Sie sind auf Seite 1von 14

CS-ASA: A NEW COMPUTATIONAL TOOL FOR ADVANCED NONLINEAR

ANALYSIS OF STEEL FRAMES

Andra Regina Dias da Silva

D.Sc., Professor, Graduate Program in Civil Engineering (PROPEC), Department of Civil

Engineering/School of Mines, UFOP, Ouro Preto, MG, Brazil

e-mail: andreadiassilva@yahoo.com.br

caro Machado Prado

M.Sc., Graduate Program in Civil Engineering (PROPEC), Department of Civil Engineering/School

of Mines, UFOP, Ouro Preto, MG, Brazil

e-mail: icaroprado@gmail.com

Ricardo Azoubel da Mota Silveira

D.Sc., Professor, Graduate Program in Civil Engineering (PROPEC), Department of Civil

Engineering/School of Mines, UFOP, Ouro Preto, MG, Brazil

e-mail: ricardo@em.ufop.br
ABSTRACT

A new computational tool for the advanced static and dynamic analyses of steel framed structures

based on the Finite Element Method has been developed and is presented herein. The factors that

influence the structural system behavior are included in the numerical models and formulations

implemented in the CA-ASA computational tool. Among these factors are those resulting from

larger displacements (second-order effects), and the effects introduced when considering semi-rigid

connections and material inelasticity. The influence of geometrical imperfections and residual

stresses are also considered. An evaluation of the static and dynamic stability of various structural

systems, with rigid and semi-rigid connections, whose analytical or numerical solutions are found in

literature, is herein presented. This latter evaluation is used for validating the numerical

methodologies and formulations developed by the authors and implemented in the computational

system proposed.

Keywords: CS-ASA, Advanced analysis, Nonlinear static and dynamic analysis, Geometric

nonlinearity, Material nonlinearity, CS-ASA Preprocessor.


1. INTRODUCTION

With the development of the civil, naval, oceanic and aeronautical industries, and the appearance of

more resistant materials and new constructive techniques, in diverse problems of the structural

mechanics, the nonlinear behavior of the systems begins to be relevant and should be considered in

the analyses. As a consequence, engineering interest in designing lighter and more slender

structures grows and the final structural system product becomes more competitive. Moreover,

these lighter and more slender structures usually cannot be analyzed and designed without their

dynamic effects also being considered. The unknown levels and characteristics of the dynamic

response can lead to system failure during the cyclic loading application due to the accumulation of

structural damages.

Therefore, with the changes of structural conceptions and requirements of more necessary

verification of the behavior of the structures, nowadays there is an increasing search for

computational systems with resources for nonlinear static and dynamic analyses. However, even

though the computational tools for nonlinear models evolve quickly, available commercial

programs with all these resources are still rarities [24]. Among these, should be mentioned the

computational programs Ansys [11], Abaqus [8] and SAP [16].

A new computational system for advanