA DE NORMAS TCNICAS.

NBR 6118. Projeto de estruturas de concreto.

Procedimento. Rio de Janeiro: ABNT, 2007.

[04] MEHTA, P. K.; MONTEIRO, P. J. M. Concreto:

microestrutura, propriedades e materiais. So Paulo:

Ibracon, 2008. 674 p.

[05] ISAIA, G. C. Concreto: ensino, pesquisa e realizaes.

So Paulo: IBRACON, 2005. 1600p.

[06] ANKA ILC; GORAN TURK; FRANCI KAVCIC;

GREGOR TRTNIK. New numerical procedure for the

prediction of temperature development in early age

concrete structures. USAL: Elsevier, 2009.

[07] NEVILLE, A M. Propriedades do concreto. Traduo

Eng. Salvador E. Giammusso. So Paulo: Pini, 1997.

828 p.

[08] KIRCHHEIM, A. P.; SOUZA, R. B.; DAL MOLIN, D. C.

C.; MONTEIRO, P. J. M. lcalis incorporados ao

aluminato triclcico: efeitos na hidratao. Ambiente

construdo. 2010. 14p.

[09] TAYLOR H. F. W. La quimica de los cementos. Bilbao:
Urmo, 1967. 512 p.

[10] Zhang, J.; Cusson, D.; Monteiro, P.; Harvey,

J.; New perspectives on maturity method and

approach for high. USAL: Elsevier, 2008.

[11] PINTO BARBOSA, M.; BERTOLUCCI, F. S.; PINTO,

R. C. A.; PERES, L. D. P. Avaliao da energia

aparente de ativao do cimento CP-II E CP-V com

adio de superplastificantes. Anais do VI Simpsio

EPUSP sobre Estruturas de Concreto, 2006. 15 p.

[12] Kim, J. K.; Han, S. H.; Song, Y. C.; Effect of

temperature and aging on the mechanical properties

of concrete Part I. Experimental results. USAL:

Elsevier, 2002.

[13] Husem, M.; Gozutok, S.; The effects of low

temperature curing on the compressive strength of

ordinary and high performance concrete. USAL:

Elsevier, 2004.

[14] CNOVAS, M. F. Patologia e terapia do concreto

armado. So Paulo: PINI, 1988. 522 p;

[15] DEMIREL, B.; KELESTEMUR, O.; Effect of elevated

temperature on the mechanical properties of concrete

produced with finely ground pumice and silica fume.
USAL: Elsevier, 2010.


NBR NM 67. Concreto Determinao da

consistncia pelo abatimento do tronco de cone. Rio

de Janeiro: ABNT, 1998.


NBR NM 23. Cimento. Determinao da massa

especfica. Rio de Janeiro: ABNT, 1998.


NBR NM 45. Determinao da Massa Unitria de

Agregados. Rio de Janeiro: ABNT, 2006.


NBR 9776. Determinao da Massa Especfica de

[20]

Agregados Midos - Por meio do frasco de Chapman.

Rio de Janeiro: ABNT, 1987.
ASSOCIAO BRASILEIRA DE NORMAS TCNICAS.
NBR NM 248. Agregados Determinao da
composio granulomtrica. Rio de Janeiro:
ABNT, 2003.
75
The influence of low temperature on the evolution

of concrete strength
A influncia das baixas temperaturas na evoluo das
resistncias do concreto
V. CECCONELLO a
vcecconello@gmail.com
B. TUTIKIAN b
bftutikian@unisinos.br
Abstract
The mechanical strength of concrete can be affected by many conditions, among them the proportion of the mixture and the influence of external
agents, such as external temperatures. Thus, the objective of the paper is to analyze the influence of low temperature on the evolution of concrete strength, from the molding of tests specimens cured at various temperatures. The specimens were submitted in the first 7 days, as curing
temperatures ranging from 0C to 25C, simulating the lowest possible value for the start of the cement hydrations reactions to the considered
ideal for development resistances. The results of the experimental program demonstrated at lower cure temperatures the strength of the mixture
develops more slowly, as might be expected, but also found after 14 days, the strength of the mixtures was reversed, i.e., concretes cured at higher
temperatures lower values. Thus, it was concluded the concrete in cold weather can be beneficial to its mechanical performance to over time, at
the expense constructive steps in normal temperature.
Keywords: external agents, low temperatures cure, evolution of resistance, hydration heat of Portland cement paste.
Resumo
As resistncias mecnicas do concreto podem ser afetadas por inmeros condicionantes, como a proporo entre os seus constituintes e a
influncia de agentes externos, como a temperatura ambiente. Com isso, o objetivo deste artigo analisar a influncia das baixas temperaturas na evoluo das resistncias do concreto a base de cimento Portland, a partir da moldagem de corpos-de-prova curados em temperaturas
variadas. Os corpos-de-prova foram submetidos, nos primeiros 7 dias, a temperaturas de cura que variaram de 0C at 25C, simulando as temperaturas mais baixas possveis para o incio das reaes de hidratao do cimento at os valores considerados ideais para o desenvolvimento
das resistncias. Os resultados do programa experimental demonstraram que em temperaturas de cura inferiores a resistncia compresso
da mistura se desenvolve mais lentamente, como era de se esperar, porm tambm verificou-se que a partir dos 14 dias, as resistncias das
misturas se inverteram, ou seja, os concretos curados a temperaturas superiores apresentaram valores menores. Com isso, concluiu-se que a
concretagem em dias frios pode ser vantajosa para o seu desempenho mecnico ao longo do tempo, apesar de prejudicar etapas construtivas
em situaes normais de temperatura.
Palavras-chave: agentes externos, curas a baixas temperaturas, evoluo de resistncia, calor de hidratao da pasta de cimento Portland.
Programa de Ps Graduao em Engenharia Civil, Ppgec, Unisinos, vcecconello@gmail.com Av. Unisinos, 950, centro 6 Exatas,
So Leopoldo / RS, Brasil;
b
Programa de Ps Graduao em Engenharia Civil, Ppgec, Unisinos, bftutikian@unisinos.br Av. Unisinos, 950, centro 6 Exatas,
So Leopoldo / RS, Brasil.
a
Received: 21 Aug 2011 Accepted: 01 Dec 2011 Available Online: 01 Feb 2012
2012 IBRACON
1. Introduo
O concreto armado tem se consagrado no decorrer dos anos como
um dos sistemas construtivos mais utilizados. Segundo Helene e
Levy (2003) [1], o concreto ter um futuro cada vez mais promissor
nas prximas dcadas, uma vez que sua versatilidade arquitetnica, aliada durabilidade crescente, possibilitar a realizao e o
desenvolvimento de qualquer projeto.
Para o concreto atingir resistncias previstas para o bom funcionamento, necessrio que reaes de hidratao se consumam
entre o cimento e a gua. Desta forma a temperatura externa comporta-se como catalisador destas reaes, quando mais quente
estiver, mais rapidamente estas reaes ocorrero.
Segundo Shoukry et al. (2010) [2], a temperatura, conjuntamente
com a umidade relativa do ar, desempenha um grande papel nas
reaes de hidratao da pasta de cimento, afetando as propriedades do concreto no estado fresco e no endurecido.
No entanto, o Brasil formado por um territrio extenso, englobando vrios tipos de condies climticas, inclusive locais com
baixas temperaturas em determinados perodos do ano. Na regio
sul, por exemplo, o clima pode ser comparado como o de outros
pases da Amrica Latina e do mundo, onde a incidncia de baixas temperaturas em longos perodos do ano caracterizada pelo
clima temperado. Ainda assim, a norma brasileira de projeto de
estruturas de concreto, a ABNT NBR 6118:2007 [3] no diferencia
as regies do pas ou temperaturas externas. Os projetos estruturais e, consequentemente, a execuo vm sendo realizados
sem cuidado no que diz respeito ao controle da temperatura de
cura nos primeiros dias, podendo surgir problemas irreversveis
nessas estruturas.
Conforme Mehta e Monteiro (2008) [4]:
Nesse aspecto, o concreto equipara-se a uma criana:
para se tornar um adulto saudvel, o recm nascido precisa de ateno especial durante os primeiros perodos
de crescimento. No entanto, em ambos os casos, no h
definio clara de quanto tempo dura esse perodo de primeiras idades.
O cimento Portland, ao reagir com a gua, forma uma pasta com
mais ou menos fluidez, dependendo do percentual de gua adicionada. Com o passar do tempo, essa mistura endurece pela
reao irreversvel da gua com o cimento, adquirindo resistncia mecnica capaz de torn-la um material de excelente desempenho estrutural, sob os mais diversos ambientes de exposio
(Isaia, 2005 [5]).
Conforme Anka et al. (2009) [6], todo o processo de hidratao do
cimento no concreto muito complexo, podendo ser influenciado
por inmeros fatores. Entre as propriedades que mais afetam o
processo de hidratao est a temperatura inicial do concreto e
tambm a temperatura do ar.
O aluminato triclcico (C3A), componente do cimento, se hidrata
em uma velocidade muito mais rpida do que os silicatos. Neville
(1997) [7] cita que as reaes do C3A puro com gua ocorrem
de uma maneira violenta e resulta em um enrijecimento imediato da pasta de cimento, conhecido tambm como pega instantnea. Com o intuito de evitar que isso acontea, adiciona-se gesso
(CaSO.2H2O) ao clnquer no processo de fabricao do cimento Portland. Segundo Kirchheim et al. (2010) [8], um exemplo da
ao do C3A pode ser visualizado na execuo do museu Iber
Camargo em Porto Alegre, onde foi utilizado cimento Portland

branco com teores de C3A acima de 9%.
Os silicatos de clcio (C3S e C2S), so responsveis por cerca
de 75% da constituio do cimento Portland comum, tm um papel dominante na determinao das caractersticas de endurecimento, tambm conhecida como a taxa de desenvolvimento da
resistncia. A hidratao do C3S e C2S no cimento Portland produz uma famlia dos silicatos de clcio hidratados que tem suas
estruturas semelhantes, que variam amplamente na relao entre clcio e slica, alm do teor de gua quimicamente combinada
(Mehta e Monteiro, 2008) [4].
J o ferroaluminato tetraclcico (C4AF) resulta em produtos de hidratao semelhantes estruturalmente aos produtos do C3A, apesar de que a reatividade do C4AF mais lenta (Neville, 1997) [7].
Em paralelo s reaes gerado o calor de hidratao do cimento
Portland, que pode ser observado como uma energia trmica resultante do contato com a gua. Essa evoluo de calor se d de
maneira rpida a partir da mistura do clinquer de cimento Portland
modo com gua. A hidratao do cimento caracterizada como
uma reao exotrmica, ou seja, reao que gera calor, com isso
o processo de hidratao est diretamente relacionado quantidade de calor gerada.
Portanto, o que se espera do cimento Portland uma boa correlao entre a velocidade de hidratao do cimento e o calor de hidratao. Cimentos com alto calor de hidratao se hidratam com
maior rapidez em comparao aos cimentos que geram menor
calor de hidratao (Taylor, 1967) [9]. Kirchheim et al. (2010) [8]
complementam que em determinadas obras que utilizam grandes
quantidades de concreto, a sua condutividade trmica pode ser
baixa e com a elevada liberao de calor devido reao de hidratao, podendo gerar fissuras por retrao trmica.
Conforme Mehta e Monteiro (2008) [4], o calor de hidratao pode,
s vezes, ser desfavorvel como, por exemplo, em estruturas de
concreto-massa. E outras vezes ser favorvel como, por exemplo, em concretagem no inverno ou em regies de temperaturas
baixas quando a temperatura do ambiente pode estar muito baixa
para fornecer energia necessria para a ativao das reaes iniciais de hidratao. Segundo Zhang et al. (2008) [10], no existe
um valor nico de energia de ativao para todos os concretos e
suas diferentes formulaes.
De acordo com Pinto Barbosa et al. (2006) [11], todo o processo
de hidratao do cimento pode ser comparado a partir de uma
taxa de evoluo que depende da concentrao e da temperatura de todos os reagentes da soluo. A hidratao acelerada
em temperaturas altas e reduzida em temperaturas mais baixas.
Kim et al. (2002) [12] ressaltam que concretos submetidos a altas
temperaturas em idades iniciais alcanam mais cedo altas resistncias, porm diminuem a taxa de crescimentos de resistncia
ao longo do tempo. Husem e Gozutok (2004) [13] complementam
que curas em altas temperaturas diminuem a resistncia do concreto, sendo inclusive aplicadas a peas moldadas em concretos
convencionais e de alto desempenho, o que pode acarretar problemas no futuro.
Quando o concreto est em perodo de pega ou em princpio de
endurecimento, e sofre a interferncia das baixas temperaturas,
essas aes apresentam a tendncia de serem desaceleradas ou,
at mesmo, anuladas, devido reduo na velocidade de hidratao dos compostos ativos do cimento (Cnovas, 1988) [14]. J em
77
Tabela 1 Composio qumica do cimento

Propriedades Mecnicas
Ensaio do cimento
Finura #200 (75mm)

Finura #325 (45mm)
2
Finura Blaine (g/cm )
Incio de tempo de pega (min)
Fim de tempo de pega (min)
Resistncia compresso aos 3 dias (MPa)
Propriedades Qumicas
Limite de norma
122600
60
600
321020
2,49
10,81
3472
190
310
26,1
31,3
40,4
Ensaio do cimento
Limite de norma
1,24
4,10
1,81
2,00
0,11
0,71
Resduo insolvel (%)

Perda ao fogo (%)
MgO (%)
So3 (%)
Na2O (%)
K2O (%)
casos de aquecimento as propriedades mecnicas do concreto
diminuem notavelmente, resultando em uma diminuio na qualidade da estrutura de concreto (Demirel e Kelestemur, 2010) [15].
2,5
6,5
4,0
6,5
-
Conforme Shoukry et al. (2010) [2], deve-se considerar em projetos estruturais a variabilidade da evoluo das propriedades do
concreto devido temperatura, uma vez que isto pode afetar o
Tabela 2 Anlise granulomtrica do agregado mido e grado

Agregado mido
Agregado grado
Abertura das peneiras (mm)
Massa
Retida (%)
Massa Retida
Acumulada
Massa
Retida (%)
Massa Retida
Acumulada
# 19
# 12,5
31,6
31,6
# 9,5
45,2
76,8
# 6,3
22,6
99,4
# 4,8
0,6
100
# 2,4
0,3
0,3
100
# 1,2
1,1
1,4
100
# 0,6
10,4
11,8
100
# 0,3
45,5
57,3
100
# 0,15
33,2
90,5
100
Fundo (< 0,075)
9,5
100
100
Mdulo de finura
1,61
6,77
Dimenso mxima
1,2 mm
19 mm
Massa especfica
2618 kg/m
2590 kg/m
Massa unitria no estado solto
1419 kg/m
1197 kg/m
78
seu comportamento estrutural. Os autores demonstraram que

para uma variao trmica de 80 C (-25 a 55 C), aos 28 dias de
cura, h uma reduo de 38% e 26% da resistncia compresso
e trao, respectivamente. Porm, o trabalho foi realizado com
condies climticas em extremos de temperatura de cura, condio raramente observada. No h um consenso sobre o comportamento do concreto em situaes de baixa temperatura, porm
positivas, at por falta de resultados com estas caractersticas.
Nesta pesquisa, em meio a isso, pretende-se analisar a influncia de
baixas temperaturas externas nas primeiras idades do concreto, a
fim de determinar suas propriedades mecnicas sob essa condio.
2. Objetivo
O objetivo geral deste trabalho foi comparar as resistncias compresso a 1, 7, 28 e 91 dias e trao aos 28 dias para concretos
curados a temperaturas variando de zero a 25C, com intervalos
de 5C, analisando assim a influncia da temperatura externa no
desenvolvimento das propriedades mecnicas do concreto.
3. Programa Experimental
Para a realizao do programa experimental buscou-se reproduzir o efeito da temperatura externa sobre o concreto, atravs do
controle da temperatura de cura dos corpos-de-prova. Com isso,
as amostras foram condicionadas a curas variadas nos primeiros
sete dias, para a realizao dos ensaios de resistncia compresso e trao.
Foi determinado um trao para obter resistncias compresso
usuais de projeto, em torno dos 25 MPa aos 28 dias. Para isso,
definiu-se o trao, em massa, de 1: 2: 2,5 (cimento: agregado mido: agregado grado), com a consistncia, realizada atravs do
procedimento especificado na ABNT NBR NM 67:2003 [16], fixada
em 90 + 10 mm.
3.1 Caracterizao fsica e qumica dos materiais

As caractersticas dos materiais empregados no concreto esto diretamente relacionadas ao desempenho e propriedades da mistura. Para a caracterizao dos materiais, foram realizados ensaios
de acordo com as normas da ABNT.
3.2.1 Cimento Portland
O cimento Portland utilizado foi o CP II E 32. Foi especificado por
se tratar de um cimento com menor ndice de adies, entre os comercialmente disponveis, e por no se tratar de um cimento com
alta resistncia inicial. A massa especifica do cimento Portland
utilizado foi de
= 3000 kg/m, determinada conforme ABNT
NBR NM 23:1998 [17], enquanto que as propriedades mecnicas
e qumicas so detalhadas na Tabela 1.
3.2.2 Agregados
O agregado mido utilizado foi areia extrada de rio, que apresentou massa unitria de
= 1419 kg/m, determinada conforme
o preconizado pela ABNT NBR NM 45:2006 [18], e massa especfica de
= 2618 kg/m, determinada conforme ABNT NBR
9776:1987 [19]. J o agregado grado constitudo por rocha
Tabela 3 Relao gua/cimento

e consistncia das moldagens
Temp. Cura (C)
Relao a/c
Consistncia (mm)
80
25 C
0,58
20 C
0,56
90
15 C
0,57
100
10 C
0,56
100
5 C
0,55
90
0 C
0,57
100
basltica britada, apresentando a massa unitria de

= 1197
kg/m, realizada conforme ABNT NBR NM 45:2006 [18]. A massa
especfica foi determinada em
=2590 kg/m, atravs do mtodo da balana hidrosttica.
O ensaio de anlise granulomtrica dos agregados foi realizado
conforme a ABNT NBR NM 248:2003 [20], que prev a determinao da composio granulomtrica de agregados midos e grados
para concreto. Os resultados esto representados na Tabela 2.
3.2.3 gua
Foi utilizada para execuo dos concretos, gua proveniente da
rede pblica de abastecimento.
3.3 Mistura do concreto e moldagem

dos corpos-de-prova
Nesta etapa foram moldados 62 corpos-de-prova cilndricos para
ensaios compresso e 10 amostras para ensaios trao, totalizando 72 corpos-de-prova moldados em formas metlicas com
dimenses de 10 cm de dimetro e 20 cm de altura, sendo dois
para cada idade, para seis temperaturas de cura.
A mistura foi realizada em betoneira com eixo vertical, com capacidade mxima de 56 litros e velocidade de rotao de 29 rotaes por minuto (RPM). O tempo da adio de todos os materiais
e homogeneizao da mistura para todas as moldagens foi de 18
2 min. A ordem de colocao dos materiais no misturador foi
seguinte:
a) Agregado grado;
b) 80% de gua;
c) Cimento;
d) Agregado mido;
e) 20% de gua;
Aps a etapa da mistura dos materiais na betoneira, foi medida
a consistncia do concreto atravs do ensaio do abatimento de
tronco de cone, conforme ABNT NBR NM 67:1996 [16]. Buscou-se fixar as relaes a/c, o tempo de cura e a consistncia dos
concretos, para avaliar unicamente a influncia das temperaturas
de cura nas propriedades mecnicas dos materiais. O tempo para
a realizao deste ensaio foi de 8 2 minutos. Os resultados dos
abatimentos e a relao gua / cimento esto apresentados na
Tabela 3. Os concretos analisados so similares, com todas as
79
Tabela 4 Condies de tempo e

temperatura ambiente de cura das moldagens
Moldagens
Figura 1 Evoluo da resistncia compresso

nos 7 primeiros dias, em cura controlada
Tempo e temperatura ambiente de cura

0 a 7 dias
8 a 91 dias
25 C
20 C
20 C
20 C
3
4
15 C
10 C
20 C
20 C
5 C
20 C
0 C
20 C
propriedades fixadas, para se analisar apenas a influncia da temperatura de cura nas propriedades finais do material.
Posteriormente, foram moldados os corpos-de-prova, em um
perodo de 25 2 minutos. Assim, o tempo total entre o contato
do cimento com a gua e o armazenamento dos corpos-de-prova no ambiente de cura com temperatura controlada foi de
51 6 minutos.
3.4 Condies de cura

A condio de cura dos corpos-de-prova foi realizada em ambiente
de baixas temperaturas controladas, simulando assim o comportamento do concreto quando submetido a essa incidncia, conforme
apresentado na Tabela 4, e umidade relativa de 50%. O equipamento
utilizado foi uma caixa vedada com controle de temperatura digital,
que permite a fixao da temperatura sem afetar a umidade relativa
do ar, por tempo pr-determinado, no caso do estudo, de 7 dias.
Nessa condio, ao conservar as amostras em temperaturas variadas de cura nos sete primeiros dias, pode-se analisar o comportamento das reaes iniciais, que ocorrem nas primeiras idades.
Aps os 7 primeiros dias, curados nas temperaturas de 0C a
25C, os cps foram acondicionados em cmara mida at as datas de rompimento.
4. Apresentao e Anlise dos Resultados

Aps os rompimentos dos corpos-de-prova nos ensaios mecnicos, pode-se comparar os resultados e chegar s concluses a
respeito do tema.
4.1 Ensaio de compresso de corpos-de-prova

cilndricos
A anlise dos resultados de resistncia compresso foi dividida
em duas etapas, por apresentar comportamentos diferentes. A primeira englobou de 1 at os 7 dias, enquanto a segunda etapa foi
realizada dos 8 aos 91 dias.
A Tabela 5 apresenta a evoluo das resistncias compresso
Tabela 5 Evoluo de resistncia

compresso em cura controlada
Tabela 6 Evoluo de resistncia

compresso em sala climatizada (20 C)
1 dia
(MPa)
7 dias
(MPa)
14 dias
(MPa)
28 dias
(MPa)
91 dias
(MPa)
25
11,4
21,5
25
24,3
26,9
29,2
20
11
23,4
20
27,5
31
31,4
15
8,3
24,4
15
28,7
21,7
27,8
10
6,5
21,7
10
27,3
31,8
28,7
3,5
18
20,9
35,6
37,6
1,2
18,8
30,3
38,6
38,8
80
nos 7 primeiros dias, quando as amostras foram condicionadas a

cura controlada em baixas temperaturas, variando de 25 a 0 C
em intervalos de 5 C.
Para uma melhor visualizao, apresenta-se os resultados na
Figura 1, relacionando a resistncia compresso, em MPa, no
eixo y com a idade de ruptura dos corpos-de-prova no eixo x,
em dias.
Neste primeiro perodo de cura pode-se observar uma maior variao de resistncia no primeiro dia de medio, que foi de 11,4
MPa para a temperatura de 25 C, a 1,2 MPa para a temperatura
de 0 C. As temperaturas de cura mais altas apresentaram resultados de resistncia compresso superiores e a diminuio de
resistncia foi gradativa at a temperatura mais baixa com o pior
desempenho, como era de se esperar.
No stimo dia essa variao foi de 24,4 MPa para a temperatura
de 15 C, a 18 MPa para a temperatura de 5 C. A diferena entre
o maior e o menor valor reduziu 37,25% em comparao com o
primeiro perodo. Ou seja, nos 7 dias de cura, a influncia das

em funo da temperatura de cura
50
fc (M P a)
Figura 2 Evoluo da resistncia

compresso em sala climatizada (20 C)
1d
40
7d
30
28d
14d
91d
20
Linear (1d)
Linear (7d)
10
Linear (14d)
Linear (28d)
0
0
10
15
20
25
30
Linear (91d)
t (C)
temperaturas iniciais de cura j foi menor, chegando, inclusive, a

sobrepor diferentes curvas. Ainda o ganho de resistncia foi mais
Tabela 7 Evoluo de resistncia compresso em 91 dias
1 dia
MPa
25
20
15
10
5
0
11,4
11
8,3
6,5
3,5
1,2
7 dias
14 dias
MPa
MPa
21,5
23,4
24,4
21,7
18
18,8
24,3
27,5
28,7
27,3
20,9
30,3
28 dias
MPa
91 dias
MPa
26,9
31
21,7
31,8
35,6
38,6
29,2
31,4
27,8
28,7
37,6
38,8
81
Tabela 8 Equaes de comportamento relacionando

a resistncia compresso e a temperatura de cura
Propriedade
Equao de comportamento
1 dia
fc = 0,4303*t + 1,6048
7 dias
fc = 1,1851*t + 18,986
14 dias
fc = -0,0503*t + 27,129
28 dias
fc = -0,4709*t + 36,819
91 dias
fc = -0,3857*t + 37,071
expressivo para os concretos curados a temperaturas inferiores,

chegando a resistncia do concreto curado a 0C aumentar cerca
de 1.600% entre o 1 e o 7 dia, contra um a ganho de 88% dos
curado a 25C.
Na Tabela 6 observa-se a evoluo da resistncia a partir dos 14
at os 91 dias de cura. Ainda ilustra-se, na Figura 2, os resultados
em forma de grfico, para melhor visualizao.
No segundo perodo de cura se observa a aproximao dos resultados a partir 7 dia, ocorrendo a inverso nos valores das resistncias a partir do 14 dia, com os concretos curados a temperaturas inferiores apresentando melhores resultados do que os
curados a temperaturas superiores.
Aos 28 dias a relao de resistncias com a idade anterior seguiu
a tendncia de maior crescimento para as temperaturas mais baixas. A variao de resistncia nesta idade foi de 21,7 MPa para
a temperatura de 15 C, a 38,6 MPa para a temperatura de 0C,
contemplando uma reduo de 43% no intervalo de 15 C, j para
o intervalo total de 25C a reduo de foi de 30%, similares aos
resultados da bibliografia referenciada. Para a idade de 91 dias a
relao de resistncias entre os concretos manteve a tendncia,
com os curados a zero e 5 C apresentando os maiores valores
de resistncia compresso. Apresenta-se ento a Tabela 7 com
todos os valores agrupados.
A Figura 3 mostra o comportamento em forma de grfico, onde observa-se as maiores resistncias compresso para os concretos
curados em temperaturas elevadas nas primeiras idades, com a
Tabela 9 Resistncia trao aos 28 dias

Resistncia trao
Temperatura (C)
28 dias (MPa)
82
25
20
9,0
10,5
15
10,0
10
10,8
11,3
13,2
Figura 5 Resistncia trao aos 28 dias
inverso ocorrendo aos 14 dias, mostrando que a hidratao lenta

uma vantagem ao longo do tempo para a estrutura de concreto.
Este comportamento fica evidenciado na Figura 4, que correlaciona a resistncia compresso, em MPa, no eixo y, com a temperatura de cura, em C, no eixo x, para as idades de rompimento
das amostras. Observa-se que o comportamento dos resultados
a 1 dia crescente, ou seja, quanto maior a temperatura de cura
maior a resistncia compresso nesta idade. Esta tendncia novamente observada aos 7 dias, porm de forma mais suave.
Aos 14 dias a equao praticamente linear, assim independente
da temperatura de cura a resistncia compresso similar. Enquanto que aos 28 e aos 91 dias o comportamento dos resultados
decrescente, ou seja, a resistncia compresso dos concretos
prejudicada pelo aumento da temperatura de cura.
Para todas as idades foram determinadas as equaes de comportamento, expressas na Tabela 8. Os comportamentos descritos podem ser analisados atravs das equaes, pois o primeiro
nmero, que multiplica o t (temperatura de cura) positivo para
resultados crescentes, a 1 e 7 dias, prxima a zero aos 14 dias,
mostrando a linearidade, e negativo, quando os valores decrescem aos 28 e 91 dias. Ainda a parcela isolada da equao representa o valor que a reta intersecciona o eixo y, ou seja, a resistncia compresso quando curado a 0C. Este valor maior para
as temperaturas de cura que propiciaram as maiores resistncias
compresso e menor para as que atingiram os menores valores,
ou seja, maior para as temperaturas maiores.
4.2 Ensaio de determinao da resistncia

trao por compresso diametral de

corpos-de-prova cilndricos
As apresentaes dos resultados de resistncia trao aos 28
dias de cura esto expressos na Tabela 9. A Figura 5 ilustra a
equao de comportamento, relacionando a resistncia trao,
no eixo y, com a temperatura de cura em C, no eixo x.
Observa-se que, novamente, os resultados aos 28 dias foram superiores aos concretos curados com temperaturas inferiores, ratificando o comportamento visto para as resistncias compresso.
A variao de resistncia para a idade foi de 9,0 MPa para a tem-
peratura de 25 C, a 13,2 MPa para a temperatura de 0 C, contemplando uma reduo de 31% no intervalo de 25 C, se aproximando dos resultados apresentados na bibliografia referenciada.
E o coeficiente de correlao foi de 0,85, considerado aceitvel
para este tipo de ensaio.
5. Concluses
Aps a realizao deste trabalho, pode-se observar que nas primeiras idades, entre 1 e 7 dia, as resistncias compresso dos
concretos curados a temperaturas mais elevadas foram superiores em relao s resistncias com temperaturas de cura inferiores, como era de se esperar. Nestas idades o aumento de resistncia foi devido ao elevado grau de hidratao dessas peas,
explicado pelo alto valor da energia de ativao. Destaca-se as
temperaturas de cura de 20 e 25C ao 1 dia, que apresentaram os
maiores valores. Porm, ao 7 dia, quando a energia de ativao
no exerce tanta influncia sob as peas, as resistncias tiveram
maior proximidade.
A partir do 14 dia verificou-se a inverso nos corpos-de-prova
de maiores resistncias, ou seja, os que inicialmente foram curados em temperaturas baixas obtiveram os melhores resultados,
porm com baixas diferenas entre as mesmas. Essas diferenas
aumentaram aos 28 dias, tendo o melhor desempenho as peas
inicialmente curadas a 0C. Por fim, comprova-se que, quanto
mais lenta for hidratao do cimento Portland, melhor ser a
formao de sua estrutura cristalina, justificando assim, desempenhos superiores para as peas curadas nas suas primeiras idades
em baixas temperaturas, ou seja, temperaturas que retardam e/ou
desaceleram o processo de hidratao.

[01]

[02]

[03]

[04]

[05]

[06]

[07]

[08]

HELENE, P. R. L.; LEVY S. M. Estado da arte do

concreto como material de construo. So Paulo:
Exacta. 2003. 8p.
Shoukry, S. N.; William, G. W.; Downie, B.;
Riad, M. Y.; Effect of moisture and temperature on the
mechanical properties of concrete. USAL: Elsevier,
2010.
ASSOCIAO BRASILEIRA DE NORMAS TCNICAS.
NBR 6118. Projeto de estruturas de concreto.
Procedimento. Rio de Janeiro: ABNT, 2007.
MEHTA, P. K.; MONTEIRO, P. J. M. Concreto:
microestrutura, propriedades e materiais. So Paulo:
Ibracon, 2008. 674 p.
ISAIA, G. C. Concreto: ensino, pesquisa e realizaes.
So Paulo: IBRACON, 2005. 1600p.
ANKA ILC; GORAN TURK; FRANCI KAVCIC;
GREGOR TRTNIK. New numerical procedure for the
prediction of temperature development in early age
concrete structures. USAL: Elsevier, 2009.
NEVILLE, A M. Propriedades do concreto. Traduo
Eng. Salvador E. Giammusso. So Paulo: Pini, 1997.
828 p.
KIRCHHEIM, A. P.; SOUZA, R. B.; DAL MOLIN, D. C.
C.; MONTEIRO, P. J. M. lcalis incorporados ao
aluminato triclcico: efeitos na hidratao. Ambiente
construdo. 2010. 14p.

[09] TAYLOR H. F. W. La quimica de los cementos. Bilbao:
Urmo, 1967. 512 p.

[10] Zhang, J.; Cusson, D.; Monteiro, P.; Harvey,

J.; New perspectives on maturity method and

approach for high. USAL: Elsevier, 2008.

[11] PINTO BARBOSA, M.; BERTOLUCCI, F. S.; PINTO,

R. C. A.; PERES, L. D. P. Avaliao da energia

aparente de ativao do cimento CP-II E CP-V com

adio de superplastificantes. Anais do VI Simpsio

EPUSP sobre Estruturas de Concreto, 2006. 15 p.

[12] Kim, J. K.; Han, S. H.; Song, Y. C.; Effect of

temperature and aging on the mechanical properties

of concrete Part I. Experimental results. USAL:
Elsevier, 2002.

[13] Husem, M.; Gozutok, S.; The effects of low

temperature curing on the compressive strength of

ordinary and high performance concrete. USAL:

Elsevier, 2004.

[14] CNOVAS, M. F. Patologia e terapia do concreto

armado. So Paulo: PINI, 1988. 522 p;

[15] DEMIREL, B.; KELESTEMUR, O.; Effect of elevated

temperature on the mechanical properties of concrete

produced with finely ground pumice and silica fume.
USAL: Elsevier, 2010.


NBR NM 67. Concreto Determinao da

consistncia pelo abatimento do tronco de cone. Rio

de Janeiro: ABNT, 1998.


NBR NM 23. Cimento. Determinao da massa

especfica. Rio de Janeiro: ABNT, 1998.


NBR NM 45. Determinao da Massa Unitria de

Agregados. Rio de Janeiro: ABNT, 2006.


NBR 9776. Determinao da Massa Especfica de

Agregados Midos - Por meio do frasco de Chapman.

Rio de Janeiro: ABNT, 1987.


NBR NM 248. Agregados Determinao da

composio granulomtrica. Rio de Janeiro:

ABNT, 2003.
83
Reliability of buildings in service limit state

for maximum horizontal displacements
Confiabilidade de edifcios no estado limite de servio
para deslocamentos horizontais mximos
A. G. B. CORELHANO a
anggio@sc.usp.br
M. R. S. CORRA b
mcorrea@sc.usp.br
A. T. BECK c
atbeck@sc.usp.br
Abstract
Brazilian design code ABNT NBR6118:2003 - Design of Concrete Structures - Procedures - [1] proposes the use of simplified models for the consideration of non-linear material behavior in the evaluation of horizontal displacements in buildings. These models penalize stiffness of columns
and beams, representing the effects of concrete cracking and avoiding costly physical non-linear analyses. The objectives of the present paper
are to investigate the accuracy and uncertainty of these simplified models, as well as to evaluate the reliabilities of structures designed following
ABNT NBR6118:2003 [1] in the service limit state for horizontal displacements. Model error statistics are obtained from 42 representative plane
frames. The reliabilities of three typical (4, 8 and 12 floor) buildings are evaluated, using the simplified models and a rigorous, physical and geometrical non-linear analysis. Results show that the 70/70 (column/beam stiffness reduction) model is more accurate and less conservative than the
80/40 model. Results also show that ABNT NBR6118:2003 [1] design criteria for horizontal displacement limit states (masonry damage according
to ACI 435.3R-68(1984) [10]) are conservative, and result in reliability indexes which are larger than those recommended in EUROCODE [2] for
irreversible service limit states.
Keywords: reinforced concrete, physical nonlinearity, structural reliability, plane frame structures, service limit state.
Resumo
A norma ABNT NBR6118:2003 - Projeto de Estruturas de Concreto - Procedimento - [1] prope o uso de modelos simplificados para a considerao da no-linearidade fsica na avaliao de deslocamentos em estruturas de concreto armado. Estes modelos penalizam a rigidez de pilares
e vigas, representando efeitos de fissurao do concreto e dispensando a realizao de anlises no-lineares fsicas de material. O presente
trabalho tem por objetivos investigar a incerteza dos modelos simplificados propostos nesta norma, bem como determinar a confiabilidade de
estruturas de edifcios projetadas segundo esta norma nos estados limites de servio para deslocamentos horizontais. Estatsticas de erro de
modelo so obtidas atravs da anlise de 42 prticos planos representativos. A confiabilidade de trs edifcios tpicos (de 4, 8 e 12 andares)
analisada, utilizando-se os modelos simplificados e a anlise no-linear fsica dita rigorosa. Os resultados mostram que o modelo 70/70 (penalizao de rigidez pilar/viga) menos conservador e mais preciso do que o modelo 80/40. Os resultados mostram ainda que os critrios de
verificao da norma ABNT NBR6118:2003 [1] para estado limite de servio de deslocamentos horizontais (tendo em vista fissurao da alvenaria
pelas prescries da ACI 435.3R-68(1984) [10]) so conservadores, e resultam em ndices de confiabilidade superiores aqueles sugeridos no
EUROCODE [2] para estados limites de servio irreversveis.
Palavras-chave: concreto armado, no-linearidade fsica, confiabilidade estrutural, prtico plano, estado limite de servio.
Doutorando em Engenharia de Estruturas, Departamento de Engenharia de Estruturas, Escola de Engenharia de So Carlos, Universidade
de So Paulo, anggio@sc.usp.br, Avenida Trabalhador Socarlense, 400, CEP 13.566.590, So Carlos, SP, Brasil
b
Professor Associado, Departamento de Engenharia de Estruturas, Escola de Engenharia de So Carlos, Universidade de So Paulo,
mcorrea@sc.usp.br, Avenida Trabalhador Socarlense, 400, CEP 13.566.590, So Carlos, SP, Brasil.
c
Professor Doutor, Departamento de Engenharia de Estruturas, Escola de Engenharia de So Carlos, Universidade de So Paulo,
atbeck@sc.usp.br, Avenida Trabalhador Socarlense, 400, CEP 13.566.590, So Carlos, SP, Brasil.
a
Received: 22 Sep 2011 Accepted: 10 Dec 2011 Available Online: 01 Feb 2012
2012 IBRACON
A. G. B. CORELHANO | M. R. S. CORRA | A. T. BECK
1. Introduction
In the design of reinforced concrete structures, it is common
practice to use simplified models which penalize the stiffness
of structural elements, in order to avoid non-linear material
analysis. A lot of research is dedicated to improve these simplified models. However, it is hard to find research works addressing the precision or errors of the simplified models. The
objective of the present article is to investigate the precision of
simplified stiffness-reducing models recommended in Brazilian code ABNT NBR6118:2003 [1] in the evaluation of horizontal displacements of plane reinforced concrete frames. The
investigation is based on a comparison, for a set of representative frames, of the displacements obtained using simplified
models and rigorous physical (material) non-linear analysis.
This article also investigates the reliability, with respect to serviceability limit states for horizontal displacements, of plane
frames representing usual reinforced concrete buildings. Reliability analyses are performed using rigorous non-linear material analysis and using the simplified models recommended
in ABNT NBR6118:2003 [1]. Geometrical non-linearities are
treated in a consistent way in all the analyses. Reliability analyses performed herein consider uncertainties in loads and in
the structural strengths, as well as the uncertainties originated
in the use of the simplified stiffness reduction models.
Non-linear structural analyses are performed using a finite
element code developed by the authors (CORELHANO [3]).
Reliability analyses are performed using the StRAnD software (BECK [4]).
2. Non-linear analyses in reinforced

concrete
2.1 Non-linear geometrical analysis
A formulation is considered based on second-order Piola Kirchhoff
tensors, developed by WEN & RAMIZADEH [5]. The deformation
tensor and deformation energy are given, respectively, by:
L
1
e x = u0 '- Y .v0 ''+ (v0 ') 2 .dx
L0
U=
(1)
2
1
1
2
(u0 '- Y .v0 ) + (v0 ') .dx E.dV
2V
L0
(2)
where:
x : longitudinal strains;
u0 and v0 : axial and transversal displacements;
Y : distance from a given fiber to the sessions gravity center (C.G.);
L : lenght of the element;
E : Youngs modulus;
U : internal strain energy.
Details of the formulation can be found in CORRA [6].
Figure 1 Details of a 2D beam element

formed by layers of steel and concrete
2.2 Rigorous material non-linear analysis

In this article, material non-linearities are considered by the method of layers, which allows independent constitutive models to be
considered for each layer. The element cross-section is divided in
steel and concrete slices, and the sum of the contribution of each
layer defines the behavior of the cross-section (Figure 1).
Properties of the cross-section (stiffness EA and EIz) are evaluated
from the sum of the contribution of each layer, at the integration
points at the extremes of each element:
EA = Ei . Ai
(3)
EI Z = Ei .I Z i
(4)
where:
Ai: area of the ith layer;
Ei: Youngs modulus of the ith layer;
Izi: inertia of the ith layer w.r.t. Z axis.
For the compressed concrete, the constitutive model of KENT &
PARK[7] is adopted, following Figure 2. Segment AB of this model
is described by:
2e e 2
s = fc ' c - c
e 0 e 0
(5)
85
Figure 2 Constitutive model for compressed (left), tensioned (centre) concretes and for steel (right)
fc' = fck + 3.5MPa
(6)
where:
f c ' : maximum compressive strength of concrete;
0 : specific deformation of concrete corresponding to maximum
tension;
c : specific deformation of concrete ;
: tension in concrete.
Segment BC is a line defined by the point of maximum compressive
strength and by the point corresponding to 50% of the maximum compressive strength. In segment CD it is admitted that the compressed con-
crete retains a tension corresponding to 20% of peak tension indefinitely.

For the tensioned concrete, the constitutive model of FIGUEIRAS [8] is adopted, following Figure 2 (centre). For the reinforcing
steel, an elasto-plastic model with hardening is considered (Figure 2, right). Details of the adopted constitutive models and of the
strategies used to solve the non-linear problem are presented in
CORELHANO [3].
2.3 Simplified material non-linear analysis

Brazilian code ABNT NBR6118:2003 [1] proposes two alternatives
for the simplified material non-linear analysis of reinforced concrete structures. These models penalize stiffness of the cross-section, in order to take into account, in a simplified way, the effects of
concrete cracking. In the first model, bending stiffness of columns
Figure 3 Geometry of the studied frames: 4 floors (left), 8 floors (centre),

and 12 floors (right), all measures in cm
86
Table 1 Evaluated horizontal displacements and model error samples
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Frame
fck
(MPa)
Reinforcement
ratio
4 floors / 1 bay
4 floors / 1 bay
4 floors / 1 bay
4 floors / 1 bay
4 floors / 1 bay
4 floors / 1 bay
8 floors / 3 bays
8 floors / 3 bays
8 floors / 3 bays
8 floors / 3 bays
8 floors / 3 bays
8 floors / 3 bays
8 floors / 3 bays
8 floors / 3 bays
8 floors / 3 bays
8 floors / 3 bays
8 floors / 3 bays
8 floors / 3 bays
8 floors / 3 bays
8 floors / 3 bays
8 floors / 3 bays
8 floors / 3 bays
8 floors / 3 bays
8 floors / 3 bays
12 floors / 3 bays
12 floors / 3 bays
12 floors / 3 bays
12 floors / 3 bays
12 floors / 3 bays
12 floors / 3 bays
12 floors / 3 bays
12 floors / 3 bays
12 floors / 3 bays
12 floors / 3 bays
12 floors / 3 bays
12 floors / 3 bays
12 floors / 3 bays
12 floors / 3 bays
12 floors / 3 bays
12 floors / 3 bays
12 floors / 3 bays
12 floors / 3 bays
30
30
35
35
40
40
23
23
23
23
23
23
30
30
30
30
30
30
40
40
40
40
40
40
22
22
22
22
22
22
30
30
30
30
30
30
40
40
40
40
40
40
High
High
Medium
Medium
Low
Low
Low
Low
Medium
Medium
High
High
Low
Low
Medium
Medium
High
High
Low
Low
Medium
Medium
High
High
Medium
Medium
High
High
Low
Low
High
High
Medium
Medium
Low
Low
Medium
Medium
High
High
Low
Low
Analysis
Model error
70/70
u (cm)
80/40
u (cm)
Rigorous
u (cm)
70/70
rigorous
u
/
80/40
simplifyed
u
1.51
2.13
1.39
1.96
1.3
1.84
1.84
2.62
4.07
5.91
6.16
9.15
1.62
2.31
3.56
5.15
5.36
7.9
1.4
1.98
3.05
4.39
4.57
6.68
4.67
6.7
5.52
7.97
4.33
6.2
4.28
5.6
3.92
6.4
3.63
5.16
3.37
4.8
3.97
5.68
3.12
4.45
1.94
2.74
1.79
2.53
1.67
2.36
2.57
3.68
5.26
7.73
7.86
11.85
2.26
3.23
4.6
6.71
6.82
10.1
1.94
2.77
3.93
5.7
5.79
8.54
6.34
9.18
7.15
10.45
5.97
8.63
5.3
7.63
5.98
8.65
4.99
7.18
4.55
6.52
5.13
7.38
4.29
6.14
1.35
2.61
1.27
2.51
1.22
2.13
1.53
2.56
3.05
5.06
4.19
6.74
1.43
2.44
3.12
5.2
4.42
7.2
1.08
1.87
2.5
4.29
3.71
6.14
4.07
7.07
4.12
7.03
3.92
6.8
3.54
6.35
3.72
6.58
3.4
6.07
2.64
3.56
2.88
3.72
2.52
3.38
0.89
1.23
0.91
1.28
0.94
1.16
0.83
0.98
0.75
0.86
0.68
0.74
0.88
1.06
0.88
1.01
0.82
0.91
0.77
0.94
0.82
0.98
0.81
0.92
0.87
1.06
0.75
0.88
0.91
1.10
0.83
1.13
0.95
1.03
0.94
1.18
0.78
0.74
0.73
0.65
0.81
0.76
0.70
0.95
0.71
0.99
0.73
0.90
0.60
0.70
0.58
0.65
0.53
0.57
0.63
0.76
0.68
0.77
0.65
0.71
0.56
0.68
0.64
0.75
0.64
0.72
0.64
0.77
0.58
0.67
0.66
0.79
0.67
0.83
0.62
0.76
0.68
0.85
0.58
0.55
0.56
0.50
0.59
0.55
87
Figure 4 Histograms and probability distribution functions of model error variables:

70/70 stiffness reduction (left), 80/40 column/beam stiffness reduction (right)
and beams are multiplied by 0,70. In the second model, equivalent

stiffness are obtained by multiplying the stiffness of columns and
beams by 0,80 and 0,40, respectively. In this article, these models
are referred to as 70/70 and 80/40, respectively.
The secant Youngs modulus of concrete is:
Esec = 0.85 5600 fck
(7)
where:
Esec: secant Youngs modulus;
fck: characteristic concrete resistance at 28 days.
3. Model errors
The simplified stiffness reducing models proposed in ABNT
NBR6118:2003 [1] are, naturally, approximations of reality. A
variable that measures the accuracy or precision of these, called
model error, is obtained by dividing the displacements obtained
via a rigorous material non-linear analysis by the displacements
obtained using the simplified model (OLIVEIRA et al. [9]):
EM =
urigorous
usimplifyed
(8)
This is a random variable as, for different structures, the simplified

model can be more or less precise. One sample (set of observations) of the model error random variable is obtained by evaluating
equation (8) for a set of different structural configurations. In this article, a sample of the two model error random variables is obtained by
evaluating 42 representative plane frames of different geometries,
materials and reinforcement ratios. Frames of four, eight and twelve
88
floors are considered, with one to three bays. The studied frames
are variations from the frames represented in Figure 3. Concrete
resistances varied from 20 to 40 MPa. Three reinforcement ratios
were considered: low, medium and high. The low reinforcement ratio
is close to the lower limit, medium is around 2% and high is close to
the upper limit (3 to 4%) allowed in ABNT NBR6118:2003 [1]. Vertical loading was determined based on the process of influence areas
(slabs, beams, columns, walls and coverings). Accidental load was
adopted as 1,5 kN/m in the influence area. Details of the studied
frames are presented in Table 1. The table also presents the model
error observations obtained for these frames.
Figure 4 shows the histograms that were obtained from the model
error samples, as well as the probability distribution functions that
were adjusted to the data. For the simplified model with 70/70 stiffness reduction, a Normal distribution was obtained with parameters:
EM70/70~N(=0.908, =0.150)
(9)
For the simplified model with 80/40 column/beam stiffness reduction, a Normal distribution was obtained with parameters:
EM80/40~N(=0.682, =0.111)
(10)
The coefficient of variation (c.o.v) is similar for both models error

variables (/=0.16).
The Normal distribution resulted in a good fit for both variables, as
indicated by the statistics shown in Table 2. For both cases, the
Normal distribution passed the Kolmogorov-Smirnov, Anderson
Darling and Chi-square goodness-of-fit tests.
Table 2 Statistics of goodness-of-fit tests for the Normal distribution

Model
error
Distribution
70/70
80/40
Normal
Normal
Kolmogorov-Smirnov
Goodness-of-fit test
Anderson Darling
0.10675
0.10290
The studied model errors (Eq. 8) compare displacements which

represent load effects. Hence, model error values smaller than
unity indicate a conservative model, as usimplified>urigorous. It can be
observed that both models are conservative, on average, as the
means are smaller than unity ( < 1). The 70/70 is moderately conservative, with mean just below unity, whereas the 80/40 model is
very conservative, with mean equal to 0.682. Since the c.o.v. is the
same for both variables (/=0.16), one concludes that the 70/70
model is more precise.
4. Structural reliability analysis

4.1 Design and verification of the frames
For the reliability analysis with respect to maximum displacements limit state, three frames were designed: with four, eight
and twelve floors (following Figure 3). Design of the frames followed guidelines of ABNT NBR6118:2003 [1] for ultimate limit
states. Once the proportioning of the frames was complete (for
ultimate limit states), their flexibility was increased until they
0.53353
0.58495
Chi Squared
1.7787
1.4824
reached the maximum horizontal displacement allowed by ABNT

NBR6118:2003 [1]. The verification with respect to horizontal displacements was made for frequent load combination, with maximum displacement of H/1700, where H is the total height of the
building. For the frequent load combination, one has:
Fser = FgiK + y 1 Fq1K + y 2 FqjK
(11)
where:
FgiK : permanent actions;
Fq1K : main variable action;
FqjK : secondary variable action;
1 : combination coefficient for the main variable action;
2 : combination coefficient for the secondary variable action.
For the studied buildings, only one equation is obtained, as the
combination factor for wind, when considered the secondary action, is null. Hence, one obtains:
Table 3 Characteristic and nominal resistance and load values considered in design
Variable
Symbol
Concrete strength
Dead load
Life load
Wind load at the floors
Characteristic or nominal values

4 floors
8 floors
12 floors
fck
25 MPa
25 MPa
30 MPa
Dn
Ln
Wn
24 kN/m
7.5 kN/m
13.5 kN
25.5 kN/m
7.5 kN/m
11.4 kN
22 kN/m
6 kN/m
13.5 kN
Table 4 Horizontal displacements at the top of the buildings

Floors
Height H (m)
4
8
12
14.4
28.8
43.2
Displacement at top (mm)

70/70
80/40
Limit displacement (H/1700)
6.62
12.09
18.29
8.10
16.50
25.05
8.47
16.94
25.41
89
(12)
Fser = Dn + 0.3Wn + 0.3Ln
where:
Fser : combined action value for service limit states;
Dn : nominal value of dead load;
Wn : nominal value of wind action;
Ln : nominal value of life action.
Table 3 summarizes the characteristic values (fck) and nominal
load values (Dn, Ln, Wn) used to verify the frames in the service limit
state. Table 4 shows the results obtained, in terms of the horizontal
displacements at the top of the studied buildings. It can be observed in this table that the representative frames were designed
for maximum flexibility.
4.2 Data for reliability analysis

For the service limit state related to horizontal displacements, the
failure condition is given by a displacement at the top of the building larger than H/500. This displacement is related to damage of
the masonry. This limit, indicated by ACI 435.3R-68(1984) [10], is
virtually equivalent to the H/1700 limit considered in the Brazilian
code [1], when the load combination factor for wind load is 0.3 and
the structural response is linear. Hence, the (service) limit state
equation for horizontal displacements is:
g (EM , f c , D, L, W ) = EM . u evaluated ( f c , D, L, W ) - H / 500
(13)
where EM, fc, D, L e W are the random variables of the problem,

described in Table 5. The parameters and probability distribution
functions of actions (D, L e W) are evaluated as indicated in Table
5, using the nominal values shown in Table 3.
The reduced loads for service verification (Eq. 12) correspond to

frequent load combinations, to which the structure will be exposed
during its design life. In principle, the reliability analysis for service
limit state could be performed for frequent loads, by combining the
arbitrary-point-in-time life load with the annual maximum of the
wind load. This analysis would result in an annual failure probability, which would have to be compared with the annual target for
irreversible service limit state (target=2.9 following the EUROCODE
[2]). Alternatively, the 50-year maximum of these actions can be
considered, in order to evaluate reliability for the same period (design life of the structure). In this case, the target reliability following the EUROCODE [2] is target=1.5 (for irreversible service limit
states). In the first case, the probability being evaluated is the probability that the limit state will occur any year during the structures
life. In the second situation, one calculates the probability of the
limit state occurring at least once during the buildings design life.
In this article, the second situation is adopted, as it is considered
to be more representative of the desired situation for a building (no
damage to masonry during the structures lifetime)
Reliability analyses, considering extreme actions, are made for two
load combinations: the first considers the 50-year extreme of the
life load, combined with the annual maximum of the wind load;
the second considers the 50-year extreme wind combined with
the arbitrary-point-in-time of the life load (the value at any point in
time). These combinations are usual, when time-dependent reliability problems are converted in time-independent problems (ELLINGWOOD et al.[11], BECK & SOUZA JR, [12]). The parameters
and probability distributions of: the 50 year extreme live and wind
loads, the annual maximum wind loads and the arbitrary-point-intime life load are presented in Table 5.
Following the First-Order Reliability Method (FORM), failure probabilities are evaluated by:
Pf =
f X ( x )dx F ( - b )
(14)
g ( x )< 0
where X is the vector of random variables, g(x) is the limit state
Table 5 Random variables, their parameters and distributions

Random varaible
Distrib.
Mdia
Desviopadro
C.V.
Fonte
70/70 model error
Normal
0.908
0.150
0.165
this work
80/40 model error
Normal
0.682
0.111
0.162
this work
fc
Normal
fck + 1.65.
4.00 MPa
0.150
MELCHERS [13]
Dead load
Normal
1.05 Dn
0.105 Dn
0.100
ELLINGWOOD et al.[11]
Arbitrary-point-in-time life load Gamma
0.25 Ln
0.148 Ln
0.55
50-year extreme life load

Annual extreme wind load
Gumbel
Gumbel
1.00 Ln
0.33 Wn
0.250 Ln
0.155 Wn
0.25
0.47
BECK & SOUZA JR. [12]
50-year extreme wind load
Gumbel
0.90 Wn
0.306 Wn
0.34
90
Table 6 Results for 50-year extreme live load combined with annual maximum wind load
N. Floors
4
12
Sensitivity coefficients
fc
D
L
Model
baprox
Pf
70/70
4.019
2.92 E-5
0.305
0.025
0.0
0.0
-0.669
80/40
4.265
9.97 E-6
0.301
0.027
0.0
0.0
-0.672
70/70
4.292
8.84 E-6
0.344
0.026
0.0
0.0
-0.630
80/40
4.331
7.43 E-6
0.308
0.030
0.0
0.0
-0.662
70/70
4.116
1.92 E-5
0.302
0.023
0.0
0.0
-0.675
80/40
4.159
1.60 E-5
0.292
0.020
0.0
0.0
-0.688
equation (Eq. 13), ( ) is the cumulative standard Gaussian distribution function and is the reliability index. In this article, equation
(14) is solved by the FORM method (MELCHERS, [13]), using the
StRAnD software (BECK [4]). In the FORM method, the original
problem is transformed to the standard normal space, and solved
as a restricted optimization problem: the reliability index is the
smallest distance between the limit state equation and the origin
of the standard normal space. The reliability index is related to the
failure probability by means of:
(15)
b = - F -1 ( Pf )
4.3 Reliability analyses using simplified stiffness

reducing models: results
Tables 6 and 7 show results of the reliability analysis, for the
50-year extreme live load combined with annual maximum wind
load (Table 6) and using the 50-year extreme wind load com-
EM
bined with the arbitrary-point-in-time live load (Table 7). Results

refer to reliability analysis using the simplified models with stiffness reduction and non-linear geometrical analysis. It can be observed that the load combination involving the 50-year extreme
wind load (Table 7) leads to larger failure probabilities than the
combination involving the 50-year extreme live load. This is to
be expected, as the wind load acts directly in the direction of the
calculated displacements.
The term failure, in this context, is used between quotes, as it
represents failure in respecting the constraint of maximum displacement (H/500) which, in theory, corresponds to a state of
masonry damage. This damage is an irreversible limit state. As
a reference, annex C of the EUROCODE [2] suggests a target
reliability index of target=1.5 for irreversible limit states and for
50 years reference. The reliability indexes found in this article,
which correspond to very flexible structures designed following
ABNT NBR6118:2003 [1], are slightly larger that this target value.
Hence, failure probabilities can be considered acceptable. These
results show that the horizontal displacement verifications of
Brazilian code ABNT NBR6118:2003 [1] (Eq. 13), together with
the maximum allowed displacement of H/1700 (for frequent load
combinations) are conservative.
Table 7 Results for 50-year extreme wind load combined with arbitrary-point-in-time life load
N. Floors
4
8
12
fc
D
L
Model
baprox
Pf
70/70
2.127
1.60 E-2
0.226
0.029
0.0
0.0
-0.745
80/40
2.369
8.90 E-3
0.216
0.030
0.0
0.0
-0.754
70/70
2.441
7.33 E-3
0.238
0.030
0.0
0.0
-0.732
80/40
2.453
7.08 E-3
0.218
0.032
0.0
0.0
-0.750
70/70
2.235
1.27 E-2
0.232
0.022
0.0
0.0
-0.746
80/40
2.253
1.21 E-2
0.214
0.023
0.0
0.0
-0.763
EM
91
Table 8 Results for 50-year extreme live load combined with annual maximum wind load
L
0.079
0.000
0.000
-0.921
2.64E-07
0.267
-0.004
-0.002
-0.727
1.46E-07
0.050
0.000
0.000
-0.950
N. Floors
brigorous
Pf
fc
4.957
3.58E-07
5.016
12
5.129
Tables 6 and 7 also show sensitivity coefficients of the problems

random variables. These coefficients show the relative importance
of each random variable towards failure. As expected, the horizontal wind action has the largest contribution towards this displacement failure. Uncertainty in concrete strength, which by way of Eq.
(7) affects the Youngs modulus, has minimal contribution. The
model error random variables have a significant contribution (from
21 to 34%) in the evaluated failure probabilities.
It is significant to note that reliability indexes obtained using the
simplified stiffness-reducing models 70/70 and 80/40 are similar.
This result is, in part, consequence of incorporating model error
random variables in the analysis. In the next section, it is verified
whether these reliability indexes agree with results of a rigorous
physical and geometrical non-linear analysis.
4.4 Reliability analysis using rigorous physical

and geometrical non-linear analysis: results
Tables 8 and 9 show results of the rigorous reliability analysis, using the 50-year extreme live load (Table 8) and the 50-year extreme wind load (Table 9). Results in both tables correspond to
reliability analysis performed using rigorous physical and geometrical non-linear analysis.
As in the simplified analysis, sensitivity coefficients show the same
behavior, with the wind load being the most important variable for
this horizontal displacement failure mode.
It is observed that reliability indexes obtained with the rigorous
analysis are significantly larger than those using the simplified
stiffness-reducing models. For the combination involving 50-year
extreme live loads (less relevant), reliability indexes obtained in
the rigorous analysis were larger than for the simplified analysis.
For the combination involving 50-year extreme wind loads, different results were obtained for the three frames studied. For the four
and twelve floor buildings, larger reliability indexes were obtained.
For the eight floor building, a smaller reliability index was obtained
in the rigorous analysis. This result may be a particularity of the
frames studied. However, since the rigorous physical analysis is
more precise, one can conclude that the stiffness-reducing simplified models can be used for design, but cannot be used for reliability analysis (even if model errors are considered).
Since reliability indexes found in the rigorous analysis are all
larger than =1.5, one concludes that the design criteria of ABNT
NBR6118:2003 [1] with respect to the service limit state for horizontal displacements (masonry damage) are conservative.
5. Concluding remarks
This article presented a study of model errors for the simplified
stiffness-reducing models proposed in ABNT NBR6118:2003 [1] for
evaluation of horizontal displacements of reinforced concrete plane
frames. A limited analysis composed of 42 plane frames of four,
eight and twelve floors has shown that the 70/70 model is more precise than the 80/40 (column/beam stiffness reduction) model.
Reliability analyses for service limit state of horizontal displacements (masonry damage) were made using the simplified stiffnessreducing models and using rigorous physical non-linear analysis.
It was observed that the simplified models are appropriate for a
verification of the structural design, but are not suitable for reliability analyses (even if model errors are considered).
It was found that the load combination involving the 50-year extreme live load is not relevant for the limit state of horizontal displacements, even when geometrical non-linear effects are considered. The combination involving the 50-year extreme wind and the
arbitrary-point-in-time live load is more critical and leads to smaller
reliability indexes. Since these reliability indexes are larger than
EUROCODE-recommended values, it is concluded that the design
and verification criteria of Brazilian code ABNT NBR6118:2003 [1]
for horizontal displacements (Eq. 13 and the maximum allowed
Table 9 Results for 50-year extreme wind load combined with arbitrary-point-in-time life load
N. Floors
92
brigorous
Pf
fc
L
2.747
3.00E-03
0.088
0.000
0.000
-0.912
2.293
1.09E-02
0.000
0.000
0.000
-1.000
12
2.955
1.56E-03
0.057
0.000
0.000
-0.943
displacement limit of H/1700) are conservative, and result in acceptable reliability indexes for the irreversible limit state of masonry
damage.
6. Acknowledgments
The authors express their gratitude to CAPES and CNPq for the
funding to this research.

Structural Design Codes, Journal of the Brazilian

Society of Mechanical Sciences and Engineering,

Vol. 32, No.2, p 119-127, 2010.

[13] MELCHERS, R.E. Structural Reliability Analysis

and Prediction, 2ndEdition, John Wiley and Sons,
NY, 1999
7. References


NBR 6118: Projeto de estruturas de concreto

Procedimento. Rio de Janeiro, 2003.

[02] EUROCODE, 2001. prEN 1990: Basis of Structural

Design Annex C: Basis for Partial Factor Design and

Reliability Analysis, European Committee for
Standardization, Brussels.

[03] CORELHANO, A.G.B. Anlise no linear geomtrica

e fsica de ncleos de edifcios altos em concreto

armado, So Carlos, 2010, Dissertao (Mestrado) -

Escola de Engenharia de So Carlos, Universidade de
So Paulo.

[04] BECK, A.T. StRAnD: Structural Reliability Analysis

and Design, Manual do Usurio, Verso 1.03, 2008,

Departamento de Engenharia de Estruturas, Escola

So Paulo.

[05] WEN, R.K.; RAHIMZADEH, J. Nonlinear elastic frame

analysis by finite element. Journal of Structural

Engineering, v.109, n.8, p.1952-1971, 1983.

[06] CORRA, M.R.S. Aperfeioamento de modelos

usualmente empregados no projeto de sistemas

estruturais de edifcios, So Carlos, 1991, Tese

(Doutorado) Escola de Engenharia de So Carlos,

Universidade de So Paulo.

[07] KENT, D. C.; PARK, R. Flexural members with

confined concrete. Journal of Structural Division

ASCE, v.97, n. ST7, p1969-1990, July, 1971

[08] FIGUEIRAS, J.A. Ultimate Load analysis of anisotropic

and reinforced concrete plates and shells. Swansea.

Ph.D. Thesis, - Department of Civil Engineering,

University College of Swansea, 1983.

[09] OLIVEIRA, W.L.; BECK, A.T.; EL DEBS, A.L.H.C.

Safety evaluation of circular concrete-filled steel

columns de-signed according to Brazilian building

code NBR 8800:2008, IBRACON Structures and

Materials Journal, vol.1, p. 212-236.

[10] AMERICAN CONCRETE INSTITUTE. Commitee 435

(1984). Allowable Deflections. Detroit.

(ACI 435.3R-68, revised 1984).

[11] ELLINGWOOD, B.; GALAMBOS, T. V.;

MACGREGOR, J. G.; CORNELL,C. A. Development

of a Probability Based Load Criterion for American

National Standard A58. US National Bureau of

Standard Special Publication 577, Washington,
D.C., 1980.

[12] BECK, A.T.; SOUZA JR., A.C.S. A First Attempt

Towards Reliability-based Calibration of Brazilian
93
Reliability of buildings in service limit state

for maximum horizontal displacements
Confiabilidade de edifcios no estado limite de servio
para deslocamentos horizontais mximos
A. G. B. CORELHANO a
anggio@sc.usp.br
M. R. S. CORRA b
mcorrea@sc.usp.br
A. T. BECK c
atbeck@sc.usp.br
Abstract
Brazilian design code ABNT NBR6118:2003 - Design of Concrete Structures - Procedures - [1] proposes the use of simplified models for the consideration of non-linear material behavior in the evaluation of horizontal displacements in buildings. These models penalize stiffness of columns
and beams, representing the effects of concrete cracking and avoiding costly physical non-linear analyses. The objectives of the present paper
are to investigate the accuracy and uncertainty of these simplified models, as well as to evaluate the reliabilities of structures designed following
ABNT NBR6118:2003 [1] in the service limit state for horizontal displacements. Model error statistics are obtained from 42 representative plane
frames. The reliabilities of three typical (4, 8 and 12 floor) buildings are evaluated, using the simplified models and a rigorous, physical and geometrical non-linear analysis. Results show that the 70/70 (column/beam stiffness reduction) model is more accurate and less conservative than the
80/40 model. Results also show that ABNT NBR6118:2003 [1] design criteria for horizontal displacement limit states (masonry damage according
to ACI 435.3R-68(1984) [10]) are conservative, and result in reliability indexes which are larger than those recommended in EUROCODE [2] for
irreversible service limit states.
Keywords: reinforced concrete, physical nonlinearity, structural reliability, plane frame structures, service limit state.
Resumo
A norma ABNT NBR6118:2003 - Projeto de Estruturas de Concreto - Procedimento - [1] prope o uso de modelos simplificados para a considerao da no-linearidade fsica na avaliao de deslocamentos em estruturas de concreto armado. Estes modelos penalizam a rigidez de pilares
e vigas, representando efeitos de fissurao do concreto e dispensando a realizao de anlises no-lineares fsicas de material. O presente
trabalho tem por objetivos investigar a incerteza dos modelos simplificados propostos nesta norma, bem como determinar a confiabilidade de
estruturas de edifcios projetadas segundo esta norma nos estados limites de servio para deslocamentos horizontais. Estatsticas de erro de
modelo so obtidas atravs da anlise de 42 prticos planos representativos. A confiabilidade de trs edifcios tpicos (de 4, 8 e 12 andares)
analisada, utilizando-se os modelos simplificados e a anlise no-linear fsica dita rigorosa. Os resultados mostram que o modelo 70/70 (penalizao de rigidez pilar/viga) menos conservador e mais preciso do que o modelo 80/40. Os resultados mostram ainda que os critrios de
verificao da norma ABNT NBR6118:2003 [1] para estado limite de servio de deslocamentos horizontais (tendo em vista fissurao da alvenaria
pelas prescries da ACI 435.3R-68(1984) [10]) so conservadores, e resultam em ndices de confiabilidade superiores aqueles sugeridos no
EUROCODE [2] para estados limites de servio irreversveis.
Palavras-chave: concreto armado, no-linearidade fsica, confiabilidade estrutural, prtico plano, estado limite de servio.
Doutorando em Engenharia de Estruturas, Departamento de Engenharia de Estruturas, Escola de Engenharia de So Carlos, Universidade
de So Paulo, anggio@sc.usp.br, Avenida Trabalhador Socarlense, 400, CEP 13.566.590, So Carlos, SP, Brasil
b
Professor Associado, Departamento de Engenharia de Estruturas, Escola de Engenharia de So Carlos, Universidade de So Paulo,
mcorrea@sc.usp.br, Avenida Trabalhador Socarlense, 400, CEP 13.566.590, So Carlos, SP, Brasil.
c
Professor Doutor, Departamento de Engenharia de Estruturas, Escola de Engenharia de So Carlos, Universidade de So Paulo,
atbeck@sc.usp.br, Avenida Trabalhador Socarlense, 400, CEP 13.566.590, So Carlos, SP, Brasil.
a
Received: 22 Sep 2011 Accepted: 10 Dec 2011 Available Online: 01 Feb 2012
2012 IBRACON
1. Introduo
prtica comum no projeto de estruturas de concreto armado o
uso de modelos simplificados que penalizam a rigidez dos elementos de uma estrutura, de forma a evitar a necessidade de
anlises no-lineares fsicas de material. Muitos trabalhos de pesquisa em concreto armado tm por objetivo estabelecer ou aperfeioar parmetros destes modelos simplificados. No entanto, a
maioria desses trabalhos no enderea os erros, ou incertezas,
decorrentes do uso de modelos simplificados (de penalizao da
rigidez), em detrimento de anlises no-lineares fsica rigorosas.
O presente trabalho tem por objetivo investigar a incerteza dos
modelos simplificados propostos na norma ABNT NBR6118:2003
[1] na avaliao de deslocamentos horizontais em prticos planos
de concreto armado. Esta investigao feita com base na comparao, para um conjunto de prticos planos representativos,
dos deslocamentos horizontais obtidos via anlises simplificada e
no-linear rigorosa (fsica).
Este trabalho tambm investiga a confiabilidade, quanto ao estado
limite de servio para deslocamentos horizontais, de prticos planos pertencentes a edifcios usuais de concreto armado. Anlises
de confiabilidade so realizadas utilizando anlise no linear fsica
com modelos simplificados (de penalizao de rigidez) e rigorosos
propostos na norma ABNT NBR6118:2003 [1]. A no-linearidade
geomtrica tratada de forma consistente em todas as anlises.
Nas anlises de confiabilidade so consideradas as incertezas nas
aes, nas propriedades dos materiais bem como as incertezas nos
modelos de penalizao da rigidez nas anlises simplificadas.
As anlises estruturais no-lineares so realizadas utilizando programa computacional (elementos finitos) desenvolvido pelos autores (CORELHANO [3]). As anlises de confiabilidade so realizadas utilizando o software StRAnD (BECK [4]).
2. Anlise no-linear em concreto armado

2.1 Anlise no-linear geomtrica
empregada uma formulao baseada no Tensor de Piola Kirchhoff de segunda espcie, desenvolvida por WEN & RAMIZADEH
[5], em que o tensor de deformaes e a expresso da energia de
deformao so dados, respectivamente, por:
e x = u0 '- Y .v0 ''+
1
(v0 ') 2 .dx
L 0
(1)
2
1
1
U = (u0 '- Y .v0 ) + (v0 ') 2 .dx E.dV
2V
L0
(2)
onde:
x : deformao longitudinal;
u0 e v0 : deslocamentos axial e transversal;
Y : distncia da fibra considerada ao centro de gravidade (C.G.)
da seo;
Figura 1 Detalhe da seo transversal

de uma barra de prtico plano
L : comprimento do elemento;
E : mdulo de elasticidade;
U : energia interna de deformao.

Maiores detalhes sobre esta formulao podem ser encontrados
em CORRA [6].
2.2 Anlise no-linear fsica rigorosa
No presente trabalho, o tratamento da no-linearidade fsica feito atravs do mtodo das fatias, que permite o uso de modelos
constitutivos independendentes para cada material componente.
O mtodo das fatias consiste na diviso da seo transversal do
elemento de barra em fatias de ao e concreto, de tal forma que o
somatrio da contribuio de todas as camadas determina o comportamento da seo (Figura 1).
As propriedades da seo transversal (rigidezes EA e EIz) so calculadas atravs do somatrio da contribuio das camadas, nos
pontos de integrao, nas sees extremas de cada elemento. Assim, as propriedades da seo adotadas para o elemento so obtidas como a mdia entre as propriedades dos pontos de integrao:
EA = Ei . Ai
(3)
EI Z = Ei .I Z i
(4)
onde:
Ai: rea da fatia i;
Ei: mdulo de elasticidade da fatia i;
Izi: inrcia da fatia i segundo o eixo Z.
95
Figura 2 Curvas tenso x deformao para concreto comprimido ( esquerda),

concreto tracionado (centro) e ao ( direita)
Para o concreto comprimido adotado o modelo constitutivo proposto por KENT & PARK[7], conforme Figura 2. O trecho AB deste
modelo constitutivo dado por:
2e e 2
s = fc ' c - c
e 0 e 0
fc' = fck + 3.5MPa
(5)
(6)
onde:
f c ' : tenso de compresso mxima do concreto;
0 : deformao especfica no concreto correspondente mxima
tenso;
c : deformao especfica no concreto;
: tenso no concreto correspondente deformao.
O ramo BC da curva corresponde a uma reta cuja inclinao
definida determinando-se a deformao quando a tenso de compresso no concreto reduz-se a 50% da tenso de pico, e o trecho
CD admite que o concreto mantm uma tenso de 20% da tenso
de pico indefinidamente.
Para o concreto tracionado adotado o modelo proposto por FIGUEIRAS [8], e para o ao adotado um modelo elasto-plstico
com encruamento positivo. Na Figura 2 so mostrados os diagramas tenso x deformao destes modelos de concreto, bem
Figura 3 Elevaes e formas dos prticos com 4 pavimentos ( esquerda),

8 pavimentos (centro), e 12 pavimentos ( direita), medidas em cm
96
Tabela 1 Deslocamentos horizontais e amostras de erros de modelo
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Descrio
dos prticos
fck
(MPa)
Taxa
armadura
4 pav. / 1 lance
4 pav. / 1 lance
4 pav. / 1 lance
4 pav. / 1 lance
4 pav. / 1 lance
4 pav. / 1 lance
8 pav. / 3 lances
8 pav. / 3 lances
8 pav. / 3 lances
8 pav. / 3 lances
8 pav. / 3 lances
8 pav. / 3 lances
8 pav. / 3 lances
8 pav. / 3 lances
8 pav. / 3 lances
8 pav. / 3 lances
8 pav. / 3 lances
8 pav. / 3 lances
8 pav. / 3 lances
8 pav. / 3 lances
8 pav. / 3 lances
8 pav. / 3 lances
8 pav. / 3 lances
8 pav. / 3 lances
12 pav. / 3 lances
12 pav. / 3 lances
12 pav. / 3 lances
12 pav. / 3 lances
12 pav. / 3 lances
12 pav. / 3 lances
12 pav. / 3 lances
12 pav. / 3 lances
12 pav. / 3 lances
12 pav. / 3 lances
12 pav. / 3 lances
12 pav. / 3 lances
12 pav. / 3 lances
12 pav. / 3 lances
12 pav. / 3 lances
12 pav. / 3 lances
12 pav. / 3 lances
12 pav. / 3 lances
30
30
35
35
40
40
23
23
23
23
23
23
30
30
30
30
30
30
40
40
40
40
40
40
22
22
22
22
22
22
30
30
30
30
30
30
40
40
40
40
40
40
Alta
Alta
Mdia
Mdia
Baixa
Baixa
Baixa
Baixa
Mdia
Mdia
Alta
Alta
Baixa
Baixa
Mdia
Mdia
Alta
Alta
Baixa
Baixa
Mdia
Mdia
Alta
Alta
Mdia
Mdia
Alta
Alta
Baixa
Baixa
Alta
Alta
Mdia
Mdia
Baixa
Baixa
Mdia
Mdia
Alta
Alta
Baixa
Baixa
Tipo de Anlise
Erro de Modelo
70/70
u (cm)
80/40
u (cm)
Rigorosa
u (cm)
70/70
rigoroso
u
/
80/40
simplificado
u
1,51
2,13
1,39
1,96
1,3
1,84
1,84
2,62
4,07
5,91
6,16
9,15
1,62
2,31
3,56
5,15
5,36
7,9
1,4
1,98
3,05
4,39
4,57
6,68
4,67
6,7
5,52
7,97
4,33
6,2
4,28
5,6
3,92
6,4
3,63
5,16
3,37
4,8
3,97
5,68
3,12
4,45
1,94
2,74
1,79
2,53
1,67
2,36
2,57
3,68
5,26
7,73
7,86
11,85
2,26
3,23
4,6
6,71
6,82
10,1
1,94
2,77
3,93
5,7
5,79
8,54
6,34
9,18
7,15
10,45
5,97
8,63
5,3
7,63
5,98
8,65
4,99
7,18
4,55
6,52
5,13
7,38
4,29
6,14
1,35
2,61
1,27
2,51
1,22
2,13
1,53
2,56
3,05
5,06
4,19
6,74
1,43
2,44
3,12
5,2
4,42
7,2
1,08
1,87
2,5
4,29
3,71
6,14
4,07
7,07
4,12
7,03
3,92
6,8
3,54
6,35
3,72
6,58
3,4
6,07
2,64
3,56
2,88
3,72
2,52
3,38
0,89
1,23
0,91
1,28
0,94
1,16
0,83
0,98
0,75
0,86
0,68
0,74
0,88
1,06
0,88
1,01
0,82
0,91
0,77
0,94
0,82
0,98
0,81
0,92
0,87
1,06
0,75
0,88
0,91
1,10
0,83
1,13
0,95
1,03
0,94
1,18
0,78
0,74
0,73
0,65
0,81
0,76
0,70
0,95
0,71
0,99
0,73
0,90
0,60
0,70
0,58
0,65
0,53
0,57
0,63
0,76
0,68
0,77
0,65
0,71
0,56
0,68
0,64
0,75
0,64
0,72
0,64
0,77
0,58
0,67
0,66
0,79
0,67
0,83
0,62
0,76
0,68
0,85
0,58
0,55
0,56
0,50
0,59
0,55
97
Figura 4 Histogramas e distribuies de probabilidades para erros de modelo:

reduo pilar/viga 70/70 ( esquerda), reduo pilar/viga 80/40 ( direita)
como do ao de reforo. Detalhes dos modelos e da estratgia

de soluo do problema no linear so apresentados em CORELHANO [3].
2.3 Anlise no-linear fsica simplificada

A norma ABNT NBR6118:2003 [1] prope duas alternativas para a
anlise no-linear fsica simplificada de estruturas de concreto armado. Estes modelos penalizam a rigidez dos membros para levar
em conta, de forma simplificada, os efeitos da fissurao do concreto. No primeiro modelo, as rigidezes flexo de pilares e vigas
so obtidas multiplicando-se os valores referentes seo bruta
por 0,70. No segundo modelo, as rigidezes equivalentes (fissuradas) de pilares e vigas so obtidas multiplicando-se as rigidezes
da seo bruta por fatores iguais a 0,80 e 0,40, respectivamente.
Neste trabalho, estes modelos so referenciados como 70/70 e
80/40, respectivamente.
O mdulo de elasticidade secante do concreto estimado como:
Esec = 0.85 5600 fck
(7)
onde:
Esec: mdulo de elasticidade secante;
fck: resistncia caracterstica do concreto aos 28 dias.
3. Erros de modelo
Os modelos simplificados propostos pela norma ABNT
NBR6118:2003 [1] so, por natureza, aproximados. Uma varivel
chamada de Erro de Modelo obtida dividindo-se os deslocamentos obtidos via anlise no-linear fsica rigorosa pelos deslocamentos obtidos via anlises simplificadas (OLIVEIRA et al. [9]):
98
EM =
urigorous
usimplifyed
(8)
Trata-se de uma varivel aleatria uma vez que, para diferentes

estruturas, o modelo simplificado pode ser mais ou menos preciso.
Uma amostra (conjunto de observaes) da varivel aleatria erro
de modelo obtida avaliando-se a equao (8) para diferentes
configuraes estruturais.
Neste trabalho, obtida uma amostra de erro de modelo correspondente a 42 prticos representativos de estruturas usuais,
variando-se geometria, material e taxa de armadura. So considerados prticos planos de quatro, oito e doze pavimentos, com
um ou trs lances de pilares. A geometria dos prticos estudados
representa variaes dos prticos apresentados na Figura 3. Foram considerados concretos com resistncias caractersticas nas
faixas de 20 a 40 MPa. Quanto s taxas de armadura dos pilares,
os prticos foram dimensionados em trs faixas: baixa, mdia, e
alta. A taxa de armadura baixa prxima armadura mnima; a
mdia tem taxas da ordem de 2% para os pilares e a taxa de
armadura mxima equivalente mxima permitida na ABNT
NBR6118:2003 [1], entre 3 e 4%. Os carregamentos verticais foram determinados pelo processo das reas de influncia (lajes, vigas, pilares, paredes e revestimentos). O carregamento acidental
foi adotado como 1,5 kN/m na rea de influncia. Detalhes das
configuraes dos prticos planos estudados so apresentados
na Tabela 1. Esta tabela tambm apresenta as observaes de
erro de modelo obtidas para cada prtico.
A Figura 4 ilustra os histogramas obtidos com base na amostra
de erros de modelo, bem como as distribuies de probabilidade
ajustadas a estes histogramas. Para o modelo simplificado com
reduo de rigidez 70/70 (pilar/viga), obteve-se ajuste a uma distribuio normal, com parmetros:
Tabela 2 Estatsticas dos testes de aderncia para a distribuio Normal

Erro de
Modelo
Distribuio
70/70
80/40
Normal
Normal
Estatsticas
Anderson Darling
Kolmogorov-Smirnov
0,10675
0,10290
(9)
EM70/70~N(=0.908, =0.150)
Para o modelo simplificado com reduo 80/40 (pilar/viga), obteve-se ajuste a uma distribuio normal, com parmetros:
(10)
EM80/40~N(=0.682, =0.111)
Chi Squared
0,53353
0,58495
1,7787
1,4824
Os erros dos modelo estudados (equao 8) comparam deslocamentos que representam efeitos de carregamentos. Portanto, valores menores do que a unidade indicam modelos conservadores,
isto , modelos que resultam em usimplificado>urigoroso. Observa-se que
os dois modelos estudados neste trabalho so conservadores, em
mdia, pois ambos apresentam mdia inferior unidade ( < 1). O
modelo 70/70 um pouco conservador, com mdia ligeiramente
menor do que um, enquanto o modelo 80/40 bastante conservador, com mdia igual a 0,682. Como o coeficiente de variao o
mesmo para os dois modelos (/=0.16), observa-se que o modelo 70/70 mais preciso, pois tem mdia mais prxima da unidade.
4. Anlises de confiabilidade estrutural

O coeficiente de variao (/=0.16) encontrado similar para as
duas variveis erro de modelo.
A distribuio Normal mostrou bom desempenho nos testes de
aderncia, conforme indicam as estatsticas apresentadas na Tabela 2. Em ambos os casos, a distribuio Normal foi aprovada nos
testes Kolmogorov-Smirnov, Anderson Darling e Chi-quadrado.
4.1 Dimensionamento e verificao dos prticos

Para fins de anlise de confiabilidade, em relao ao estado
limite de servio de deslocamento horizontal no topo, foram dimensionados trs prticos: de quatro, oito e doze andares (conforme Figura 3). O dimensionamento dos prticos foi realizado
Tabela 3 Valores de projeto de resistncia e aes usados no dimensionamento dos prticos

Varivel
Smbolo
Resistncia do concreto
Ao permanente
Ao varivel (acidental)
Ao de vento ao nvel dos pavimentos
Valores caractersticos ou nominais

4 pavimentos
8 pavimentos
12 pavimentos
fck
25 MPa
25 MPa
30 MPa
Dn
Ln
Wn
24 kN/m
7,5 kN/m
13,5 kN
25,5 kN/m
7,5 kN/m
11,4 kN
22 kN/m
6 kN/m
13,5 kN
Tabela 4 Deslocamentos horizontais no topo dos prticos

Nmero de
pavimentos
Altura H (m)
4
8
12
14,4
28,8
43,2
Deslocamentos no topo do prdio (mm)

70/70
80/40
Deslocamento limite (H/1700)
6,62
12,09
18,29
8,10
16,50
25,05
8,47
16,94
25,41
99
segundo as recomendaes da norma ABNT NBR6118:2003 [1]

para estado limite ltimo. Uma vez dimensionados os prticos,
buscou-se aumentar a sua flexibilidade, at se atingir o deslocamento horizontal mximo permitido em norma. A verificao
quanto ao deslocamento horizontal foi realizada para combinao freqente, com deslocamento mximo admitido igual a
H/1700, onde H a altura total do edifcio. Para a combinao
freqente, tem-se:
Fser = FgiK + y 1 Fq1K + y 2 FqjK
(11)
onde:
FgiK : aes permanentes;

Fq1K : ao varivel principal;
FqjK : aes variveis secundrias;
verificar nesta tabela que os prticos foram dimensionados para

flexibilidade mxima admitida em norma.
4.2 Dados para as anlises de confiabilidade

Para o estado limite de servio para deslocamentos horizontais, a condio de falha dada para um deslocamento no
topo do edifcio superior a H/500, deslocamento que leva a um
comprometimento das alvenarias de fechamento por fissurao. Este limite, indicado pelo ACI 435.3R-68(1984) [10],
praticamente equivalente ao limite H/1700 com ponderao da
ao do vento utilizando fator 0,3, admitindo-se comportamento linear da estrutura. Portanto, a equao de estado limite,
para anlise de confiabilidade em relao a deslocamentos,
dada por:
g (EM , f c , D, L, W ) = EM . u evaluated ( f c , D, L, W ) - H / 500
1 : coeficiente de combinao para ao varivel principal;

2 : coeficiente de combinao para aes variveis secundrias.
Para os edifcios analisados uma nica equao obtida, pois o

vento, quando carregamento varivel secundrio, tem multiplicador nulo. Assim, obtm-se:
(12)
Fser = Dn + 0.3Wn + 0.3Ln
onde:
Fser : valor combinado das aes para estado limite de servio;
Dn : ao permanente nominal;
Wn : ao do vento nominal;
Ln : ao varivel (acidental) nominal.
A Tabela 3 resume os valores caractersticos de resistncia (fck) e
valores nominais das aes (Dn, Ln, Wn) utilizados na verificao
dos prticos para estado limite de servio. A Tabela 4 mostra os
resultados obtidos na verificao do estado limite de servio, para
deslocamento horizontal no topo dos prdios estudados. Pode-se
(13)
onde EM, fc, D, L e W so as variveis aleatrias do problema,

descritas na Tabela 5. Os parmetros das distribuies de probabilidade das aes (D, L e W) so calculados, conforme indicado
na Tabela 5, a partir dos valores nominais utilizados no dimensionamento (indicados na Tabela 3).
Os carregamentos reduzidos para condio de servio (Eq.
12) correspondem a carregamentos freqentes, aos quais a
estrutura estar sujeita durante a vida til. Da mesma maneira, a princpio, a anlise de confiabilidade poderia ser realizada
para combinaes freqentes de carregamento, combinando o
carregamento acidental de ponto arbitrrio com o carregamento
extremo anual de vento. No entanto, esta anlise resultaria em
uma probabilidade de falha anual, que teria que ser comparada
com o ndice de confiabilidade alvo anual para estado limite de
servio irreversvel (alvo=2,9 segundo o EUROCODE [2]). Alternativamente, pode-se considerar as distribuies de extremos
(50 anos) das mesmas aes, a fim de se avaliar a confiabilidade para a vida de projeto da estrutura. Neste caso, o ndice de
Tabela 5 Variveis aleatrias, distribuies de probabilidades e parmetros

Varivel Aleatria
Distrib.
Mdia
Desviopadro
C.V.
Fonte
Erro de modelo 70/70
Normal
0,908
0,150
0,165
este trabalho
Erro de modelo 80/40
Normal
0,682
0,111
0,162
este trabalho
fc
Normal
fck + 1,65.
4,00 MPa
0,150
MELCHERS [13]
Ao permanente
Normal
1,05 Dn
0,105 Dn
0,100
Ao varivel - ponto arbitrrio
Gamma
0,25 Ln
0,148 Ln
0,55
Ao varivel - extremo de 50 anos

Ao do vento - mximo anual
Gumbel
Gumbel
1,00 Ln
0,33 Wn
0,250 Ln
0,155 Wn
0,25
0,47
Ao do vento - mximo de 50 anos
Gumbel
0,90 Wn
0,306 Wn
0,34
100
Tabela 6 Resultados para combinaes de ao acidental de 50 anos e vento mximo anual

N. Pav.
4
12
Coeficientes de sensibilidade das variveis aleatrias

fc
D
L
W
EM
Caso
baprox
Pf
70/70
4,019
2,92 E-5
0,305
0,025
0,0
0,0
-0,669
80/40
4,265
9,97 E-6
0,301
0,027
0,0
0,0
-0,672
70/70
4,292
8,84 E-6
0,344
0,026
0,0
0,0
-0,630
80/40
4,331
7,43 E-6
0,308
0,030
0,0
0,0
-0,662
70/70
4,116
1,92 E-5
0,302
0,023
0,0
0,0
-0,675
80/40
4,159
1,60 E-5
0,292
0,020
0,0
0,0
-0,688
confiabilidade alvo recomendado pelo EUROCODE [2] igual a

alvo=1,5 (estados limites irreversveis de servio). Na primeira
condio, calcula-se a probabilidade do estado limite acontecer
em qualquer ano durante a vida da estrutura. Na segunda condio, calcula-se a probabilidade do estado limite ocorrer pelo
menos uma vez durante a vida de projeto (50 anos) da estrutura.
Neste trabalho, adota-se a segunda condio, por se entender
que mais representativa da situao desejvel para a estrutura
(nenhuma fissurao da alvenaria de fechamento durante a vida
de projeto da estrutura).
As anlises de confiabilidade, considerando aes extremas, so
realizadas para duas combinaes de aes: a primeira, considerando carregamento acidental extremo para perodo de 50
anos, combinado com vento mximo anual; a segunda, considerado carregamento acidental de ponto arbitrrio (aquele que
estaria atuando em qualquer instante) combinado com o vento
extremo para perodo de 50 anos. Estas combinaes de aes
so usuais, ao se converterem problemas de confiabilidade dependentes do tempo em problemas independentes do tempo
(ELLINGWOOD et al.[11], BECK & SOUZA JR, [12]). As distribuies de extremos de 50 anos, mximo vento anual e ao
acidental de ponto arbitrrio so apresentadas na Tabela 5.
A probabilidade de falha calculada por:
Pf =
f X ( x )dx F ( - b )
(14)
g ( x )< 0
onde X o vetor de variveis aleatrias, g(x) a equao de estado limite (Eq. 13), ( ) a funo cumulativa de probabilidade normal padro e
o ndice de confiabilidade. Neste trabalho, a equao (14) resolvida
atravs do mtodo de confiabilidade de primeira ordem - FORM (MELCHERS, [13]), utilizando o software StRAnD (BECK [4]). Neste mtodo,
o problema transformado para o espao normal padro, e resolvido
como um problema de otimizao com restrio: o ndice de confiabilidade torna-se a menor distncia entre a equao de estado limite e a
origem do espao normal padro. O ndice de confiabilidade est relacionado com a probabilidade de falha atravs da equao (15):
(15)
b = - F -1 ( Pf )
Tabela 7 Resultados para combinaes de ao acidental de ponto arbitrrio e vento extremo de 50 anos
N. Pav.
4
8
12

fc
D
L
W
EM
Caso
baprox
Pf
70/70
2,127
1,60 E-2
0,226
0,029
0,0
0,0
-0,745
80/40
2,369
8,90 E-3
0,216
0,030
0,0
0,0
-0,754
70/70
2,441
7,33 E-3
0,238
0,030
0,0
0,0
-0,732
80/40
2,453
7,08 E-3
0,218
0,032
0,0
0,0
-0,750
70/70
2,235
1,27 E-2
0,232
0,022
0,0
0,0
-0,746
80/40
2,253
1,21 E-2
0,214
0,023
0,0
0,0
-0,763
101
Tabela 8 Resultados para combinaes de ao acidental de 50 anos e vento mximo anual

D
L
W
N. Pav.
brigorous
Pf
fc
4,957
3,58E-07
0,079
0,000
0,000
-0,921
5,016
2,64E-07
0,267
-0,004
-0,002
-0,727
12
5,129
1,46E-07
0,050
0,000
0,000
-0,950
4.3 Anlise de confiabilidade usando modelos

simplificados: resultados
As Tabelas 6 e 7 apresentam resultados das anlises de confiabilidade, utilizando as combinaes da ao acidental mxima
de 50 anos com vento mximo anual (Tabela 6) e ao acidental de ponto arbitrrio com vento extremo de 50 anos (Tabela 7).
Os resultados referem-se anlise de confiabilidade utilizando os
modelos simplificados, com penalizao da rigidez, linearidade de
material e no-linearidade geomtrica. Percebe-se que a combinao envolvendo vento extremo de 50 anos (Tabela 7) leva a
valores maiores da probabilidade de falha do que a combinao
envolvendo ao acidental mxima de 50 anos, mesmo levando-se em considerao a no-linearidade geomtrica do problema.
Tal resultado era de se esperar, uma vez que a ao do vento atua
diretamente na direo dos deslocamentos calculados.
O termo falha, neste contexto, utilizado entre aspas, pois representa falha em atender ao critrio de deslocamento mximo
(H/500), o que (em teoria) corresponde a um estado de fissurao das alvenarias de fechamento. Tal fissurao representa um
estado limite de servio irreversvel. Como referncia, o anexo
C do EUROCODE [2] sugere um ndice de confiabilidade alvo
alvo=1,5 para estado limite de servio irreversvel e perodo de
referncia de 50 anos. Os ndices de confiabilidade encontrados
so ligeiramente maiores (Tabela 7), e, portanto, as probabilidades de falha encontradas podem ser consideradas aceitveis.
Estes resultados mostram que a equao de verificao da norma
ABNT NBR6118:2003 [1] para condio de servio (Equao 13),
juntamente com o deslocamento mximo admitido (H/1700 para
combinao freqente) so conservadores.
As Tabelas 6 e 7 mostram ainda os coeficientes de sensibilidade
das variveis aleatrias do problema. Estes coeficientes revelam
quais variveis aleatrias tm maior contribuio nas probabilidades de falha calculadas. Como era de se esperar, a ao do vento,
horizontal, tem a maior contribuio neste modo de falha por deslocamentos horizontais. A incerteza na resistncia do concreto, que
atravs da equao (7) afeta a elasticidade do material, tem relevncia mnima. As aes verticais tm efeito desprezvel nas probabilidades de falha calculadas, mesmo levando-se em considerao
a no-linearidade geomtrica do problema. Importante destacar que
a incerteza nos modelos de penalizao da rigidez tem contribuio
significativa (de 21 a 34%) nas probabilidades de falha calculadas.
relevante observar que os ndices de confiabilidade encontrados utilizando os modelos simplificados 70/70 e 80/40 so semelhantes entre si. Este resultado , em parte, conseqncia da incorporao das
variveis erro de modelo na anlise. Na prxima seo, verifica-se se
esses ndices de confiabilidade esto de acordo com aqueles encontrados em uma anlise no-linear rigorosa (fsica e geomtrica).
4.4 Anlises de confiabilidade usando modelo

rigoroso: resultados
As Tabelas 8 e 9 apresentam resultados das anlises de confiabilidade, utilizando as combinaes ao acidental mxima de
50 anos com vento mximo anual (Tabela 8) e ao acidental
de ponto arbitrrio com vento extremo de 50 anos (Tabela 9). Os
resultados destas tabelas se referem anlise de confiabilidade
utilizando o modelo rigoroso de anlise no-linear fsica e no-linearidade geomtrica.
Assim como nas anlises simplificadas, os coeficientes de sensibilidade das variveis aleatrias apresentam o mesmo comportamento, com a ao do vento (horizontal) tendo a maior contribuio neste modo de falha por deslocamentos horizontais.
Observa-se que os ndices de confiabilidade obtidos na anlise rigorosa so razoavelmente maiores do que aqueles encontrados utilizando
os modelos simplificados. Para a combinao envolvendo ao acidental extrema (menos relevante), os ndices de confiabilidade obtidos
na anlise rigorosa foram maiores do que para as anlises simplifica-
Tabela 9 Resultados para combinaes de ao acidental de ponto arbitrrio e vento extremo de 50 anos
N. Pav.
102
brigorous
Pf
fc

D
L
W
2,747
3,00E-03
0,088
0,000
0,000
-0,912
2,293
1,09E-02
0,000
0,000
0,000
-1,000
12
2,955
1,56E-03
0,057
0,000
0,000
-0,943
das. Para a combinao envolvendo vento extremo de 50 anos, resultados diferentes foram obtidos para os trs prdios analisados. Para
os prdios de 4 e 12 andares, ndices de confiabilidade maiores foram
obtidos. Para o prdio de 8 andares, o ndice de confiabilidade obtido
na anlise no-linear rigorosa foi menor do que aqueles encontrados
nas anlises simplificadas. Tal resultado pode ser uma particularidade dos prdios estudados. No entanto, considerando-se que a anlise
no-linear rigorosa mais precisa, pode-se concluir que os modelos
simplificados de penalizao da rigidez servem para realizar o projeto estrutural, mas no so suficientemente precisos para se realizar
anlises de confiabilidade (mesmo considerando os erros de modelo).
Como os ndices de confiabilidade encontrados nas anlises rigorosas foram todos maiores do que =1,5, verifica-se que os critrios da ABNT NBR6118:2003 [1] para condio de servio em
deslocamentos (fissurao da alvenaria) so conservadores.
5. Concluses
O presente trabalho apresentou um estudo do erro dos modelos
simplificados, propostos na norma ABNT NBR6118:2003 [1], para
avaliao de deslocamentos horizontais em prticos planos de
concreto armado. Uma anlise limitada para um conjunto de 42
prticos planos, de 4, 8 e 12 andares, mostrou que o modelo 70/70
mais preciso do que o modelo 80/40.
Anlises de confiabilidade estrutural no estado limite de deslocamento horizontal do topo (fissurao de alvenaria) foram realizadas utilizando os modelos simplificados e de anlise no linear
fsica dita rigorosa. Observou-se que os modelos simplificados de
penalizao da rigidez so apropriados para verificao do projeto
estrutural, mas so imprecisos para se realizar anlises de confiabilidade (mesmo quando a varivel erro de modelo considerada).
Verificou-se que a combinao de aes envolvendo a ao acidental mxima de 50 anos e vento mximo anual pouco representativa para o estado limite de deslocamento horizontal, mesmo
quando efeitos de no-linearidade geomtrica so levados em
considerao. A combinao envolvendo ao acidental de ponto
arbitrrio e vento extremo de 50 anos mostrou-se mais relevante,
apresentando ndices de confiabilidade menores, mas ainda dentro
de limites recomendados no EUROCODE [2]. A investigao sugere que os critrios de verificao da ABNT NBR6118:2003 [1] para
condio de servio em deslocamentos horizontais (Equao 13
e deslocamento mximo H/1700 para combinao freqente) so
conservadores, e resultam em ndices de confiabilidade aceitveis
para um estado limite (irreversvel) de fissurao de alvenarias.
A anlise de coeficientes de sensibilidade das variveis aleatrias
mostrou a influncia preponderante da ao do vento sobre as demais variveis, bem como a grande influncia dos erros de modelo
nas anlises simplificadas.
6. Agradecimentos
Os autores agradecem CAPES e ao CNPq pelo apoio financeiro
a este projeto de pesquisa.
[02] EUROCODE, 2001. prEN 1990: Basis of Structural

Design Annex C: Basis for Partial Factor Design and

Reliability Analysis, European Committee for
Standardization, Brussels.

[03] CORELHANO, A.G.B. Anlise no linear geomtrica e

fsica de ncleos de edifcios altos em concreto

armado, So Carlos, 2010, Dissertao (Mestrado) -

Escola de Engenharia de So Carlos, Universidade de
So Paulo.

[04] BECK, A.T. StRAnD Structural Reliability Analysis and

Design, Manual do Usurio, Verso 1.03, 2008,

Departamento de Engenharia de Estruturas, Escola

So Paulo.

[05] WEN, R.K.; RAHIMZADEH, J. Nonlinear elastic frame

analysis by finite element. Journal of Structural

Engineering, v.109, n.8, p.1952-1971, 1983.

[06] CORRA, M.R.S. Aperfeioamento de modelos

usualmente empregados no projeto de sistemas

estruturais de edifcios, So Carlos, 1991, Tese

(Doutorado) Escola de Engenharia de So Carlos,

Universidade de So Paulo.

[07] KENT, D. C.; PARK, R. Flexural members with

confined concrete. Journal of Structural Division

ASCE, v.97, n. ST7, p1969-1990, July, 1971

[08] FIGUEIRAS, J.A. Ultimate Load analysis of

anisotropic and reinforced concrete plates and shells.

Swansea. Ph.D. Thesis, - Department of Civil

Engineering, University College of Swansea, 1983.

[09] OLIVEIRA, W.L.; BECK, A.T.; EL DEBS, A.L.H.C.

Safety evaluation of circular concrete-filled steel

columns de-signed according to Brazilian building

code NBR 8800:2008, IBRACON Structures and

Materials Journal, vol.1, p. 212-236.

[10] AMERICAN CONCRETE INSTITUTE. Commitee 435

(1984). Allowable Deflections. Detroit.

(ACI 435.3R-68, revised 1984).

[11] ELLINGWOOD, B.; GALAMBOS, T. V.;

MACGREGOR, J. G.; CORNELL,C. A. Development

of a Probability Based Load Criterion for American

National Standard A58. US National Bureau of

Standard Special Publication 577, Washington,
D.C., 1980.

[12] BECK, A.T.; SOUZA JR., A.C.S. A First Attempt

Towards Reliability-based Calibration of Brazilian

Structural Design Codes, Journal of the Brazilian

Society of Mechanical Sciences and Engineering,

Vol. 32, No.2, p 119-127, 2010.

[13] MELCHERS, R.E. Structural Reliability Analysis

and Prediction, 2nd Edition, John Wiley and Sons,
NY, 1999.
7. Referncias bibliogrficas


NBR 6118: Projeto de estruturas de concreto

Procedimento. Rio de Janeiro, 2003.
103
A variable limit for the instability parameter of

wall-frame or core-frame bracing structures
Um limite varivel para o parmetro de instabilidade
de estruturas de contraventamento formadas por
associaes de prticos com paredes ou ncleos
R. J. ELLWANGER a
rjellwanger@pop.com.br
Abstract
This work aims to investigate the viability and convenience of adopting a variable limit a1 for the instability parameter of buildings with reinforced
concrete wall-frame or core-frame structures. Initially, the evolution of tall buildings global stability theory is summarized, giving emphasis to
define when a second order analysis is needed. The treatment given to this subject by the present Brazilian code for concrete structures design
(NBR 6118:2007) is also showed. It follows a detailed analytical study that led to the derivation of an equation for the variable limit a1; a series of
examples is presented to check its accuracy. Results are analyzed, showing the validity bounds of the equation and research directions are suggested, in order to improve it.
Keywords: instability, bracing structures, second order.
Resumo
O presente trabalho tem por objetivo investigar a viabilidade e a convenincia de se adotar um limite varivel a1 para o parmetro de instabilidade
de edifcios com estruturas de concreto armado, constitudas por associaes de prticos com paredes ou ncleos. Inicialmente, feito um resumo da evoluo da teoria sobre a anlise da estabilidade global de edifcios altos, especialmente sobre a definio da necessidade ou no de
se realizar uma anlise de segunda ordem; mostra-se tambm como esta questo tratada pela atual norma de projeto de estruturas de concreto
(NBR 6118:2007). Na seqncia, apresenta-se um detalhado estudo analtico que levou ao estabelecimento de uma frmula para o limite varivel
a1, seguido de uma srie de exemplos para testar a validade da mesma. Os resultados so analisados, mostrando-se os limites de validade da
frmula e indicando-se linhas de investigao no sentido de aperfeio-la.
Palavras-chave: instabilidade, estruturas de contraventamento, segunda ordem.
Professor Associado, Departamento de Engenharia Civil, Universidade Federal do Rio Grande do Sul, e-mail: rjellwanger@pop.com.br,
endereo postal: Rua Marcelo Gama 1189/401, CEP 90540-041, Porto Alegre-RS, Brasil.
Received: 14 Sep 2011 Accepted: 18 Nov 2011 Available Online: 01 Feb 2012
2012 IBRACON
R. J. ELLWANGER
1. Introduction
1.1 Second order effects in building structures
Depending on its flexibleness, a building bracing structure, when
simultaneously subject to gravity and wind loads, may develop additional effects to those usually obtained in a first order linear analysis
(in which the equilibrium is verified in the non deformed structure).
They are the second order effects, in whose computation the material nonlinear behavior (physical nonlinearity) and the structure deflected shape (geometric nonlinearity) must be considered.
The work of Beck and Knig [1], brought in 1967, represented an important advance in the development of tall buildings global stability
analysis. A very easy criterion to apply was established, determining
that the second order effects may be neglected, provided that they
dont represent an increase more than 10% on the first order effects.
Figure 1 shows the simplified model for the bracing system. At first,
all bracing substructures are grouped in a single column, while all
braced elements (bearing elements that dont belong to the bracing
system) are replaced by an assemblage of hinged bars, as shown
in figure 1-a. The wind is considered by means of a w rate uniform
load. P and V are the floor vertical loads, applied on the bracing substructures and braced elements, respectively. The loads w, P and V
are considered with their characteristic values. Thereafter, in order
to make possible to determine the second order effects by means of
a continuum analysis, an equivalent approximate model, shown in
figure 1-b, is adopted, with a continuous and uniform distribution of
floors and vertical loads (p = P/h e v = V/h).
Concerning to the influence of the loads V, acting on the braced
elements, Beck and Knig [1] proved that, when the system distorts laterally, horizontal forces are transmitted through the floor
members to the bracing system, increasing the bending moment

on its support. It can be proved that this increase is given by the
sum of the forces V multiplied by the horizontal displacements of
the respective floors. Therefore, in order to compute this bending
moment including second order effects, the vertical loads acting
on the bracing system would be given by the sum of its proper P
loads and V loads.
In 1978, the criterion proposed by Beck and Knig [1] was included
in the Comit Euro-International du Bton recommendations (CEB
[2]). Its application consists in comparing the global bending moments at the bracing system support M I (considering only first order effects) and M II (including second order effects):
M II 1,1 M I
(1)
2
1,75wH tot
(2)
or
2
wH tot
1
1
1
1
75
,
3
2
1,75(p + v)H tot
18EJ
It can be noted that M I and M II are due to factored loads, since

the rates w, p and v are multiplied by 1,75. On the other hand,
the physical nonlinearity is regarded taking EJ = 0,7 Ecm J for the
structural members, where Ecm J represents the sum of the bracing
substructures stiffness coefficients at the non cracked stage. Thus,
performing this substitution leads to the condition:
Figure 1 Simplified model for the bracing system
105
H tot ( p + v) H tot /( Ecm J ) 0,54
(3)
According to Vasconcelos [3], the results obtained by Beck and

Knig [1] could be applied only to building structures whose lateral stiffness was concentrated in few columns rigidly connected
among themselves, in order to be considered as a single column.
The correspondence of this model with other types of bracing systems (variable section walls, rigid frames etc.) came to be done
through the equality of horizontal displacements due to horizontal
loads. The equivalent column would be that one with a stiffness
factor EJ such that the resulting horizontal displacements were
the same of the structure under consideration, for the same horizontal loading. With the purpose of simplification, this equivalent
stiffness came to be determined based on the actuation of a unit
horizontal load at the building top. In Brazil, the procedure came
to be applied changing the load factor from 1,75 to 1,40 see, for
example, Sussekind [4] and came to be known as minimal stiffness check. Consequently, inequality (3) changed to:
H tot ( p + v) H tot /( ES- 28 J I ) 0,60
(4)
where ES-28 is the concrete secant elasticity modulus at 28 days and

JI is the sum of bracing substructures inertias at non cracked stage.
In 1985, Franco [5] proposed that the equivalent column stiffness have to be obtained based on the actuation of a uniformly distributed horizontal load, in place of the top unit load. Furthermore,
he preconized that the deflected shape of the bracing structure can
affect the Beck and Knig [1] criterion application. Thus, the coefficient on the inequality (4) right hand would have its value defined
as a function of the bracing type:
- walls or cores: coefficient 0,7;
- wall-frame or core-frame structures: coefficient 0,6;
- only rigid frames: coefficient 0,5.
In 1995, Franco [6], dealing with the physical nonlinearity consideration through structural members stiffness reduction, proposed
different reduction factors, specific for slabs, column members and
beam members with symmetrical and asymmetrical reinforcement.
Although not belonging to this work purpose, a mention deserves
to be made to the method based on the moment amplification factor gz, presented in 1991 by Franco and Vasconcelos [7]. It also applies the criterion of 10% increase in relation to first order effects,
to define if a second order analysis is or not needed; however, in
this case it is done for each combination of horizontal and vertical loads. Furthermore, under certain conditions, this method may
itself constitute a second order analysis. These features caused
this method to be rapidly disseminated and largely employed in
buildings structures design.
1.2 ABNT NBR 6118 prescriptions

The ABNT NBR 6118 [8], present Brazilian code for concrete structures design, adopted the fundamental idea presented in [1] and
106
[2], on determining in its section 15 that the second order global

effects are negligible when lower than 10% of the respective first
order effects (fixed nodes structure). In order to verify the possibility of dispensing the consideration of second order global efforts,
in other words, to define if the structure may be classified as a
fixed nodes one, without the need of a rigorous analysis, ABNT [8]
presents two approximate procedures, based respectively on the
instability parameter and the gz factor. The first one just consists of
the Beck and Knig [1] criterion application and determines that: A
symmetrical framed structure may be considered as a fixed nodes
one, if its instability parameter a will be lesser than the a1 value,
according to the expressions:
(5)
a = H tot N k /( E C S I C )
a1 = 0,2 + 0,1n ' n 3
a1 = 0,6 ' n 4
(6)
n is the number of horizontal bars levels (floors) above the foundation or a slightly displaceable subsoil level. Htot is the structure total
height, measured from the foundation top or a slightly displaceable
subsoil level. Nk is the summation of all vertical loads acting on
the structure (from the level considered for Htot computation), with
their characteristic values. ECSIC represents the summation of all
column stiffness values in the considered direction. In the case
of framed, trussed or mixed structures, or columns with variable
stiffness along the height, the ECSIC value of an equivalent column
with constant section may be considered. The determination of
this equivalence will be seen in section 2.1. IC is the moment of
inertia considering columns gross sections. ECS is the secant elasticity modulus, expressed by:
ECS = 0,85 ECi = 0,85 5600 f ck1/ 2
(7)
ECS, ECi (tangent elasticity modulus) and fck (compressive characteristic strength) are given in MPa. The NBR 6118 code also adopted the Franco [5] propositions on determining different a1 values, depending on the bracing structure type: The limit value a1 =
0,6, prescribed for n > 4, is generally applicable to the building
usual structures. It may be adopted for wall-columns assemblages
and rigid frames associated to wall-columns. It may be increased
until a1 = 0,7 in the case of bracing systems composed exclusively
by wall-columns and must be reduced to a1 = 0,5 if there are only
rigid frames.
In a second order analysis, the effects of both physical and geometric nonlinearities must be considered. ABNT [8], in its item 15.7.3,
allows that the physical nonlinearity can be considered in an approximated manner, on calculating second order global efforts in
framed structures with four or more floors. This is done by means
of a reduction of the structural members (EI)sec stiffness factors as
R. J. ELLWANGER
a function of ECiIC, or of ECSIC if equation (7) is used. Representing

by As and As, respectively, the tensile and compressive longitudinal reinforcements areas, the following expressions can be written:
- slabs:
(8)
( EI ) sec = 0,3 ECi I C = 0,353 ECS I C

- beams:
( EI )sec = 0,4 ECi I C = 0,471ECS IC
' As As'
(9)
( EI )sec = 0,5 ECi I C = 0,588 ECS IC
' As = As'
(10)
2. Linear analysis
2.1 Equivalence between bracing substructures
- columns:
( EI )sec = 0,8 ECi I C = 0,941 ECS I C
nance of frames, when a value lesser than 0,6, possibly close

to 0,5, should be adopted.
These errors, if expressed in relation to a1, are apparently small.
However, is must be remembered that the instability parameter
computation requires a square root extraction. Consequently, on
verifying the exemption of performing a second order analysis, the
error on determining the needed stiffness can become significant.
This work aims to research a way of defining the instability parameter
limit a1 for associations of rigid frames and walls/cores, variable with
their stiffness factors proportion. At first, the linear behavior formulation for these associations is presented, followed by an analytical study
about the geometric nonlinear behavior of isolated walls/cores and rigid
frames and then of their association. This study is based on the simplified model presented in section 1.1, applying the criterion expressed by
inequality (1); the differential equations are solved by Galerkin method.
Right away, the formula found for the variable limit a1 is tested in a series of examples of buildings braced by wall-frame associations. 88
tests are performed, varying the number of floors, of frame spans and
the proportion between the stiffness factors of frames and walls.
(11)
Furthermore, when the bracing substructure is exclusively constituted by beams and columns (rigid frame) and the importance
factor of the second order global efforts (gz) is lesser than 1,3 (corresponding to a bland nonlinearity) it is allowed to consider the
stiffness of the rigid frame members as a whole, as follows:
The substructures of the wall or core types distinguish themselves

by a high stiffness to shear, predominating flexural deflections.
They may be modeled by simple beams, fixed on the building support, behaving as columns. Figure 2-a shows a wall or core, modeled by a cantilever bar of length l, subject to an uniformly distributed horizontal load of ratio w. Representing the material longitudinal
elasticity modulus, the constant section moment of inertia and the
bending moment function respectively by E, J and M(x), the differential equation of motion may be expressed as:
E J d 2 y / dx 2 = E J df / dx = M ( x) = w(l - x) 2 / 2
(12)
1.3 Reasons and targets of the research

The ABNT [8] code represented an improvement in relation to
the preceding one, on establishing procedures for checking if
second order global effects are unnecessary to consider. Concerning to the instability parameter for buildings with four or
more floors, it treated differently the various types of bracing
systems, on determining different values for the a1 limit. However, the prescription of a fixed limit (a1 = 0,6) for associations
of walls and/or cores with rigid frames is questionable. As the
relation between the stiffness factors of walls/cores and frames
can vary, a1 also can vary from 0,5 to 0,7. This can lead to two
types of errors:
n on behalf of safety: in associations with predominance of walls/
cores, the code restricts a1 to 0,6, when a larger value, possibly
close to 0,7, could be adopted;
n contrary to safety: in the case of associations with predomi-
(13)
Bending moments inducing tension on the bar left side are positive, the deflected concavity becoming turned to right (f(x) is its
slope). Introducing the appropriate boundary conditions, y(x) and
the top horizontal displacement DH are obtained:
y ( x) =
wl 4
24 EJ
4
4x
x
- 1
+
l
l
D H = y (l ) = wl 4 / 8EJ
(14)
(15)
In substructures of the rigid frame type, the deflections due to bending of the individual beam and column members are predominant.
When the frame is subject to horizontal loads, the global bending
moment is mainly carried to the columns as axial efforts, for which
107
Figure 2 Bracing substructure equivalent bars
the structure has a high stiffness. The horizontal deflections are

mostly caused by global shear. Therefore, the rigid frames may be
modeled as vertical bars extremely stiff to global bending, prevailing shear distortions.
Figure 2-b shows a rigid plane frame subject to an uniformly distributed horizontal load of ratio w. It is modeled by a vertical bar
predominantly deformable by shear. As shown in the figure, the
deflected shape of this bar characterizes itself by a maximum f(x)
slope at basis and tending to zero at top, just the contrary that happens to the bar simulating the wall or core. This slope is related
with the differences between horizontal displacements at adjacent
floors. In their turn, these differences are proportional to the global
shear Q(x). According to Stamato [9], the deflected shape for this
case is described by the following equation:
S dy / dx = S f ( x) = Q ( x) = w ( l - x)
(16)
The proportionality factor S represents the system (plane frame) stiffness to global shear; it corresponds to the G A / c factor of a bar
with shear deformation, where G, A and c are, respectively, the shear
modulus, the section area and the section shape factor. Solving equation (16), y(x) is obtained, leading to the top horizontal displacement:
D H = y (l ) = wl 2 / 2S
(17)
The relations established in this section have as their purpose to

obtain the inertia of a bar equivalent to a given rigid plane frame.
The item 15.5.2 of ABNT [8] code, on dealing with the instability
parameter, establishes a methodology to determine the ECSIC factor of a constant section column, equivalent to a given rigid plane
frame. According to this methodology, the above-mentioned stiff-
108
ness factor should be obtained computing initially the horizontal

displacement on the bracing structure (frame) top, under the horizontal loading, which is just DH given by (17). The next step is to
obtain the stiffness of an equivalent column with constant section
such that, under the same loading, undergoes the same top horizontal displacement which, in this case, is DH given by (15). This
implies in equality between the two expressions for DH, resulting:
S = 4E J /l 2
(18)
2.2 Association of rigid frames with walls and/or cores

This section presents the formulation of the linear response of
frame-wall/core structures, in order to that it have further on to
be used by Galerkin method in obtaining an approximate solution
for the nonlinear behavior of these structures. Figure 3-a shows
the simplified model of a bracing system composed by substructures of the frame and wall/core types. The model consists in a
wall (representing all system walls and cores) and a rigid frame
(representing all system frames) connected among themselves by
hinges (representing the floor slabs). An uniform distribution of rate
w is admitted for the wind loads. EJ1 represents the stiffness of the
frames set, according to equation (18). EJ2 represents the stiffness
of the walls/cores set.
Figures 3-b and 3-c show the loads to that the wall and the frame
will respectively be subject. These loads consist in top concentrated forces (QT for the frame and QT for the wall) and distributed
forces that can be decomposed in constant and variable (along
the height) parcels. The constant parcels (w1 for the frame and w2
for the wall) are such that w1 + w2 = w. The variable parcels (rate
(u(x) for the frame and u(x) for the wall), jointly with the forces QT,
represent internal forces originated from the wall-frame interaction;
since they are connected by the hinges, the wall and the frame
R. J. ELLWANGER
Figure 3 Association of rigid frames with walls or cores
are impeded to develop their natural deflected shapes, as shown

in figure 3-d. The frame will be subject to a global shear forces
distribution given by:
Q( x) = w1 ( l - x) + QT +
u (x )dx
(19)
As was seen in section 2.1, the frame behavior is described by

equation (16). Writing this equation, introducing (18) and (19) and
isolating the terms regarding to the internal forces, gives:
4 EJ1
QT + u (x )dx =
f ( x) - w1 ( l - x)
x
2
(20)
In its turn, the wall will be subject to a bending moments distribution given by:
M ( x) = w2 (l - x) 2 / 2 - QT (l - x) -
u(x )(x - x)dx (21)

x
As was also seen in section 2.1, the wall behavior is described by

equation (13). Writing this equation, introducing M(x) given by (21)
and deriving both members, gives successively:
EJ 2
df
= w2 (l - x) 2 / 2 - QT (l - x) dx
l
x
u (x )(x - x)dx
(22)
EJ 2
d 2f
= - w2 (l - x) + QT +
dx 2
l
x
u (x )dx
(23)
Substituting (20) into (23) and re-arranging:
EJ 2
d 2f 4 EJ1
- 2 f ( x) + w1 (l - x) + w2 (l - x) = 0
dx 2
l
(24)
Considering that w1 + w2 = w (total wind load acting on the system)

and defining a new variable K = J 1/J 2 , the solution for equation (24) may be expressed as follows:
f ( x) = C1 e 2 Kx / l + C2 e - 2 Kx / l +
wl 2
(l - x)
4 EJ1
(25)
where
C1 =
wl 3 e 2 K - 2 K
8EJ 1K e 4 K + 1
(26)
C2 =
- wl 3 2 K e 4 K + e 2 K
8EJ1K
e4K + 1
(27)
109
3. The Galerkin method

In many engineering problems, as the ones that are presented
in the next sections, there arises the need to solve an equation of
the type L(y) = 0, where L is a differential operator, whose solution satisfies to homogeneous boundary conditions. The Galerkin
method consists in obtaining an approximate solution of the form:
y ( x) =
a j ( x)
i =1
(28)
i i
where i ( x ) (i = 1, 2,..., n) are functions, previously chosen

and satisfying to the same boundary conditions; the ai are coefficients to be determined. The n functions i ( x ) must be linearly
independent and belong to a system, represented by { i ( x ) } (i
= 1, 2,..., n) and endowed of the completeness property in the
solution domain. In order to y (x ) be the exact solution of the
given equation, it is necessary that L( y ) be identically null. This
requirement, if L( y ) is considered to be continuous, is equivalent to the requirement of the orthogonality of the expression L(
y ) to all the functions i ( x) (i = 1, 2,..., n). However, having at
disposal only n constants ai, only n orthogonality conditions can
be satisfied. Applying these conditions, the following system of
equations is obtained:
L( y( x)) ji ( x)dx =
n
L a jj j ( x) ji ( x)dx = 0
j =1
(i = 1, 2, . . . , n)
(29)
The solution of this system (a linear one, in the case of a linear

operator L) provides the values of the coefficients ai, from which
the approximate solution y (x ) is obtained. The proof of convergence, as well as more detailed considerations about the Galerkin
method can be seen in Kantorovich and Krylov [10].
4. Exemption of the second order effects

consideration
The sections 4.1 and 4.2 present the formulation of the geometric
nonlinear behavior, respectively of wall/cores and rigid frames assemblages. For both cases, the limits a1 of the instability parameter are deduced, comparing them with the values prescribed by ABNT [8] code.
The section 4.3 does the same for the associations of these types of
substructures, obtaining an expression for the variable limit a1, main
objective of this work.
Figure 4-a shows the deflected shape of a bar equivalent to a bracing
system, subject to uniformly distributed loads of rates w and q, respectively in the horizontal and vertical directions; q is given by the sum of
the rates p and v of figure 1-b. Taking into account the bar deflections
(geometric nonlinearity) and representing by Y the primitive function of
the displacements y(x), the bending moment will be given by:
M ( x) = w(l - x) 2 / 2 +
l
x
q[ y(x ) - y ( x)]dx
(30)
or
M ( x) = w(l - x) 2 / 2 + q[Y (l ) - Y ( x) - (l - x) y ( x)]
(31)
Considering that y(0) = 0, the bending moment at support will be

expressed by:
M (0) = wl 2 2 + q[Y (l ) - Y (0)]
(32)
Figure 4 Deformations influence in the structure response
A Bracing system equivalent bar
110
Shear deformation at the infinitesimal level
R. J. ELLWANGER
4.1 Substructures of shear wall or shear core types

In the case of bracing systems formed exclusively by shear walls
and/or shear cores, the differential equation of motion will be obtained equalizing EJ d 2 y / dx 2 to M(x) given by (31):
E J d 2 y / dx 2 = w(l - x) 2 / 2 + q[Y (l ) - Y ( x) - (l - x) y ( x)]
(33)
Deriving it in relation to x and considering that the rotations are

given by f(x) = dy/dx, changes equation (33) into:
E J d 2f dx 2 + q(l - x)f ( x) + w(l - x) = 0
(34)
An approximate solution for equation (34) can be obtained

through the Galerkin method. Assuming that this solution is
proportional to f(x) due exclusively to first order effects, it may
be written:

x
f ( x) = a1j1 ( x) = a1 1 - 1 -
l

3
(35)
where 1 ( x ) was obtained deriving equation (14) in relation to x

and suppressing the constant that would remain in evidence. Applying equation (29) with n = 1, leads successively to:
l
0
L(f ) j1 ( x) dx =
L(a j ( x)) j ( x)dx = 0 (36)

0
1 1
displacement at support, leads to the displacements function. Integrating again, leads to:
Y ( x) =
l x 5 2 x 2
wl 4
- x + C
- 1 - +
l
24 EJ - 3ql 3 5 l
where C is the integration constant. The bending moment at support can be obtained on applying equations (32) and (40):
M ( 0) =
wl 2
2qwl 5
+
2
5(8 EJ - ql 3 )
x 3
x 3
6EJa1 x
2 - 1 + q(l - x )a1 1 - 1 - + w(l - x) 1 - 1 - dx = 0
l
l
0 l l
(37)
Performing the integration and isolating a1, results:
a1 =
4 wl 3
24 EJ - 3ql 3
(38)
On substituting (38) into (35), the approximate solution is obtained:
f ( x) =
4 wl 3
24 EJ - 3ql 3

1 - 1
(39)
(41)
Equation (41) can be transformed successively into:
M (0) =
wl 2 8EJ - ql 3 / 5
4ql 3
wl 2
=
1 +
2 5(8EJ - ql 3 )
2 8 EJ - ql 3
(42)
The condition of that, in the ultimate limit state (loads multiplied by

1,4), the second order effects may not exceed the first order effects
in more than 10% (inequality (1)), is applied to the support bending
moment, obtaining:
1,4 wl 2 8EJ - 1,4ql 3 /5

1,4 wl 2
1,1
2
2
8 EJ - 1,4ql 3
(43)
The terms wl2, multiplying both sides of the inequality, do vanish.

Performing the required algebraic transformations, results:
ql 3 / EJ 0,6349
l
(40)
(44)
Since a wall or core has a behavior equivalent to the one of a

column, the physical nonlinearity may be considered adopting for
EJ the expression 0,941 ECSIC, according to equation (11). On the
other hand, remembering that ql is the total vertical load Nk and l is
the total height Htot, inequality (44) becomes:
2
H tot
N k / ECS I C 0,5974
(45)
Extracting the square root of both members:
H tot N k / E CS I C 0,773
(46)
Integrating (39) in relation to x and applying the condition of zero
111
Thus, inequality (46) denotes a value of 0,773 for a1. In its turn, the
ABNT [8] code allows the coefficient a1 to be increased until 0,7 if the
bracing system is composed exclusively by shear walls or shear cores.
4.2 Substructures of the rigid plane frame type

In the case of bracing systems formed exclusively by rigid frames,
the equivalent bar of figure 4-a will deform predominantly by shear.
Also in this case, the efforts expression must take the deflected
shape into account. It can be proved that the infinitesimals ds and
dx shown in figure 4-b are related by:
ds = dx2 + dy2 = dx2 (1+ dy2 dx2 ) = dx 1+f 2
(47)
The shear effort can be obtained from the derivation of equation (31) in
relation to the bar deflected axis. Introducing ds given by (47), results:
dM
- dM dx w(l - x) + q(l - x)f ( x)
Q( x) = =
=
ds
1 + f 2 ( x)
1 + f 2 ( x)
(48)
It is an inclined shear effort, as shown in figure 4-b. The shear deformation caused by it has the same slope, given at the infinitesimal level by:
dy
dy
dy
=
=
= dy 1 + f 2
cos f dx ds 1 1 + f 2
(49)
On establishing the differential equation of motion for this case,

two changes must be performed with relation to equation (16): to
introduce dy/cosf given by (49), in place of dy, and Q(x) given by
(48). On doing so, results:
dy
Q(x)
w( l - x) + q( l - x)f(x)
1 + f 2 ( x) =
=
dx
S
S 1 + f 2(x)
(50)
Thus:
dy
w(l - x) + q(l - x)f ( x)
f ( x) =
=
dx
S[1 + f 2 ( x)]
(51)
In cases of bland geometric nonlinearity, as the ones treated by this work,

the rotationsf(x) present values much lesser than unit; therefore, f 2(x) may
be neglected in face of 1 and equation (51) may be put in the form:
Sf ( x) = w(l - x) + q(l - x) f ( x)
112
Isolating f(x):
f ( x) =
w(l - x)
S - q (l - x )
(53)
Integrating equation (53) in relation to x and applying the condition

of zero displacement at support, leads to the displacements function. Integrating again, gives:
Y(x) =
Sw
[S - q(l - x)] {ln[S - q(l - x)]-1}- Sw2 x ln(S - q l) - w (l 2 + x2 ) + C
3
2q
q
q
(54)
where C is the integration constant. Applying equation (54) for x =

0 and x = l, leads to the difference below:
Y(l ) - Y( 0 ) =
S 2w
S
wl l S
+
ln
3
S - ql
q 2 q
q
(55)
Thus, the bending moment at support can be expressed, substituting equation (55) into (32):
M( 0 ) =
l S S 2w
wl 2 S 2 w S
Swl
1
+ 2 ln
- wl + = 2 ln
2 q S - ql
2 q q 1 - ql / S q
(56)
Applying the condition expressed by inequality (1) to this bending

moment, results:
wl 2
1,4 S 2 w
1
1,4 Swl
,
ln
1
1
1
4
1,4 q
2
1,4 2 q 2 1 - 1,4ql S
(57)
Performing the required algebraic transformations, inequality (57)

changes into:
1
1
1
ln
0,77
2
l
l
1
1
,
4
q
/S
q
/S
1,4(ql /S)
(58)
Taking the factor ql /Sas an unknown, inequality (58) can be solved

by means of trials, obtaining:
ql / S 0,0962
(59)
(52)
R. J. ELLWANGER
Substituting S by expression (18), results:
l 2 ql /EJ 0,3848
Considering this expression of (EI)sec and following the same deductive sequence that led to the inequalities (45) and (46), results:
(60)
According to the ABNT [8] code, the physical nonlinearity might be

considered, substituting EJ by (EI)sec given by (12). However, the
(EI)sec / ECSIC ratio of the frame bars assemblage, as a function of
the individual bar (EI)sec / ECSIC ratios cannot be considered a constant value; it can vary due to many factors, such as the numbers
of floors and spans, story heights, span lengths, relation between
the cross section dimensions of beams and columns etc. Pinto and
Ramalho [11] show that the influence of physical nonlinearity in the
frame lateral stiffness depends mainly on the reinforcement ratios
and the loading magnitude; they obtained (EI)sec / ECSIC ratios varying from 0,51 until 0,75 for the ultimate limit state.
On the other hand, Schueler [12] states that the contribution of beams
flexibility for the lateral deflections of a rigid frame can reach 65%,
remaining 35% due to columns flexibility. Furthermore, in a slender
frame the beam reinforcements As and As tend to be the same, due
to the predominance of wind effects. Thus, in this case equations (10)
and (11) may be employed to relate the components of yNL (frame
horizontal displacements including physical nonlinearity), due to the
beams (yNLBEAMS) and columns (yNLCOLUMNS), with the corresponding
components (yLBEAMS) and (yLCOLUMNS) of the horizontal displacements
resulting from linear analysis (yL). Simultaneously, the above-mentioned share factors (35% and 65%) of these components in the total
displacements may be used, leading to the following expressions:
COLUMNS
y NL
=
BEAMS
y NL
yLCOLUMNS
0,35 y L
=
0,941
0,941
y BEAMS
0,65 y L
= L
=
0,588
0,588
(61)
(62)
Performing the sum of the components expressed by (61) and

(62), leads to the following relation between the total horizontal
displacements yNL and yL:
y NL =
yL
0,35 y L
0,65 y L
+
=
0,941
0,588
0,677
(63)
As the frame lateral stiffness is inversely proportional to these displacements, it may be written:
( EI ) sec = 0,677 ECS I C
(64)
(65)
It can be noticed that this inequality is coherent with ABNT [8] code,
which appoints the value of 0,5 for the coefficient a1, if the bracing
system is constituted exclusively by rigid frames. In fact, in order to
obtain a1 = 0,5, (EI)sec must satisfy the following:
( EI ) sec = 0,650 ECS I C = 0,552 ECi I C
(66)
4.3 Associations of rigid frames with shear walls

and/or shear cores
The same model of figure 3 is adopted and the same definitions
of section 2.2 are considered. In order to deduce the differential
equation of motion for the frames assemblage (figure 3-c), equation (50) is applied, adding the terms due to the wall-frame interaction to the shear effort, as was done in equation (19):
Q(x)
=
S
w1 ( l- x) + QT + u (x ) dx + q1 (l - x)f ( x)
x
S 1 + f ( x)
= f ( x) 1 + f 2 ( x )
(67)
Considering that f 2(x) may be neglected in face of 1 and isolating

the terms due to the interaction forces:
QT +
u (x )dx
= Sf ( x) - w1 (l - x) - q1 (l - x)f ( x)
(68)
In order to deduce the differential equation of motion for the walls

assemblage (figure 3-b), the bending moment given by (31) is introduced, added to the terms due to the interaction forces, as was
done in equation (21):
l
w2 (l - x)2
EJ d 2 y
- QT (l - x) - u(x )(x - x)dx + q2 [Y (l ) - Y ( x) - (l - x) y( x)]= 2 2
x
2
dx
(69)
Deriving equation (69) in relation to x, gives:
EJ 2
l
d3y
d 2f
=
EJ
2 2 = - w2 (l - x) + QT + u (x ) dx - q2 (l - x) f ( x)
3
x
dx
dx
(70)
113
Substituting (68) into (70) and re-arranging, results:
EJ 2
d 2f
= - (w1 + w2 ) (l - x) - (q1 + q2 ) (l - x) f ( x) + Sf ( x)
dx 2
(71)
f (x) = a1j1 ( x)
Considering that w1 + w2 = w (total wind load), q1 + q2 = q (total

gravity load) and re-arranging again, leads to the differential equation describing the behavior of a system composed by rigid frames
and walls/cores, including the deflections influence:
d 2f
EJ 2 2 - [ S - q (l - x)]f ( x) + w(l - x) = 0
dx
(72)
(73)
wl 2
EJ 2 f " ( x) C1 e2 Kx / l + C2 e-2 Kx / l + (l - x) +
4EJ1
2K
wl 2 4 K 2
2EJ 2 f ' ( x) C1 e 2Kx / l - C2 e -2 Kx / l - + 2 EJ 2 f ( x) C1 e2 Kx / l + C2 e-2Kx / l 4EJ1 l
l
) (74)
wl 2
f ( x) [S - q(l - x)]C1 e2 Kx / l + C2 e-2Kx / l + (l - x) + w(l - x) = 0
4EJ1
Assuming that f(x) is a constant function, leads the first and second terms of equation (74) to be null, since they are multiplied by
the derivatives f ' (x ) and f " (x ) . Furthermore, considering the
preceding definitions of K (section 2.2) and S (equation 18), it may
be written:
4K 2
4 EJ1
EJ 2 =
= S
2
l
l2
(75)
2 Kx/l
+ C2 e -2 Kx/l + wl 2 (l - x) 4EJ1 + w (1 - a1 ) dx = 0
wl 2
q f ( x) C1 e 2 Kx / l + C2 e -2 Kx / l +
(l - x) + w [1 - f ( x)] = 0
4 EJ1
(78)
Performing the integration, a1 can be isolated, giving:
a1 =
1-
ql
q
C1 (e 2 K - 1) - C2 (e - 2 K - 1)
8EJ 1 2 Kw
] (79)
Therefore, the approximate solution for equation (72) will be given by:
(80)
with a1 given by (79). Integrating twice leads to the primitive of the

displacements function:
l2
wl 2 x 2 l x
Y ( x) = a1 2 C1 e 2 Kx / l + C2 e - 2Kx / l +
- + C3 x + C 4
4EJ1 2 6
4K
(81)
where C4 is an undetermined constant and C3 results from the condition of zero displacement at support:
C3 =
- wl 4 K(e 4 K - 1 ) + e 2 K
8 EJ1
K 2(e 4 K + 1 )
(82)
The bending moment at support is obtained, applying equation (32):
M(0) =
Consequently, the third term of equation (74) cancels with some

parts of the fourth one, reducing the equation to:
114
{a q [C e
wl 2
f ( x) = a1 C1 e 2 Kx / l + C2 e - 2 Kx / l +
(l - x)
4 EJ 1
Substituting (73) into (72), results:
(77)
' j1 ( x) = 1
For this case, equation (29) is applied in the following form:

l
In order to apply the Galerkin method to equation (72), it will be

assumed a solution given by a function f(x) multiplied by the linear
solution, expressed by equations (25), (26) and (27):
wl 2
f ( x) = f ( x) C1 e 2 Kx / l + C2 e -2 Kx / l +
(l - x )
4 EJ1
The Galerkin method will be used in order to find a constant function f (x) that has to be a good approximation for f(x) appearing in
equation (76). According to (28), it may be written:
l2
wl 2
wl 5
+ a1 q 2 C1 (e 2 K - 1) + C2 (e - 2 K - 1) +
+ C3 l
2
12 EJ1
4K
(83)
(76)
R. J. ELLWANGER
It can be observed that the difference Y(l) Y(0) made the constant C4 to vanish. Substituting C1, C2 and C3, respectively, by
equations (26), (27) and (82), transforms equation (83) into:
M(0) =
wl 2 a1qwl 5 (6K + 8K 3 )(e4 K + 1) + (3 - 12K 2 )(e4 K - 1) - 24Ke 2K

+
2 96 EJ1K 3
e 4K + 1
(84)
Introducing the expressions for C1 and C2 also in equation (79) and

substituting the formula of a1 obtained in such a manner into equation (84), leads to the expression of the bending moment at support
of the system composed by frames and walls/cores:
wl 2
(2K + 8K 3 / 3)(e 4 K + 1) + (1 - 4K 2 )(e4 K - 1) - 8Ke 2K
M(0) = 1 +
3 3 4K
3 4K
2 4K
2K
2 (16EJ1 /ql )K (e + 1) - ( K + 2K )(e + 1) + 2K (e - 1) + 2Ke
(86)
(87)
(88)
(89)
IC1 and IC2 are defined as the gross inertias, respectively, of the
frames and walls assemblages. Calling IC the sum of IC1 and IC2 and
applying the relations (11) and (66), gives:
I C = I C1 + I C 2 =
J1 J 2
J
J
1,5385 K 2 + 1,0625
+
= 1 + 1 2=
J1
0,65 0,941 0,65 0,941 K
K2
a1
IC1/IC
a1
0
0,10
0,20
0,30
0,40
0,773
0,772
0,771
0,768
0,763
0,50
0,60
0,70
0,80
0,85
0,755
0,744
0,726
0,699
0,679
0,90
0,95
0,98
0,99
1,00
0,651
0,611
0,574
0,555
0,509
(90)
K2
ECS I C
1,5385 K 2 + 1,0625
(91)
Equation (87) can be rewritten, substituting EJ1 by expression (91),

ql by Nk (total vertical load) and l by Htot (building height). After,
extracting the square root of both members, results:
H tot
Isolating J1:
J1 = K 2 J ( K 2 + 1)
IC1/IC
EJ1 =
Calling J the sum of inertias of the walls/cores and frames assemblages, and considering the preceding definition of K, it may be written:
J = J1 + J 2 = J1 + J1 K 2 = J1 ( K 2 + 1) K 2
a1
(85)
In inequality (86), the factor ql3/EJ1 can be isolated, giving:
ql 3
(24 / 7) K 3 (e 4 K + 1)
EJ1 (6,3K + 8,6K 3 ) (e 4 K + 1) + (3 - 12,6K 2 ) (e 4 K - 1) - 24,6 Ke 2 K
IC1/IC
From (90), the factor EJ1 may be expressed by:
Applying the condition expressed by inequality (1) to this bending

moment, results:
1,4 wl 2
(2K + 8K 3 / 3)(e 4K + 1) + (1 - 4K 2 )(e 4K - 1) - 8Ke 2K
1,4 wl 2
1
,
1
1 +
2 (16EJ1K 3 / 1,4ql 3 - 2K 3 - K )(e 4K + 1) + 2K 2 (e 4K -1) + 2 Ke 2K

2
Table 1 Values of a1, varying the IC1/ICratio
Nk
1
ECS I C
(92)
where
1 =
(24/7) K 5 (e4 K + 1)
(1,5385K 2 + 1,0625)[(6,3K + 8,6K 3 )(e4 K + 1) + (3 - 12,6K 2 )(e4K - 1) - 24,6Ke2 K ]
(93)
In this manner, an expression for the limit a1 of the instability parameter, variable with K (relation between the reduced inertias of
frames and walls/cores), was obtained. However, in order to obtain
a1, it is more practical to deal with the gross inertias. Combining
equations (66) and (90), the following relation between K and IC1/
IC (proportion between the frames gross inertia and the total one)
is obtained:
K = 0,831 ( I C1 / I C ) /(1 - I C 1 / I C )
(94)
Thus, given any proportion I C1/IC, K is obtained applying equation (94); soon after, a 1 is obtained applying equation (93). The
sequence of a 1 values, presented in table 1 and graphically
represented in figures 7 and 8, shows a rough variation for
I C1/IC close to 1 (predominance of frames) and a more smooth
one for IC1/IC close to 0 (predominance of walls). It can also be
observed that equation (93), at its domain ends, reproduces
equation (46) exactly, but presents a discrepancy of 1,8% for
bracing systems composed exclusively by rigid frames. The
115
Figure 5 Transversal bracing system: examples 1, 3, 5 and 7
next sections present a series of examples assaying the feasibility of equation (93).
5. Examples
5.1 Description of the tests
The plan of figure 5 shows the basic configuration of the transversal bracing system of a rectangular on plane building (examples 1,
3, 5 and 7); it is constituted by walls 1 and 5 on the lateral faces
and the rigid frames 2, 3 and 4 spanning over a single bay (7,5 m
between column axes). In the same way, figure 6 shows the basic
configuration of the transversal bracing system of a building with
an oblong octagonal shape on plane (examples 2, 4, 6 and 8); it is
also composed by two walls and three frames which, in this case,
span over three equal bays (5m between column axes). Each of
these systems was employed in buildings having 5, 10, 20 and 30
floors with a 3 m height, constituting examples 1 to 8, whose general information is mentioned in table 2.
Tests were performed for each of the eight buildings, varying the
rectangular cross sections of walls and frame members, in such
a way to result the sequence of IC1/IC ratios mentioned in table 3.
In some cases, this required changes in the basic configurations
of figures 5 and 6, but keeping the double symmetry of the bracing system on plane. For IC1/IC = 1, the walls were suppressed; for
decreasing values of IC1/IC, the frames were gradually suppressed,
becoming excluded for IC1/IC = 0. In the examples with 30 floors,
some additional frames were included in some cases and some
additional walls in others. The cross sections dimensions adopted
in the tests are listed in table 4.
Figure 6 Transversal bracing system: examples 2, 4, 6 and 8
116
R. J. ELLWANGER
Figure 7 Graphs a1 x IC1/IC ratio for

the examples with 5 and 10 floors
Figure 8 Graphs a1 x IC1/IC ratio for

the examples with 20 and 30 floors
Table 2 General information about examples 1 to 8

Example
Floors number
Building height (m)
Spans number perframe
Frames number
Walls number
5
15
1
0 to 3
0 to 2
5
15
3
0 to 3
0 to 2
10
30
1
0 to 3
0 to 2
10
30
3
0 to 3
0 to 2
20
60
1
0 to 3
0 to 2
20
60
3
0 to 3
0 to 2
30
90
1
0 to 5
0 to 2
30
90
3
0 to 5
0 to 4
Table 3 IC1/IC ratios adopted in the examples

1,00
0,95
0,90
0,85
0,80
0,70
0,60
0,50
0,40
0,20
Table 4 Cross sections dimensions (cm)

Example
Beams
Columns
Walls
1
2
3
4
5
6
7
8
20 x 50 to 25 x 60
15 x 34 to 20 x 50
18 x 68 to 24 x 68
15 x 51 to 24 x 59,5
34 x 85
21 x 68
34 x 85
21 x 85
20 x 50 to 25x 60
15 x 34 to 20x 50
30 x 85 to 40 x 85
20 x 77 to 34x 90
42,5 x 140
34 x 96
42,5 x 140
36 x 119
20 x 88,5 to 20 x 201
20 x 102 to 20 x 243
20 x 158,5 to 20 x 422,5
20 x 224 to 20 x 481,5
20 x 307,5 to 25 x 838,5
20 x 351 to 25 x 957
20 x 418 to 35 x 1200
20 x 588 to 31 x 998,5
A concrete with fck = 25 MPa was adopted, resulting in an elasticity modulus ECS = 23800 MPa. A total vertical load of 10 kN/m2
per floor (characteristic value) was considered. A wind pressure
of 1,5 kN/m2 (characteristic value), constant along the height, was

adopted, since it was the first experience with a formulation based
on a model with constant wind load.
117
Table 5 Values of a1, varying the IC1/IC ratio and the numbers of floors and spans
IC1/IC
1,00
0,95
0,90
0,85
0,80
0,70
0,60
0,50
0,40
0,20
0
Example:
Floors:
1
5
2
5
3
10
4
10
5
20
6
20
7
30
8
30
0,515
0,552
0,572
0,590
0,610
0,626
0,641
0,652
0,662
0,675
0,683
0,514
0,557
0,584
0,603
0,619
0,641
0,656
0,665
0,671
0,681
0,683
0,528
0,567
0,594
0,613
0,629
0,653
0,671
0,685
0,695
0,714
0,726
0,519
0,563
0,592
0,614
0,632
0,657
0,676
0,689
0,699
0,715
0,726
0,569
0,605
0,630
0,650
0,663
0,687
0,702
0,716
0,724
0,738
0,749
0,534
0,590
0,621
0,644
0,662
0,687
0,702
0,716
0,726
0,736
0,749
0,608
0,639
0,656
0,675
0,690
0,710
0,724
0,734
0,743
0,753
0,764
0,591
0,635
0,656
0,676
0,690
0,711
0,727
0,735
0,740
0,752
0,764
Thus, 88 different bracing systems were tested. Each test aimed

to determine the relation between vertical loads and horizontal stiffness that would result in a 10% increase on the global
bending moment at building support, concerning to first order
analysis; in this way, the limit a1 for the instability parameter
was determined. The procedure applied in each test consisted
in, at first, to assign cross section dimensions for the members
of the frames assemblage and compute its horizontal stiffness
IC1, according to item 15.5.2 of ABNT [8] code (relation between
the wind load and the horizontal displacement at structure top).
After, the cross section of the walls assemblage was adjusted in
order to obtain the desired IC1/IC ratio.
The test proceeded with an initial second order analysis of the
frame-wall system, employing the P-D method. After, this second
order analysis was successively repeated, adjusting the values
of the vertical loads, until achieving the desired 10% increase on
the support global moment. Although an adjustment of the horizontal stiffness would be more logical, the loads adjustment was
preferred because it made the 88 tests more agile to perform and
didnt affect the results. The physical nonlinearity was considered
by means of the individual bar stiffness reductions expressed
by equations (10) and (11). The analysis was performed using
the same plane frame model of figure 3, with the sets of frames
and walls joined by hinges, since the formulation proposed in this
work is based just on that model. Later on, some cases were reanalyzed by means of a method employing geometric stiffness
matrices, in order to confirm the results obtained by P-D method.
5.2 Results discussion

The values of a1 obtained in the tests are listed in table 5. In order to better interpret the results, it is appropriate to arrange the
eight examples according to the floors number and to consider two
ranges of the relative stiffness values: IC1/IC < 0,9 and IC1/IC > 0,9.
Figures 7 and 8 show, for each floors number, a graph representing the variation of the parameter a1 found in the tests, as well as
the graph of a1 corresponding to formula (93).
On examining the results regarding to IC1/IC < 0,9, it is verified that
almost all the values for a1 obtained in the examples are below the
values predicted by formula (93) and listed in table 1. Thus, the
application of this formula results in upward errors, whose maximum values are mentioned in table 6. It can be clearly observed
that these errors decrease as the number of floors increases; they
are of 16,2% at 5 floors and drop to 3,1% at 30 floors, indicating a
trend to become null for a little more than 30 floors. This behavior
of equation (93) is much probably due to the adoption of figure 1-b
model, in place of figure 1-a one; in other words, the model with
an uniform and continuous distribution of floors and vertical loads
provides a reasonable accuracy only for buildings with more than
30 floors.
On the other hand, for values of IC1/IC larger than 0,9, indicative of
a high predominance of frames, the trend of decreasing errors with
the increase of floors also exists. However, in this case, it comes
along with another trend, consisting of downward errors generated by equation (93), increasing with the floors number, as can be
Table 6 Maximum errors (%)

IC1/IC
5 floors
10 floors
20 floors
30 floors
0,90
= 0,95
= 1,00
+16,2
+10,7
1,2
+11,2
+8,5
3,6
+6,0
+3,6
10,5
+3,1
4,4
16,3
118
R. J. ELLWANGER
observed in table 6 for IC1/IC = 0,95 and 1,00. This trend is due to
equation (93) taking into account that the frames distort only due to
global shear. Smith and Coull [13] state that, in slender buildings, the
frames global bending, due to the columns axial deformation, can
contribute significantly to the horizontal displacements. It is about
the same distortion pattern of walls, making the limit coefficient a1 to
increase. Thus, as much higher the building and the greater presence of frames, greater will be the errors generated by a formulation
that neglects this effect. In the case of bracing systems composed
exclusively by frames (IC1/IC = 1), table 6 shows maximum errors
varying between 1,2% (5 floors) and 16,3% (30 floors).
Still regarding to the case IC1/IC = 1, the values of a1 mentioned
in the first line of table 5 hint that the limit a1 = 0,5, prescribed by
ABNT [8] for bracing systems composed exclusively by frames, is
conservative, especially in buildings with more than 10 floors. On
the other hand, the values of a1 found for IC1/IC = 0, mentioned in
the last line of table 5, indicate that the limit a1 = 0,7, prescribed for
bracing systems composed exclusively by walls/cores, is also conservative for buildings with more than 10 floors. However, in buildings with less than 10 floors, the contrary can occur; values slightly
lower than 0,7 were found in buildings with 5 floors. Furthermore,
the adoption of the fixed value a1 = 0,6 for wall-frame and coreframe bracing systems should be conditioned to a minimum limit of
the walls contribution for the bracing stiffness, especially in lower
buildings. Interpolations performed in the values of table 5 indicate
that the proportion of the walls gross inertia in relation to the total
one should be at least 18% in example 1, 14% in example 2 and
12% in examples 3 and 4.
6. Conclusions
The limit values a1 for the instability parameter, obtained in the examples of this work and mentioned in table 5, vary from a minimum
of 0,514 in example 2 until a maximum of 0,764 in examples 7
and 8. The proportion between these extreme values is close to
1,5:1. Since their computation includes a square root extraction,
the proportion between the radicands (vertical load/horizontal stiffness relations) associated to these extreme values is more than
2:1. The extent of this variability shows the importance of having
a way of predicting a limit a1 appropriated to the relation IC1/IC and
the floors number of a given building to be designed, in place of the
fixed values prescribed by ABNT [8].
Equation (93) represents an initial attempt to accomplish such prediction. The relatively good accuracy attained in examples 7 and
8 for IC1/IC < 0,9 denotes that this aim is possible to be achieved
and efforts deserve to be done in order to carry it out. In order to
remedy the errors occurred in the other cases (predominance of
frames and low number of floors), the effect of the columns axial
flexibility has to be introduced into equations (16) and (50), making the curve of a1, corresponding to equation (93) and depicted
in figures 7 and 8, decline not so much in its final segment; it has
also to be searched a way to adjust the formulation to the variation
of the floors number. Another subject to investigate is the viability
of including the variability of physical nonlinearity influence in the
frames horizontal stiffness (actually, this influence is considered
through a constant factor).
It must be emphasized that all of this has to be done in such a manner to keep the formulation simplicity, just one of the greater advantages of the instability parameter utilization. Finally, it must be ac-
centuated the need of adopting a more realistic analysis model for

the tests: modeling of the structure as a three-dimensional frame,
considering the floors as rigid diaphragms; variation of the wind
load along the building height; effectuation of the nonlinear analysis through an incremental-iterative method; and a more accurate
consideration of physical nonlinearity, for example, by means of
moment-curvature relations.
7. Acknowledgment
I would like to thank Prof. Eng. Mrio Franco, for providing precious material for this work achievement.
8. References

[01] Beck, H. and Knig, G.; Haltekrfte im Skeletbau.
In: Beton- und Stahlbetonbau, n. 62, tome 1

(pp. 7-15) and tome 2 (pp. 37-42), Berlin, 1967.

[02] CEB Comit Euro-International du Bton; CEB/FIP

Manual of Buckling and Instability, The Construction

Press, Lancaster, 1978.

[03] Vasconcelos, A.C.; Origem dos Parmetros de

Estabilidade a e gz. In: Revista IBRACON de

Estruturas, n. 20, pp. 56-60, So Paulo, 1998.

[04] Sussekind, J.C.; Curso de Concreto, vol. 2, Porto

Alegre, Ed. Globo, 1984.

[05] Franco, M.; Problemas de Estabilidade nos Edifcios

de Concreto Armado. In Colquio IBRACON sobre

Estabilidade Global das Estruturas de Concreto

Armado, So Paulo, 1985.

[06] Franco, M.; Global and Local Instability of Concrete

Tall Buildings, In: International Symposium for Shell

and Spatial Structures, Proceedings, vol. 2,

pp. 1327-36, Milan, 1995.

[07] Franco, M. and Vasconcelos, A.C.; Practical

Assessment of Second Order Effects in Tall Buildings.

In: Colquio do CEB-FIP Model Code 1990,

pp. 307-24, Rio de Janeiro, 1991.

[08] ABNT Associao Brasileira de Normas Tcnicas;

NBR 6118 Projeto de Estruturas de Concreto

Procedimento, Rio de Janeiro, 2007.

[09] Stamato, M.C.; Associao Contnua de Painis de

Contraventamento (Publicao 157), So Carlos,

EESC/USP, 1972.

[10] Kantorovich, L.V. and Krylov, V.I.; Approximate

Methods of Higher Analysis, New York, Interscience

Publishers Inc., 1964.

[11] Pinto, R.S. and Ramalho, M.A.; Inrcia equivalente

das estruturas de contraventamento de edifcios em

concreto armado. In: Cadernos de Engenharia de

Estruturas, So Carlos, v. 9, n. 38, p. 107-136, 2007.

[12] Schueler, W.; High-rise Building Structures, New

York, John Wiley & Sons, 1977.

[13] Stafford Smith, B. and Coull, A.; Tall Building

Structures: Analysis and Design, New York, John

Wiley & Sons Inc., 1991.
119
A variable limit for the instability parameter of

wall-frame or core-frame bracing structures
Um limite varivel para o parmetro de instabilidade
de estruturas de contraventamento formadas por
associaes de prticos com paredes ou ncleos
R. J. ELLWANGER a
rjellwanger@pop.com.br
Abstract
This work aims to investigate the viability and convenience of adopting a variable limit a1 for the instability parameter of buildings with reinforced
concrete wall-frame or core-frame structures. Initially, the evolution of tall buildings global stability theory is summarized, giving emphasis to
define when a second order analysis is needed. The treatment given to this subject by the present Brazilian code for concrete structures design
(NBR 6118:2007) is also showed. It follows a detailed analytical study that led to the derivation of an equation for the variable limit a1; a series of
examples is presented to check its accuracy. Results are analyzed, showing the validity bounds of the equation and research directions are suggested, in order to improve it.
Keywords: instability, bracing structures, second order.
Resumo
O presente trabalho tem por objetivo investigar a viabilidade e a convenincia de se adotar um limite varivel a1 para o parmetro de instabilidade
de edifcios com estruturas de concreto armado, constitudas por associaes de prticos com paredes ou ncleos. Inicialmente, feito um resumo da evoluo da teoria sobre a anlise da estabilidade global de edifcios altos, especialmente sobre a definio da necessidade ou no de
se realizar uma anlise de segunda ordem; mostra-se tambm como esta questo tratada pela atual norma de projeto de estruturas de concreto
(NBR 6118:2007). Na seqncia, apresenta-se um detalhado estudo analtico que levou ao estabelecimento de uma frmula para o limite varivel
a1, seguido de uma srie de exemplos para testar a validade da mesma. Os resultados so analisados, mostrando-se os limites de validade da
frmula e indicando-se linhas de investigao no sentido de aperfeio-la.
Palavras-chave: instabilidade, estruturas de contraventamento, segunda ordem.
Professor Associado, Departamento de Engenharia Civil, Universidade Federal do Rio Grande do Sul, e-mail: rjellwanger@pop.com.br,
endereo postal: Rua Marcelo Gama 1189/401, CEP 90540-041, Porto Alegre-RS, Brasil.
Received: 14 Sep 2011 Accepted: 18 Nov 2011 Available Online: 01 Feb 2012
2012 IBRACON
R. J. ELLWANGER
1. Introduo
1.1 Os efeitos de segunda ordem em estruturas

de edifcios
Dependendo de seu grau de deformabilidade, a estrutura de contraventamento de um edifcio, quando submetida simultaneamente a aes verticais e de vento, pode desenvolver efeitos adicionais em relao queles que so usualmente determinados em
uma anlise linear ou de primeira ordem (na qual o equilbrio
estudado na configurao geomtrica inicial da estrutura). Tratam-se dos efeitos de segunda ordem, em cuja determinao devem
ser considerados o comportamento no linear dos materiais (no
linearidade fsica) e a configurao deformada na anlise do equilbrio (no linearidade geomtrica).
O trabalho de Beck e Knig [1], apresentado em 1967, representou um marco histrico no desenvolvimento da teoria e
prtica da anlise da estabilidade global de edifcios altos. Foi
estabelecido um critrio, de grande simplicidade de aplicao,
o qual determina que os efeitos de segunda ordem podem ser
desprezados sempre que no representem acrscimo superior
a 10% em relao aos efeitos de primeira ordem. Foi adotado
um modelo simplificado para o sistema de contraventamento,
mostrado na figura 1. Inicialmente, feito o agrupamento de
todas as subestruturas de contraventamento num nico pilar
e de todos os elementos contraventados (elementos portantes que no participam do contraventamento) num conjunto
de barras bi-rotuladas, conforme mostrado na figura 1-a.
admitida uma distribuio uniforme de taxa w para as aes
do vento. P e V representam as aes verticais totais, por andar, transmitidas, respectivamente, s subestruturas de con-
traventamento e aos elementos contraventados. As aes w,

P e V so consideradas com seus valores caractersticos. Em
seguida, para possibilitar a determinao dos efeitos de segunda ordem por meio da tcnica do meio contnuo, adota-se
um sistema aproximado equivalente, mostrado na figura 1-b,
no qual se admite uma distribuio contnua e uniforme de
andares e aes verticais (p = P/h e v = V/h).
Com relao influncia das aes V, atuantes nos elementos
contraventados, Beck e Knig [1] mostraram que, quando o sistema se deforma lateralmente, foras horizontais so transmitidas
atravs das barras dos pavimentos ao sistema de contraventamento, fazendo aumentar o momento fletor na base do mesmo.
Pode-se mostrar que este aumento igual soma dos produtos das foras V pelos deslocamentos horizontais dos respectivos pavimentos. Portanto, no que se refere determinao deste
momento fletor, incluindo os efeitos de segunda ordem, tudo se
passa como se as cargas verticais atuantes no sistema de contraventamento fossem dadas pela soma de suas prprias aes
P com as aes V.
Em 1978, o critrio proposto por Beck e Knig [1] passou
a fazer parte das recomendaes do Comit Euro-International du Bton (CEB [2]). A aplicao do mesmo consiste em comparar os momentos fletores globais na base do
sistema de contraventamento M I (considerando apenas os
efeitos de primeira ordem) e M II (incluindo os efeitos de
segunda ordem):
M II 1,1 M I
(1)
Figura 1 Modelo simplificado do sistema de contraventamento
121
ou
2
1,75wH tot
2
wH tot
1
1
1
1
75
,
3
2
1,75(p + v)H tot
18EJ
(2)
Observe-se que M I e M II so devidos a aes de clculo, uma vez

que as taxas w, p e v esto majoradas pelo coeficiente 1,75. Por
outro lado, a no linearidade fsica considerada, tomando-se para
a rigidez dos elementos estruturais o valor EJ = 0,7 Ecm J, onde Ecm
J representa a soma das rigidezes das subestruturas de contraventamento no estdio I (sem fissurao). Assim, efetuando esta substituio e realizando os devidos algebrismos, chega-se condio:
H tot ( p + v) H tot /( Ecm J ) 0,54
(3)
Segundo Vasconcelos [3], os resultados obtidos por Beck e Knig

[1] s poderiam ser aplicados a estruturas de edifcios cuja rigidez
lateral se concentrasse em poucos pilares, ligados rigidamente entre si, de maneira a poderem ser considerados equivalentes a um
pilar nico, como o da figura 1. A correspondncia deste modelo
com outros tipos de contraventamento (paredes de seo varivel,
prticos etc.) passou a ser feita atravs da igualdade dos deslocamentos horizontais provocados pelas aes horizontais. O pilar
equivalente seria aquele com um fator de rigidez EJ tal que resultassem os mesmos deslocamentos horizontais ocorridos na estrutura
em considerao, para o mesmo carregamento horizontal. Para fins
de simplificao, passou-se a determinar esta rigidez equivalente
com base na atuao de uma carga horizontal unitria no topo do
edifcio. No Brasil, o procedimento passou a ser utilizado com a alterao do coeficiente majorador de 1,75 para 1,40 vide, por exemplo, Sussekind [4] passando a ser conhecido por teste de robustez
mnima. Em conseqncia, a inequao (3) alterou-se para:
H tot ( p + v) H tot /( ES- 28 J I ) 0,60
(4)
onde ES-28 o mdulo de elasticidade secante do concreto aos 28

dias e JI a soma das inrcias das subestruturas de contraventamento no estdio I.
Em 1985, Franco [5] prope que a rigidez do pilar equivalente
seja determinada com base na atuao de uma carga horizontal
uniformemente distribuda, no lugar da carga unitria concentrada
no topo. Alm disso, preconiza que a forma da linha elstica da
estrutura de contraventamento pode afetar a aplicao do critrio
de Beck e Knig [1]. Assim, o coeficiente numrico direita da inequao (4) seria definido em funo do tipo de contraventamento:
n paredes ou ncleos: coeficiente 0,7;
n associaes de paredes ou ncleos com prticos: coeficiente 0,6;
n exclusivamente prticos: coeficiente 0,5.
Em 1995, Franco [6], tratando da considerao da no-linearidade fsica atravs da reduo das rigidezes dos elementos estruturais, prope valores para esta reduo, diferenciados para lajes,
122
barras de pilares e de vigas com armadura simtrica e assimtrica.

Apesar de no fazer parte do escopo do presente trabalho, merece
meno o mtodo baseado no coeficiente de amplificao de momentos gz, apresentado em 1991 por Franco e Vasconcelos [7]. Ele tambm aplica o critrio do acrscimo de 10% em relao aos efeitos de 1
ordem para definir a necessidade ou no de uma anlise de 2 ordem;
aqui, porm, isto feito para cada combinao entre aes horizontais
e verticais. Alm disso, sob certas condies, o prprio mtodo pode
se constituir em uma anlise de 2 ordem. Estas caractersticas fizeram
com que este mtodo fosse rapidamente difundido e passasse a ser
amplamente utilizado no projeto de estruturas de edifcios.
1.2 Prescries da ABNT NBR 6118

A NBR 6118 (ABNT [8]), atual norma para projeto de estruturas de
concreto, incorporou a idia fundamental contida em [1] e [2], ao
estabelecer, em sua seo 15, que os efeitos globais de segunda
ordem so desprezveis quando inferiores a 10% dos respectivos
efeitos de primeira ordem (estrutura com ns fixos). Para verificar a
possibilidade da dispensa da considerao dos esforos globais de
segunda ordem, ou seja, para indicar se a estrutura pode ser classificada como de ns fixos, sem a necessidade de clculo rigoroso,
a ABNT [8] apresenta dois processos aproximados, baseados respectivamente no parmetro de instabilidade e no coeficiente gz. O
primeiro consiste justamente na aplicao do critrio de Beck e Knig [1] e estabelece que: Uma estrutura reticulada simtrica pode
ser considerada como sendo de ns fixos se seu parmetro de instabilidade a for menor que o valor a1, conforme as expresses:
(5)
a = H tot N k /( E C S I C )
a1 = 0,2 + 0,1n ' n 3
a1 = 0,6 ' n 4
(6)
n o nmero de nveis de barras horizontais (andares) acima da

fundao ou de um nvel pouco deslocvel do subsolo. Htot a altura
total da estrutura, medida a partir do topo da fundao ou de um nvel
pouco deslocvel do subsolo. Nk o somatrio de todas as cargas
verticais atuantes na estrutura (a partir do nvel considerado para o
clculo de Htot), com seu valor caracterstico. ECSIC representa o somatrio dos valores de rigidez de todos os pilares na direo considerada. No caso de estruturas de prticos, de trelias ou mistas, ou com
pilares de rigidez varivel ao longo da altura, pode ser considerado o
valor da expresso ECSIC de um pilar equivalente de seo constante. A determinao desta equivalncia ser vista na seo 2.1. IC o
momento de inrcia considerando as sees brutas dos pilares. ECS
o mdulo de elasticidade secante, expresso por:
ECS = 0,85 ECi = 0,85 5600 f ck1/ 2
(7)
R. J. ELLWANGER
ECS, ECi (mdulo de elasticidade tangente) e fck (resistncia caracterstica compresso) so dados em MPa. A norma NBR 6118
tambm incorporou as proposies de Franco [5], ao estabelecer
diferentes valores de a1 em funo do tipo de estrutura de contraventamento: O valor limite a1 = 0,6 prescrito para n > 4 , em geral, aplicvel s estruturas usuais de edifcios. Pode ser adotado
para associaes de pilares-parede e para prticos associados a
pilares-parede. Pode ser aumentado para a1 = 0,7 no caso de contraventamento constitudo exclusivamente por pilares-parede e
deve ser reduzido para a1 = 0,5 quando s houver prticos.
Em uma anlise de segunda ordem, devem ser considerados simultaneamente os efeitos das no-linearidades fsica e geomtrica. A ABNT [8], em seu item 15.7.3, permite que, na obteno dos
esforos globais de segunda ordem em estruturas reticuladas com
quatro ou mais andares, a no-linearidade fsica seja considerada
de forma aproximada. Isso se d mediante uma reduo das rigidezes dos elementos estruturais em funo de ECiIC, ou de ECSIC se
for aplicada a equao (7). Representando as reas das armaduras
longitudinais de trao e de compresso, respectivamente, por As e
As, resultam as seguintes expresses para as rigidezes reduzidas:
n lajes:
(8)

n
vigas:
( EI )sec = 0,4 ECi I C = 0,471ECS IC
' As As'
(9)
( EI )sec = 0,5 ECi I C = 0,588 ECS IC
' As = As'
(10)
pilares:
( EI )sec = 0,8 ECi I C = 0,941 ECS I C
(11)
Alm disso, quando a subestrutura de contraventamento for constituda exclusivamente por vigas e pilares (prtico) e o coeficiente
de avaliao da importncia dos esforos de segunda ordem global (gz) for menor que 1,3 (o que corresponde a uma no-linearidade branda), permite-se considerar a rigidez do conjunto de
barras do prtico como sendo:
(12)
1.3 Justificativa e objetivos da pesquisa

Ao introduzir em seu texto os processos para verificao da
dispensa de considerao dos efeitos globais de 2 ordem,
a ABNT [8] representou um avano em relao norma anterior. No que diz respeito ao parmetro de instabilidade
para edifcios com quatro ou mais andares, deu um tratamento diferenciado aos diferentes tipos de sistemas de contraventamento, ao fixar diferentes valores para o limite a 1.
Todavia, a prescrio de um limite fixo (a 1 = 0,6) para as
associaes de paredes e/ou ncleos com prticos questionvel. medida que pode variar a relao entre as rigidezes de prticos e paredes/ncleos, a 1 tambm pode variar
entre 0,5 e 0,7. Isto pode acarretar dois tipos de erros:
a favor da segurana: em associaes com predominncia
de paredes, a norma limita a 1 em 0,6, quando poderia ser
adotado um valor maior, possivelmente prximo de 0,7;
contra a segurana: no caso de associaes com predominncia de prticos, quando deveria ser adotado um valor
menor do que 0,6, possivelmente prximo de 0,5.
Esses erros, se expressos em funo de a 1, aparentemente
so pequenos. Todavia, convm lembrar que a determinao do parmetro de instabilidade envolve a extrao de
uma raiz quadrada. Consequentemente, ao se verificar a
dispensa da necessidade de uma anlise de 2 ordem, o
erro na determinao da rigidez necessria pode tornar-se
relevante.
Este trabalho tem por objetivo pesquisar uma forma de definio do limite a 1 do parmetro de instabilidade para associaes de prticos com paredes e/ou ncleos, varivel
com a relao entre as rigidezes dos mesmos. Inicialmente, apresentada a formulao do comportamento linear
dessas associaes. Segue-se um estudo analtico sobre o
comportamento no-linear geomtrico de paredes/ncleos
e prticos isolados e, aps, da associao entre os mesmos. O estudo tem por base o modelo simplificado apresentado na seo 1.1, sendo aplicado o critrio expresso
pela inequao (1); as equaes diferenciais so resolvidas
pelo mtodo de Galerkin. A frmula deduzida para o limite
varivel a 1 ento testada em uma srie de exemplos de
edifcios contraventados por associaes de paredes com
prticos. So realizados 88 testes, variando-se os nmeros
de andares, de vos dos prticos e a proporo entre as
rigidezes de prticos e paredes.
2. Anlise linear
2.1 Equivalncia entre as subestruturas

de contraventamento
As subestruturas do tipo paredes ou ncleos caracterizam-se por
serem bastante rgidas ao esforo cortante, predominando as deformaes por flexo. Elas podem ser modeladas por simples barras, engastadas na base do edifcio, comportando-se como pilares. A figura 2-a mostra uma parede ou ncleo modelado por uma
barra engastada-livre, de comprimento l e submetida a uma carga
horizontal uniformemente distribuda de taxa w. Representando
por E, J e M(x), respectivamente, o mdulo de elasticidade longitudinal do material, o momento de inrcia da seo transversal
123
Figura 2 Barras equivalentes s subestruturas de contraventamento
(constante) e a funo de momentos fletores, pode-se expressar a

equao diferencial da linha elstica:
E J d 2 y / dx 2 = E J df / dx = M ( x) = w(l - x) 2 / 2
(13)
Consideram-se positivos os momentos causando trao no lado

esquerdo da barra, ficando a concavidade da deformada voltada
para a direita; f(x) sua declividade. Introduzindo-se as condies
de contorno apropriadas, obtem-se y(x) e o deslocamento horizontal no topo DH:
y ( x) =
wl 4
24 EJ
x 4 4x
- 1
1 - +
l
l
D H = y (l ) = wl 4 / 8EJ
(14)
(15)
por uma barra vertical com predominncia de deformaes por

corte. Conforme mostrado na figura, a deformada desta barra
caracteriza-se por uma declividade f(x) mxima junto base e
tendendo a zero no topo, justamente o contrrio do que ocorre
com a deformada da barra simulando a parede ou ncleo. Esta declividade est relacionada com as diferenas entre deslocamentos
horizontais de andares adjacentes as quais, por sua vez, so proporcionais ao esforo cortante global Q(x). Conforme Stamato [9],
a deformada para este caso descrita pela equao:
S dy / dx = S f ( x) = Q ( x) = w ( l - x)
A constante de proporcionalidade S representa a rigidez do sistema

(prtico plano) ao esforo cortante global; ela anloga ao fator G
A / c de uma barra com deformao por corte, onde G, A e c so,
respectivamente, o mdulo de elasticidade transversal, a rea e o
coeficiente de forma da seo transversal. Resolvendo-se a equao (16), obtem-se y(x) e o deslocamento horizontal no topo:
D H = y (l ) = wl 2 / 2S
Nas subestruturas do tipo prtico plano, predominam as deformaes por flexo das barras individuais de viga e pilar. Quando o
prtico submetido s aes horizontais, o momento fletor global
predominantemente absorvido na forma de esforos normais
nos pilares, para os quais a estrutura proporciona uma grande
rigidez. Por outro lado, o esforo cortante global que causa a
maior parte das deformaes horizontais da estrutura. Portanto,
os prticos podem ser modelados como barras verticais extremamente rgidas ao momento fletor global, nas quais predominam
deformaes por corte.
A figura 2-b mostra um prtico plano submetido a uma carga
horizontal uniformemente distribuda de taxa w. Ele modelado
124
(16)
(17)
As relaes estabelecidas nesta seo tm por objetivo obter a

inrcia de uma barra equivalente a um prtico plano. O item 15.5.2
da ABNT [8], ao tratar do parmetro de instabilidade, estabelece
uma metodologia de determinao do fator ECSIC de um pilar de
seo constante, equivalente a um dado prtico plano. Segundo
esta metodologia, a referida rigidez deve ser obtida calculando-se,
inicialmente, o deslocamento horizontal no topo da estrutura de
contraventamento (prtico) sob a ao do carregamento horizontal, que vem a ser o DH dado por (17). Em seguida, obtem-se a
rigidez de um pilar equivalente de seo constante tal que, sob a
R. J. ELLWANGER
Figura 3 Associao de prticos com paredes ou ncleos
ao do mesmo carregamento, sofra o mesmo deslocamento horizontal no topo o que, neste caso, vem a ser o DH da equao (15).
Isto implica na igualdade entre as duas expresses, o que leva a:
S = 4E J /l 2
(18)
2.2 Associao de prticos com paredes e/ou ncleos

Apresenta-se nesta seo a formulao da resposta linear das
associaes de prticos com paredes/ncleos, para que a mesma seja posteriormente utilizada pelo mtodo de Galerkin na
obteno de uma soluo aproximada para o comportamento
no-linear das mesmas. A figura 3-a mostra o modelo simplificado de um sistema de contraventamento formado por subestruturas dos tipos prtico e parede/ncleo. O modelo consiste
numa parede (representando todas as paredes e ncleos do
sistema) e num prtico (representando todos os prticos do sistema) ligados entre si por bielas (representando as lajes dos
pavimentos). admitida uma distribuio uniforme de taxa w
para as aes do vento. EJ1 representa a rigidez do conjunto de
prticos, conforme a equao (18). EJ2 representa a rigidez do
conjunto de paredes/ncleos.
As figuras 3-b e 3-c mostram as aes s quais estaro submetidos a parede e o prtico, respectivamente. Estas aes consistem em foras concentradas no topo (QT para o prtico e QT para
a parede) e em foras distribudas que podem ser decompostas
em parcelas constantes e variveis. As parcelas constantes (w1
para o prtico e w2 para a parede) so tais que w1 + w2 = w. As
parcelas variveis com a altura (taxa (u(x) para o prtico e u(x)
para a parede), juntamente com as aes QT, representam foras
internas decorrentes da interao entre a parede e o prtico, os
quais, por estarem ligados pelas bielas, ficam impedidos de de-
senvolver suas deformadas naturais, mostradas na figura 3-d. O

prtico estar submetido a uma distribuio de esforos cortantes
globais dada por:
Q( x) = w1 ( l - x) + QT +
u (x )dx
(19)
Conforme foi visto na seo 2.1, o comportamento do prtico descrito pela equao (16). Escrevendo esta equao, introduzindo (18)
e (19) e isolando as parcelas referentes s foras internas, obtem-se:
QT +
u(x )dx
x
4 EJ1
f ( x) - w1 ( l - x)
2
(20)
Por sua vez, a parede estar submetida a uma distribuio de

momentos fletores dada por:
M ( x) = w2 (l - x) 2 / 2 - QT (l - x) -
u(x )(x - x)dx (21)

x
Conforme tambm foi visto na seo 2.1, o comportamento da

parede descrito pela equao (13). Escrevendo esta equao,
entrando com M(x) dado por (21) e derivando ambos os membros,
obtem-se sucessivamente:
EJ 2
df
= w2 (l - x) 2 / 2 - QT (l - x) dx
u (x )(x - x)dx (22)

x
125
3. O mtodo de Galerkin
EJ 2
d f
= - w2 (l - x) + QT +
dx 2
l
x
u (x )dx
(23)
Substituindo (20) em (23) e rearranjando:
EJ 2
d 2f 4 EJ1
- 2 f ( x) + w1 (l - x) + w2 (l - x) = 0
dx 2
l
(24)
wl 2
(l - x)
4 EJ1
(25)
onde:
C1 =
wl 3 e 2 K - 2 K
8EJ 1K e 4 K + 1
y ( x) =
(26)
C2 =
4K
- wl 2 K e + e
8EJ1K
e4K + 1
2K
(27)
(28)
i i
onde i ( x ) (i = 1, 2,..., n) so funes, escolhidas previamente e satisfazendo s mesmas condies de contorno; os ai so

coeficientes a serem determinados. As n funes i ( x ) devem
ser linearmente independentes e pertencer a um sistema, representado por { i ( x ) } (i = 1, 2,..., n) e dotado da propriedade de
plenitude no domnio da soluo. Para que y (x ) seja a soluo
exata da equao dada, necessrio que L( y ) seja identicamente nulo. Esta condio, se L( y ) for contnuo, equivalente condio de ortogonalidade entre a expresso de L( y ) e todas as
funes i ( x ) (i = 1, 2,..., n). Todavia, tendo disponveis apenas
n constantes ai, podem-se satisfazer somente a n condies de
ortogonalidade. Aplicando estas condies, obtem-se o seguinte
sistema de equaes:
a j ( x)
i =1
Considerando que w1 + w2 = w (ao total de vento atuante no

sistema), definindo uma nova varivel K = J 1/J 2 , pode-se expressar a soluo da equao (24) na forma:
f ( x) = C1 e 2 Kx / l + C2 e - 2 Kx / l +
Em muitos problemas de Engenharia, como os apresentados nas

prximas sees, surge a necessidade de se resolver uma equao do tipo L(y) = 0, onde L um operador diferencial, cuja soluo satisfaz a condies de contorno homogneas. O mtodo de
Galerkin consiste em obter uma soluo aproximada, na forma:
L( y( x)) ji ( x)dx =
n
L a jj j ( x) ji ( x)dx = 0
j =1
(i = 1, 2, . . . , n)
(29)
A soluo deste sistema (linear, no caso de L ser um operador

linear) fornece os valores dos coeficientes ai, a partir dos quais se
chega soluo aproximada y (x ) . A demonstrao da convergncia, bem como consideraes mais detalhadas sobre o mtodo de Galerkin podem ser vistas em Kantorovich e Krylov [10].
Figura 4 Influncia das deformaes na resposta da estrutura
A Barra equivalente ao contraventamento
126
Deformao por corte a nvel infinitesimal
R. J. ELLWANGER
4. Dispensa da considerao dos efeitos

de segunda ordem
As sees 4.1 e 4.2 apresentam a formulao do comportamento
no-linear geomtrico, respectivamente, de conjuntos de paredes/
ncleos e de prticos. Para ambos os casos, so deduzidos os
limites a1 do parmetro de instabilidade, confrontando-os com os
valores prescritos pela ABNT [8]. A seo 4.3 realiza o mesmo para
as associaes destes tipos de subestruturas, obtendo-se uma expresso para o limite varivel a1, objetivo principal deste trabalho.
A figura 4-a mostra a deformada de uma barra equivalente a
um sistema de contraventamento, submetida a cargas uniformemente distribudas de taxas w e q, respectivamente, nas direes
horizontal e vertical; q dado pela soma das taxas p e v da figura
1-b. Levando em considerao a deformao da barra (no-linearidade geomtrica) e representando por Y a funo primitiva dos
deslocamentos y(x), o momento fletor ser dado por:
M ( x) = w(l - x) / 2 +
l
x
q[ y(x ) - y ( x)]dx
(30)
mtodo de Galerkin. Assumindo que esta soluo seja proporcional f(x)

devida exclusivamente aos efeitos de primeira ordem, pode-se escrever:
3

x
f ( x) = a1j1 ( x) = a1 1 - 1 -
l

(35)
onde 1 ( x ) foi obtido derivando-se a equao (14) em relao a

x e suprimindo-se a constante que ficaria em evidncia. Aplicando
a equao (29) com n = 1, obtem-se sucessivamente:
l
0
L(f ) j1 ( x) dx =
L(a j ( x)) j ( x)dx = 0 (36)

0
1 1
x 3
x 3
6EJa1 x
2 - 1 + q(l - x )a1 1 - 1 - + w(l - x) 1 - 1 - dx = 0
l
l
0 l l
(37)
Efetuando-se a integrao e isolando a1, resulta:
ou
M ( x) = w(l - x) / 2 + q[Y (l ) - Y ( x) - (l - x) y ( x)]

2
(31)
Considerando que y(0) = 0, o momento fletor na base ser expresso por:
M (0) = wl 2 2 + q[Y (l ) - Y (0)]
(32)
No caso de contraventamento formado exclusivamente por paredes e/ou ncleos, a equao diferencial da linha elstica ser obtida substituindo-se M(x) dado por (31) na equao (13):
(33)
Derivando-a em relao a x e considerando que as rotaes so

dadas por f(x) = dy/dx, a equao (33) transforma-se em:
E J d 2f dx 2 + q(l - x)f ( x) + w(l - x) = 0
(34)
Pode-se obter uma soluo aproximada para a equao (34) por meio do
4 wl 3
24 EJ - 3ql 3
(38)
Substituindo (38) em (35), obtem-se a soluo aproximada:
f ( x) =
4 wl 3
24 EJ - 3ql 3

1 - 1
(39)
Integrando (39) em relao a x e aplicando a condio de deslocamento nulo na base, obtem-se a funo dos deslocamentos.
Integrando novamente, obtem-se:
4.1 Subestruturas do tipo paredes ou ncleos
E J d 2 y / dx 2 = w(l - x) 2 / 2 + q[Y (l ) - Y ( x) - (l - x) y ( x)]
a1 =
Y ( x) =
l x 5 2 x 2
wl 4
- x + C
- 1 - +
l
24 EJ - 3ql 3 5 l
(40)
onde C a constante de integrao. O momento fletor na base

pode ser obtido, combinando-se as equaes (32) e (40):
M ( 0) =
wl 2
2qwl 5
+
2
5(8 EJ - ql 3 )
(41)
A equao (41) pode ser transformada sucessivamente em:
M (0) =
wl 2 8EJ - ql 3 / 5
4ql 3
wl 2
=
1 +
2 5(8EJ - ql 3 )
2 8 EJ - ql 3
(42)
127
A condio de que, no estado limite ltimo (cargas majoradas em

1,4), os efeitos de 2 ordem no podem superar em mais de 10%
os efeitos de 1 ordem (inequao (1)), aplicada ao momento
fletor na base, obtendo-se:
1,4 wl 2 8EJ - 1,4ql 3 /5

1,4 wl 2
1,1
2
2
8 EJ - 1,4ql 3
(43)
Os termos wl2, em evidncia em ambos os lados da inequao, desaparecem. Efetuando-se os devidos algebrismos, obtem-se:
ql 3 / EJ 0,6349
(44)
Como uma parede ou ncleo tem comportamento equivalente

ao de um pilar, a considerao da no-linearidade fsica pode
ser feita adotando-se para EJ a expresso 0,941 ECSIC, de acordo com a equao (11). Por outro lado, lembrando que ql a
carga vertical total Nk e que l a altura Htot, a inequao (44)
transforma-se em:
dM
- dM dx w(l - x) + q(l - x)f ( x)
=
=
ds
1 + f 2 ( x)
1 + f 2 ( x)
Q( x) = -
(48)
Trata-se de um esforo cortante inclinado, conforme mostra a figura 4-b. A deformao por corte provocado por ele tem a mesma
inclinao, sendo dada, a nvel infinitesimal, por:
dy
dy
dy
=
=
= dy 1 + f 2
cos f dx ds 1 1 + f 2
(49)
Ao estabelecer a equao diferencial da linha elstica para este

caso, devem-se realizar duas adaptaes em relao equao
(16): entrar com dy/cosf, dado por (49), no lugar de dy e com Q(x)
dado por (48). Desta forma, obtem-se:
dy
Q(x)
w( l - x) + q( l - x)f(x)
1 + f 2 ( x) =
=
dx
S
S 1 + f 2(x)
(50)
Assim:
2
tot
N k / ECS I C 0,5974
(45)
f ( x) =
Extraindo a raiz quadrada de ambos os membros:
(46)
A inequao (46) indica, assim, um valor a1 = 0,773. Por sua vez,

a ABNT [8] permite que seja aumentado para at 0,7 o coeficiente
a1, quando o contraventamento for constitudo exclusivamente por
paredes ou ncleos.
4.2 Subestruturas do tipo prticos planos
(47)
O esforo cortante pode ser obtido a partir da derivao da equao (31) em relao ao eixo deformado da barra. Introduzindo ds
dado por (47), obtem-se:
128
(51)
Nos casos de que trata o presente trabalho, isto , de

no-linearidade geomtrica branda, as rotaes f(x) apresentam valores bem menores do que a unidade; portanto, f 2(x)
pode ser desprezado na presena de 1 e a equao (51) pode
ser posta na forma:
Sf ( x) = w(l - x) + q(l - x) f ( x)
(52)
Isolando f(x):
No caso de contraventamento formado exclusivamente por prticos, a barra equivalente da figura 4-a ter predominncia de deformaes por corte. Tambm neste caso, a expresso das solicitaes deve levar em considerao a configurao deformada.
Pode-se provar que os infinitsimos ds e dx da figura 4-b esto
relacionados por:
ds = dx2 + dy2 = dx2 (1+ dy2 dx2 ) = dx 1+f 2
dy
w(l - x) + q(l - x)f ( x)
=
dx
S[1 + f 2 ( x)]
f ( x) =
w(l - x)
S - q (l - x )
(53)
Integrando a equao (53) em relao a x e aplicando a condio

de deslocamento nulo na base, obtem-se a funo dos deslocamentos. Integrando novamente, obtem-se:
Y(x) =
Sw
[S - q(l - x)] {ln[S - q(l - x)]-1}- Sw2 x ln(S - q l) - w (l 2 + x2 ) + C
3
2q
q
q
(54)
R. J. ELLWANGER
onde C a constante de integrao. Aplicando a equao (54)

para x = 0 e x = l, pode-se expressar a diferena:
Y(l ) - Y( 0 ) =
S 2w
S
wl l S
+
ln
3
S - ql
q 2 q
q
(55)
Assim, o momento fletor na base pode ser expresso, substituindo-se a equao (55) na (32):
M( 0 ) =
l S S 2w
wl 2 S 2 w S
Swl
1
+ 2 ln
- wl + = 2 ln
2 q S - ql
2 q q 1 - ql / S q
(56)
das vigas para a deformao lateral de um prtico pode chegar a 65%,

restando 35% devidos flexibilidade dos pilares. Assim, considerando
a situao de projeto de um prtico esbelto (na qual a predominncia
dos efeitos do vento leva a uma tendncia de igualdade entre as armaduras As e As das vigas), podem-se utilizar as equaes (10) e (11)
para relacionar as parcelas dos deslocamentos horizontais do prtico
com a considerao da no-linearidade fsica (yNL), relativas s vigas
(yNLVIGAS) e aos pilares (yNLPILARES), com as correspondentes parcelas
(yLVIGAS) e (yLPILARES) dos deslocamentos horizontais resultantes da anlise linear (yL). Simultaneamente, podem-se aplicar as recm mencionadas propores de 35% e 65% de participao destas parcelas em relao aos deslocamentos totais, resultando as seguintes expresses:
PILARES
y NL
=
Aplicando-se a este momento fletor a condio expressa pela inequao (1), resulta:
wl 2
1,4 S 2 w
1
1,4 Swl
,
ln
1
1
1
4
1,4 q
2
1,4 2 q 2 1 - 1,4ql S
(57)
Efetuando-se os devidos algebrismos, a inequao (57) transforma-se em:
1
1
1
ln
0,77
2
l
l
1
1
,
4
q
/S
q
/S
1,4(ql /S)
(58)
Tomando-se o fator ql/S como incgnita, pode-se resolver a inequao (58) por meio de tentativas, obtendo-se:
ql / S 0,0962
(59)
Substituindo S pela expresso (18), resulta:
l 2 ql /EJ 0,3848
(60)
VIGAS
y NL
=
yLPILARES
0,35 y L
=
0,941
0,941
y LVIGAS
0,65 y L
=
0,588
0,588
(62)
Em seguida, efetuando a soma das parcelas expressas por (61)

e (62), obtem-se a seguinte relao entre os deslocamentos horizontais totais yNL e yL:
y NL =
yL
0,35 y L
0,65 y L
+
=
0,941
0,588
0,677
(63)
Como a rigidez lateral do prtico inversamente proporcional a

esses deslocamentos, pode-se escrever:
( EI ) sec = 0,677 ECS I C
(64)
Considerando esta expresso de (EI)sec e seguindo a mesma linha

dedutiva que levou s inequaes (45) e (46), obtem-se:
De acordo com a ABNT [8], a no-linearidade fsica poderia ser
considerada, substituindo-se EJ por (EI)sec dado por (12). Todavia,
a relao (EI)sec / ECSIC do conjunto de barras do prtico, como uma
funo das relaes (EI)sec / ECSIC das barras individuais, no pode ser
considerada fixa; ela pode variar em funo de vrios fatores, como
nmero e altura dos andares, nmero e extenso dos vos, relao
entre as dimenses transversais de vigas e pilares etc. Pinto e Ramalho [11] mostram que a influncia da no linearidade fsica na rigidez
lateral dos prticos depende principalmente das taxas de armadura e
da magnitude do carregamento aplicado, tendo obtido relaes (EI)sec
/ ECSIC para o estado limite ltimo variando entre 0,51 e 0,75.
Por outro lado, Schueler [12] afirma que a contribuio da flexibilidade
(61)
(65)
Observe-se a coerncia desta inequao com a ABNT [8], a qual

fixa em 0,5 o coeficiente a1, quando o contraventamento for constitudo exclusivamente por prticos. Na verdade, para obter-se a1
= 0,5, dever-se-ia ter:
( EI ) sec = 0,650 ECS I C = 0,552 ECi I C
(66)
129
4.3 Associaes de prticos com paredes

e/ou ncleos
soluo linear, expressa pelas equaes (25), (26) e (27):
adotado o mesmo modelo da figura 3 e so consideradas as

mesmas definies da seo 2.2. Ao estabelecer a equao diferencial da linha elstica para o conjunto de prticos (figura 3-c),
aplica-se a equao (50), acrescentando ao esforo cortante as
parcelas relativas s foras de interao entre paredes e prticos,
tal como foi feito na equao (19):
Q(x)
=
S
w1 ( l- x) + QT + u (x ) dx + q1 (l - x)f ( x)
x
S 1 + f ( x)
= f ( x) 1 + f 2 ( x )
(67)
Considerando que f (x) pode ser desprezado na presena de 1 e

isolando as parcelas referentes s foras de interao:
2
QT +
u (x )dx
= Sf ( x) - w1 (l - x) - q1 (l - x)f ( x)
(68)
Ao estabelecer a equao diferencial da linha elstica para o conjunto de paredes (figura 3-b), entra-se com o momento fletor dado
por (31) acrescido das parcelas relativas s foras de interao,
tal como foi feito na equao (21):
2
l
w2 (l - x)
EJ d y
- QT (l - x) - u(x )(x - x)dx + q2 [Y (l ) - Y ( x) - (l - x) y( x)]= 2 2
x
2
dx
(69)
l
d3y
d 2f
=
EJ
2 2 = - w2 (l - x) + QT + u (x ) dx - q2 (l - x) f ( x)
3
x
dx
dx
(70)
Substituindo (68) em (70) e rearranjando, resulta:
EJ 2
d 2f
= - (w1 + w2 ) (l - x) - (q1 + q2 ) (l - x) f ( x) + Sf ( x)
dx 2
(71)
Considerando que w1 + w2 = w (ao total do vento), q1 + q2 = q (ao

gravitacional total) e rearranjando novamente, obtem-se a equao
diferencial que governa o comportamento de um sistema de prticos
e paredes/ncleos incluindo a influncia das deformaes:
EJ 2
d 2f
- [ S - q (l - x)]f ( x) + w(l - x) = 0
dx 2
(73)
Substituindo (73) em (72), resulta:
wl 2
EJ 2 f " ( x) C1 e2 Kx / l + C2 e-2 Kx / l + (l - x) +
4EJ1
2K
wl 2 4 K 2
2EJ 2 f ' ( x) C1 e 2Kx / l - C2 e -2 Kx / l - + 2 EJ 2 f ( x) C1 e2 Kx / l + C2 e-2Kx / l 4EJ1 l
l
) (74)
wl 2
f ( x) [S - q(l - x)]C1 e2 Kx / l + C2 e-2Kx / l + (l - x) + w(l - x) = 0
4EJ1
Assumindo que f(x) seja uma funo constante, implica em serem

nulas a primeira e a segunda parcelas da equao (74), uma vez
que elas esto com as derivadas f (x) e f (x) em evidncia. Alm
disso, considerando a definio j existente de K, bem como a de
S (equao 18), pode-se escrever:
4K 2
4 EJ1
EJ 2 =
= S
2
l
l2
(75)
Em conseqncia disto, a terceira parcela da equao (74) cancela-se com alguns termos da quarta e a equao fica reduzida a:
Derivando a equao (69) em relao a x, obtem-se:
EJ 2
wl 2
f ( x) = f ( x) C1 e 2 Kx / l + C2 e -2 Kx / l +
(l - x )
4 EJ1
(72)
wl 2
q f ( x) C1 e 2 Kx / l + C2 e -2 Kx / l +
(l - x) + w [1 - f ( x)] = 0
4 EJ1
(76)
O mtodo de Galerkin ser utilizado para encontrar uma funo f (x) = constante, que seja uma boa aproximao para a
f(x) envolvida na equao (76). De acordo com (28), pode-se
escrever:
f (x) = a1j1 ( x)
(77)
' j1 ( x) = 1
A aplicao da equao (29) para este caso assume a forma:
{a q [C e
l
2 Kx/l
+ C2 e -2 Kx/l + wl 2 (l - x) 4EJ1 + w (1 - a1 ) dx = 0
(78)
Para a aplicao do mtodo de Galerkin equao (72), ser assumida uma soluo dada pelo produto de uma funo f(x) pela
130
R. J. ELLWANGER
Efetuando-se a integral, pode-se isolar a1, obtendo-se:
a1 =
ql
q
C1 (e 2 K - 1) - C2 (e - 2 K - 1)
8EJ 1 2 Kw
1-
Aplicando-se a este momento fletor a condio expressa pela inequao (1), resulta:
] (79)
Portanto, a soluo aproximada para a equao (72) ser dada por:
wl 2
f ( x) = a1 C1 e 2 Kx / l + C2 e - 2 Kx / l +
(l - x)
4 EJ 1
(80)
com a1 dado por (79). Integrando-a duas vezes, obtem-se a primitiva dos deslocamentos:
l2
wl 2 x 2 l x
Y ( x) = a1 2 C1 e 2 Kx / l + C2 e - 2Kx / l +
- + C3 x + C 4
4EJ1 2 6
4K
(81)
- wl 4 K(e 4 K - 1 ) + e 2 K
8 EJ1
K 2(e 4 K + 1 )
(82)
O momento fletor na base obtido, aplicando-se a equao (32):
M(0) =
wl
wl
+ a1 q 2 C1 (e 2 K - 1) + C2 (e - 2 K - 1) +
+ C3 l
2
12 EJ1
4K
(83)
Observe-se que a diferena Y(l) Y(0) fez desaparecer a constante C4. Substituindo C1, C2 e C3, respectivamente, pelas equaes
(26), (27) e (82), leva a equao (83) a assumir a forma:
M(0) =
wl 2 a1qwl 5 (6K + 8K 3 )(e4 K + 1) + (3 - 12K 2 )(e4 K - 1) - 24Ke 2K

+
2 96 EJ1K 3
e 4K + 1
(84)
Introduzindo as expresses de C1 e C2 tambm na equao (79) e

substituindo a frmula de a1 assim obtida na equao (84), obtem-se a expresso do momento fletor na base do sistema formado
por prticos e paredes/ncleos:
M(0) =
wl 2
(2K + 8K 3 / 3)(e 4 K + 1) + (1 - 4K 2 )(e4 K - 1) - 8Ke 2K
1 +
3 3 4K
3 4K
2 4K
2K
2 (16EJ1 /ql )K (e + 1) - ( K + 2K )(e + 1) + 2K (e - 1) + 2Ke
(85)
(86)
Na inequao (86), pode-se isolar o fator ql3/EJ1, obtendo-se:
ql 3
(24 / 7) K 3 (e 4 K + 1)
EJ1 (6,3K + 8,6K 3 ) (e 4 K + 1) + (3 - 12,6K 2 ) (e 4 K - 1) - 24,6 Ke 2 K
(87)
Chamando de J a soma das inrcias dos conjuntos de paredes/

ncleos e de prticos e considerando a definio anterior de K,
pode-se escrever:
J = J1 + J 2 = J1 + J1 K 2 = J1 ( K 2 + 1) K 2
sendo C4 uma constante indeterminada e C3 resultante da condio de deslocamento nulo na base:
C3 =
1,4 wl 2
(2K + 8K 3 / 3)(e 4K + 1) + (1 - 4K 2 )(e 4K - 1) - 8Ke 2K
1,4 wl 2
1
,
1
1 +
2 (16EJ1K 3 / 1,4ql 3 - 2K 3 - K )(e 4K + 1) + 2K 2 (e 4K -1) + 2 Ke 2K

2
(88)
Isolando J1:
J1 = K 2 J ( K 2 + 1)
(89)
Definem-se IC1 e IC2 como as inrcias brutas, respectivamente, dos

conjuntos de prticos e de paredes. Chamando de IC a soma de IC1
e IC2 e aplicando as relaes (11) e (66), tem-se que:
I C = I C1 + I C 2 =
J1 J 2
J
J
1,5385 K 2 + 1,0625
+
= 1 + 1 2=
J1
0,65 0,941 0,65 0,941 K
K2
(90)
A partir de (90), o fator EJ1 pode ser expresso por:
EJ1 =
K2
ECS I C
1,5385 K 2 + 1,0625
(91)
A equao (87) pode ser reescrita, substituindo-se EJ1 pela expresso

(91), ql por Nk (carga vertical total) e l por Htot (altura do edifcio). Em
seguida, extraindo a raiz quadrada de ambos os membros, resulta:
H tot
Nk
1
ECS I C
(92)
131
onde
(24/7) K 5 (e4 K + 1)
1 =
(1,5385K 2 + 1,0625)[(6,3K + 8,6K 3 )(e4 K + 1) + (3 - 12,6K 2 )(e4K - 1) - 24,6Ke2 K ]
(93)
Obteve-se assim uma expresso para o limite a1 do parmetro

de instabilidade, varivel com K (relao entre as inrcias reduzidas de prticos e paredes/ncleos). Para fins de obteno
de a1, no entanto, mais prtico trabalhar com as inrcias
brutas. Combinando as equaes (66) e (90), obtem-se a seguinte relao entre K e I C1/IC (razo entre a inrcia bruta dos
prticos e a total):
K = 0,831 ( I C1 / I C ) /(1 - I C 1 / I C )
(94)
Assim, dada uma razo I C1/I C qualquer, obtem-se K pela aplicao da equao (94) e em seguida a 1 pela aplicao da
(93). A seqncia de valores de a 1, apresentada na tabela 1
e representada graficamente nas figuras 7 e 8, mostra uma
variao mais brusca para I C1/I C prximo de 1 (predominncia
de prticos) e mais suave para I C1/I C prximo de 0 (predominncia de paredes). Observa-se tambm que a equao
(93), nos extremos de seu intervalo de aplicao, reproduz
fielmente a equao (46), mas apresenta uma diferena de
1,8% para contraventamento formado exclusivamente por
prticos. A seguir, a equao (93) ser testada em uma srie
de exemplos.
5. Exemplos
5.1 Descrio dos testes
A figura 5 mostra a configurao bsica, em planta, do sistema de
contraventamento transversal de um edifcio de planta retangular
(exemplos 1, 3, 5 e 7), o qual constitudo pelas paredes 1 e 5
junto s fachadas laterais e pelos prticos 2, 3 e 4, de vo nico
(7,5 m de eixo a eixo de pilar). Da mesma forma, pode-se observar
na figura 6 a configurao bsica do sistema de contraventamento
transversal de um edifcio de planta octogonal alongada (exem-
Table 1 Values of a1, varying the IC1/ICratio

IC1/IC
a1
IC1/IC
a1
IC1/IC
a1
0
0,10
0,20
0,30
0,40
0,773
0,772
0,771
0,768
0,763
0,50
0,60
0,70
0,80
0,85
0,755
0,744
0,726
0,699
0,679
0,90
0,95
0,98
0,99
1,00
0,651
0,611
0,574
0,555
0,509
plos 2, 4, 6 e 8), sendo tambm formado por duas paredes e trs

prticos que, neste caso, possuem trs vos iguais (5 m de eixo a
eixo de pilar). Cada um destes sistemas foi empregado em edifcios de 5, 10, 20 e 30 andares com p direito de 3 m, constituindo
os exemplos 1 a 8, cujas informaes gerais constam na tabela 2.
Para cada um dos oito edifcios, foram realizados testes variando as sees retangulares das paredes e das barras dos prticos,
de forma a resultar na srie de razes IC1/IC apresentada na tabela
3. Isto exigiu, em alguns casos, alteraes nas configuraes bsicas das figuras 5 e 6, mantendo-se, porm, a dupla simetria do
contraventamento em planta. Para IC1/IC = 1 foram suprimidas as
paredes; para valores decrescentes de IC1/IC, os prticos eram gradativamente suprimidos, ficando completamente eliminados para
IC1/IC = 0. Nos exemplos de 30 andares, foram includos prticos
adicionais em alguns casos e paredes adicionais em outros. As
dimenses das sees transversais adotadas nos testes encontram-se na tabela 4.
Considerou-se a utilizao de um concreto com fck = 25 MPa, resultando em um mdulo ECS = 23800 MPa. Foi considerada uma
carga vertical total (valor caracterstico) de 10 kN/m2 por pavimento.
Adotou-se uma carga devida presso do vento de 1,5 kN/m2 (valor
caracterstico), constante ao longo da altura. Esta considerao foi
feita por tratar-se da primeira experincia com uma formulao baseada num modelo com carga de vento de taxa constante.
Assim, foram testados 88 sistemas de contraventamento. Cada
teste teve por objetivo determinar a relao entre cargas verticais
e rigidez horizontal que resultasse em um acrscimo de 10% no
momento global da base do edifcio, em relao anlise de primeira ordem; com isso, determinava-se o limite a1 do parmetro
de instabilidade. O procedimento aplicado em cada teste consistiu
em, inicialmente, fixar as dimenses das sees das barras do
Figura 5 Sistema de contraventamento transversal: exemplos 1, 3, 5 e 7
132
R. J. ELLWANGER
Figura 6 Sistema de contraventamento transversal: exemplos 2, 4, 6 e 8
Tabela 2 Informaes gerais dos exemplos 1 a 8

Exemplo
N de andares
5
Altura (m)
15
N de vos de cada prtico
1
N de prticos
0a3
N de paredes
0a2
5
15
3
0a3
0a2
10
30
1
0a3
0a2
10
30
3
0a3
0a2
20
60
1
0a3
0a2
20
60
3
0a3
0a2
30
90
1
0a5
0a2
30
90
3
0a5
0a4
conjunto de prticos e calcular sua rigidez horizontal IC1, de acordo

com o item 15.5.2 da ABNT [8] (relao entre a carga de vento
e o deslocamento horizontal ocorrido no topo da estrutura). Em
seguida, ajustou-se a seo do conjunto de paredes de forma a
obter-se a razo IC1/IC desejada.
Na seqncia do teste, realizou-se uma anlise inicial de segunda

ordem do conjunto prticos-paredes, empregando-se o mtodo
P-D. Em seguida, realizaram-se novas anlises de segunda ordem, ajustando-se o valor das cargas verticais at que resultasse
o acrscimo desejado de 10% no momento global da base. Embo-
Tabela 3 Razes IC1/IC adotadas nos exemplos

1,00
0,95
0,90
0,85
0,80
0,70
0,60
0,50
0,40
0,20
Tabela 4 Dimenses (cm) das sees transversais

Exemplo
Vigas
Pilares
Paredes
1
2
3
4
5
6
7
8
20 x 50 a 25 x 60
15 x 34 a 20 x 50
18 x 68 a 24 x 68
15 x 51 a 24 x 59,5
34 x 85
21 x 68
34 x 85
21 x 85
20 x 50 a 25x 60
15 x 34 a 20x 50
30 x 85 a 40 x 85
20 x 77 a 34x 90
42,5 x 140
34 x 96
42,5 x 140
36 x 119
20 x 88,5 a 20 x 201
20 x 102 a 20 x 243
20 x 158,5 a 20 x 422,5
20 x 224 a 20 x 481,5
20 x 307,5 a 25 x 838,5
20 x 351 a 25 x 957
20 x 418 a 35 x 1200
20 x 588 a 31 x 998,5
133
ra a lgica indique que o ajustamento devesse ser feito na rigidez

horizontal, optou-se pelo ajustamento das cargas, pois, alm de
no afetar os resultados, tornou mais gil a realizao dos 88 testes. A no-linearidade fsica foi considerada por meio da reduo
da rigidez das barras individuais, expressa pelas equaes (10) e
(11). Adotou-se para a anlise um modelo de prtico plano como o
da figura 3, com os conjuntos de prticos e de paredes unidos por
bielas, por tratar-se do modelo no qual est baseada a formulao proposta neste trabalho. Posteriormente, alguns casos foram
re-analisados por um mtodo com emprego de matriz de rigidez
geomtrica, a fim de confirmar os resultados obtidos pelo P-D.
5.2 Anlise dos resultados

Os valores de a1 obtidos nos testes encontram-se na tabela 5.
Para a interpretao dos resultados, conveniente agrupar os oito
exemplos por nmero de andares e considerar dois intervalos de
valores da rigidez relativa: IC1/IC < 0,9 e IC1/IC > 0,9. As figuras 7 e
8 mostram, para cada nmero de andares, uma curva representando a variao do limite a1 encontrado nos testes, bem como a
curva de a1 correspondente frmula (93).
Ao analisar os resultados referentes ao intervalo IC1/IC < 0,9, constata-se que os valores de a1 obtidos nos exemplos esto, em sua quase
totalidade, abaixo dos valores previstos pela frmula (93), apresentados na tabela 1. Assim, a aplicao desta frmula resulta em erros
para mais, cujos valores mximos so apresentados na tabela 6.
Observa-se claramente que estes erros decrescem com o aumento
de andares; iniciam com 16,2% aos 5 andares e diminuem para 3,1%
aos 30 andares, mostrando uma tendncia de se anularem para um
nmero de andares pouco superior a 30. Esse desempenho da equao (93) deve-se muito provavelmente adoo do modelo da figura 1-b no lugar do da figura 1-a; ou seja, o modelo com distribuio
uniforme e contnua de andares e aes verticais proporciona uma
preciso aceitvel apenas para edifcios a partir de 30 andares.
Tabela 5 Valores de a1, variando-se a razo IC1/IC, o nmero de andares e o de vos

IC1/IC
1,00
0,95
0,90
0,85
0,80
0,70
0,60
0,50
0,40
0,20
0
Exemplo:
Andares:
1
5
2
5
3
10
4
10
5
20
6
20
7
30
8
30
0,515
0,552
0,572
0,590
0,610
0,626
0,641
0,652
0,662
0,675
0,683
0,514
0,557
0,584
0,603
0,619
0,641
0,656
0,665
0,671
0,681
0,683
0,528
0,567
0,594
0,613
0,629
0,653
0,671
0,685
0,695
0,714
0,726
0,519
0,563
0,592
0,614
0,632
0,657
0,676
0,689
0,699
0,715
0,726
0,569
0,605
0,630
0,650
0,663
0,687
0,702
0,716
0,724
0,738
0,749
0,534
0,590
0,621
0,644
0,662
0,687
0,702
0,716
0,726
0,736
0,749
0,608
0,639
0,656
0,675
0,690
0,710
0,724
0,734
0,743
0,753
0,764
0,591
0,635
0,656
0,676
0,690
0,711
0,727
0,735
0,740
0,752
0,764
Figura 7 Grficos a1 x razo IC1/IC

para os exemplos de 5 e 10 andares
134
Figura 8 Grficos a1 x razo IC1/IC

para os exemplos de 20 e 30 andares
R. J. ELLWANGER
Tabela 6 Erros mximos (%)

IC1/IC
5 andares
10 andares
20 andares
30 andares
0,90
= 0,95
= 1,00
+16,2
+10,7
1,2
+11,2
+8,5
3,6
+6,0
+3,6
10,5
+3,1
4,4
16,3
Por outro lado, para valores de IC1/IC superiores a 0,9, indicativos

de uma alta predominncia de prticos, a tendncia de erros decrescentes com o aumento de andares tambm existe. Todavia,
aqui ela est acompanhada por outra tendncia que a de erros
para menos produzidos pela equao (93), crescentes com o nmero de andares, conforme se pode observar na tabela 6 para
IC1/IC = 0,95 e 1,00. Essa tendncia devida ao fato da equao (93) levar em considerao que os prticos se deformam
somente devido ao esforo cortante global. Smith e Coull [13]
afirmam que, em edifcios esbeltos, a flexo global dos prticos,
devida deformao axial dos pilares, pode ter uma contribuio significativa para os deslocamentos horizontais. Trata-se do
mesmo padro de deformao das paredes, o que faz aumentar
o coeficiente limite a1. Assim, quanto mais alto o edifcio e maior
a presena de prticos, maiores sero os erros gerados por uma
formulao que ignore esse efeito. No caso de contraventamento
formado exclusivamente por prticos (IC1/IC = 1), observa-se na
tabela 6 que os erros mximos variam entre 1,2% (5 andares) e
16,3% (30 andares).
Com relao, ainda, ao caso IC1/IC = 1, os valores de a1, mostrados na primeira linha da tabela 5, sugerem ser conservador o
limite a1 = 0,5 prescrito pela ABNT [8] para o caso de contraventamento formado exclusivamente por prticos, principalmente
em edifcios com mais de 10 andares. Por outro lado, os valores
de a1 encontrados para IC1/IC = 0, mostrados na ltima linha
da tabela 5, indicam ser tambm conservador o valor de a1 =
0,7 prescrito para o contraventamento formado exclusivamente por paredes/ncleos, em edifcios com mais de 10 andares.
Todavia, em edifcios com menos de 10 andares, pode ocorrer
o contrrio. Nos exemplos analisados, encontraram-se valores
ligeiramente inferiores a 0,7 nos edifcios de 5 andares. Alm
disso, a adoo do valor fixo a1 = 0,6 para associaes de prticos com paredes e/ ou ncleos deveria estar condicionada a
um limite mnimo de contribuio das paredes para a rigidez
do contraventamento, principalmente em edifcios mais baixos.
Nos exemplos analisados, interpolaes feitas na tabela 5 mostram que a inrcia bruta das paredes em relao inrcia total
deveria ser de pelo menos 18% no exemplo 1, 14% no 2 e 12%
nos exemplos 3 e 4.
envolve a extrao de uma raiz quadrada, a proporo entre

os radicandos (relaes carga vertical/rigidez horizontal) associados a esses extremos superior a 2:1. Esta variabilidade
mostra a importncia de se ter uma forma de prever-se um
limite a 1 apropriado relao I C1/IC e ao nmero de andares
de um dado edifcio a ser projetado, no lugar dos valores fixos
estabelecidos pela ABNT [8].
A equao (93) representa uma tentativa inicial de se realizar
tal previso. A preciso relativamente boa obtida nos exemplos 7
e 8 para IC1/IC < 0,9 mostra que o objetivo vivel e que esforos
merecem ser realizados no sentido de atingi-lo. Para fazer frente
aos erros encontrados nos demais casos (predominncia de prticos e menor nmero de andares), deve-se introduzir o efeito da
deformabilidade axial dos pilares dos prticos nas equaes (16)
e (50), fazendo com que a curva de a1 , correspondente equao
(93) e mostrada nas figuras 7 e 8, no decline tanto em seu trecho
final; deve-se tambm encontrar um meio de adequar a formulao variao do nmero de andares. Outra questo a investigar
a viabilidade de se incorporar formulao a variabilidade da
influncia da no-linearidade fsica na rigidez lateral dos prticos
(atualmente, esta influncia considerada na forma de um coeficiente constante).
Convm salientar que tudo isto deve ser feito de forma a manter a simplicidade da formulao, justamente uma das maiores
virtudes da utilizao do parmetro de instabilidade. Finalmente,
convm destacar a necessidade de se adotar para os testes um
modelo mais realstico de anlise: simulao da estrutura como
um reticulado tridimensional considerando-se os pavimentos
como diafragmas rgidos; variao da carga de vento ao longo
da altura do edifcio; realizao da anlise no-linear atravs de
um mtodo incremental-iterativo; e considerao mais acurada da
no-linearidade fsica, por exemplo, por meio das relaes momento-curvatura.
6. Concluses

[01] Beck, H. and Knig, G.; Haltekrfte im Skeletbau.
In: Beton- und Stahlbetonbau, n. 62, tome 1

(pp. 7-15) and tome 2 (pp. 37-42), Berlin, 1967.

[02] CEB Comit Euro-International du Bton; CEB/FIP

Manual of Buckling and Instability, The Construction

Press, Lancaster, 1978.

[03] Vasconcelos, A.C.; Origem dos Parmetros de

Estabilidade a e gz. In: Revista IBRACON de
Os valores limites a1 do parmetro de instabilidade, obtidos nos exemplos do presente trabalho e apresentados na
tabela 5, mostram uma variabilidade que vai de um mnimo de
0,514 no exemplo 2 at um mximo de 0,764 nos exemplos
7 e 8. A proporo entre estes valores extremos de aproximadamente 1,5:1. Considerando que a obteno dos mesmos
7. Agradecimento
Ao Prof. Eng. Mrio Franco, pelo fornecimento de material valioso

realizao deste trabalho.
135

Estruturas, n. 20, pp. 56-60, So Paulo, 1998.

[04] Sussekind, J.C.; Curso de Concreto, vol. 2, Porto

Alegre, Ed. Globo, 1984.

[05] Franco, M.; Problemas de Estabilidade nos Edifcios

de Concreto Armado. In Colquio IBRACON sobre

Estabilidade Global das Estruturas de Concreto

Armado, So Paulo, 1985.

[06] Franco, M.; Global and Local Instability of Concrete

Tall Buildings, In: International Symposium for Shell

and Spatial Structures, Proceedings, vol. 2,

pp. 1327-36, Milan, 1995.

[07] Franco, M. and Vasconcelos, A.C.; Practical

Assessment of Second Order Effects in Tall Buildings.

In: Colquio do CEB-FIP Model Code 1990,

pp. 307-24, Rio de Janeiro, 1991.

[08] ABNT Associao Brasileira de Normas Tcnicas;

NBR 6118 Projeto de Estruturas de Concreto

Procedimento, Rio de Janeiro, 2007.

[09] Stamato, M.C.; Associao Contnua de Painis de

Contraventamento (Publicao 157), So Carlos,

EESC/USP, 1972.

[10] Kantorovich, L.V. and Krylov, V.I.; Approximate

Methods of Higher Analysis, New York, Interscience

Publishers Inc., 1964.

[11] Pinto, R.S. and Ramalho, M.A.; Inrcia equivalente

das estruturas de contraventamento de edifcios em

concreto armado. In: Cadernos de Engenharia de

Estruturas, So Carlos, v. 9, n. 38, p. 107-136, 2007.

[12] Schueler, W.; High-rise Building Structures, New

York, John Wiley & Sons, 1977.

[13] Stafford Smith, B. and Coull, A.; Tall Building

Structures: Analysis and Design, New York, John

Wiley & Sons Inc., 1991.
136
Volume 5, Number 1
February 2012
ISSN: 1983-4195

A. D. de Figueiredo, A. de la Fuente, A. Aguado, C. Molins

A. de la Fuente, A. D. de Figueiredo, A. Aguado, C. Molins
12
26
38
The influence of low temperature on the evolution of concrete
strength
68
Reliability of buildings in service limit state for maximum
horizontal displacements
A. G. B. CORELHANO,M. R. S. CORRA and A. T. BECK
84
A variable limit for the instability parameter of wall-frame
or core-frame bracing structures
R. J. ELLWANGER
104

Ibracon Structures and Materials Journal: February, 2012 ISSN 1983-4195 Volume 5, Number 1

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Ibracon Structures and Materials Journal: February, 2012 ISSN 1983-4195 Volume 5, Number 1

Hochgeladen von

Copyright:

Verfügbare Formate

REVISTA IBRACON DE ESTRUTURAS E MATERIAIS

IBRACON STRUCTURES AND MATERIALS JOURNAL

Burj El Arab, Dubai

R. Julieta do Espirito Santo Pinheiro, 68

Steel fiber reinforced concrete pipes. Part 1:

Steel fibre reinforced concrete pipes. Part 2: Numerical

Aims and Scope

works and products under development. The instructions to submit a

instituies de pesquisa, pesquisadores e profissionais que desejem

Acesso via Internet

Volume 5, Number 1 (February, 2012) p. 1-11 ISSN 1983-4195

Steel fiber reinforced concrete pipes. Part 1:

Figure 1 Front and side views of the configuration adopted

IBRACON Structures and Materials Journal 2012 vol. 5 n 1

A. D. de Figueiredo | A. de la Fuente | A. Aguado | C. Molins | P. J. Chama Neto

Figure 2 Configurations adopted for the measurement of displacements

Figure 3 LVDT positioned at the upper part

IBRACON Structures and Materials Journal 2012 vol. 5 n 1

Figure 4. Rebars set up used in SBRCP production that served as

Figure 4 Rebars set up used in SBRCP

Table 1 Material and its consumption

Fine aggregate I (Natural river sand)

* A different brand of cement was used for each series of pipes.

The main steel reinforcement (transverse reinforcement) was made

3. Results and Analysis

IBRACON Structures and Materials Journal 2012 vol. 5 n 1

A. D. de Figueiredo | A. de la Fuente | A. Aguado | C. Molins | P. J. Chama Neto

At the spigot and socket

Fixed on 141 liters/m

Figure 6 Geometric characterization

Figure 5 General appearance of a

Figure 7 Load versus displacement curves

IBRACON Structures and Materials Journal 2012 vol. 5 n 1

Figure 8 Load versus displacement curves

In order to facilitate the analysis of the influence of the setups for

Figure 10 Load versus displacement curves

Figure 9 Load versus displacement curves

Figure 11 Load versus displacement curves

hardening pattern, which consists of increasing resisted load with

IBRACON Structures and Materials Journal 2012 vol. 5 n 1

A. D. de Figueiredo | A. de la Fuente | A. Aguado | C. Molins | P. J. Chama Neto

Figure 12 Load versus displacement curves

Figure 14 Load versus displacement curves

Figure 13 Load versus displacement curves

Figure 15 Average load versus diametrical

of the pipe. The difference observed in the pattern of the curves

IBRACON Structures and Materials Journal 2012 vol. 5 n 1

Figure 16 Average load versus diametrical

ening behavior of these pipes. In this particular case, the drop in

Figure 17 Comparison between the average

IBRACON Structures and Materials Journal 2012 vol. 5 n 1

A. D. de Figueiredo | A. de la Fuente | A. Aguado | C. Molins | P. J. Chama Neto

F1.2mm y F3mm (kN)

a situation which took place when load F reached average values

Figure 18 Correlations between ultimate

IBRACON Structures and Materials Journal 2012 vol. 5 n 1

Figure 19 Correlations between fiber

The performance demonstrated by the FRCP was higher than

The authors of this paper wish to express their appreciation for

VIOLAS, V., AGUADO, A., JOSA, A. Evaluacin de

IBRACON Structures and Materials Journal 2012 vol. 5 n 1

A. D. de Figueiredo | A. de la Fuente | A. Aguado | C. Molins | P. J. Chama Neto

shop on Fiber Reinforced Concrete and Other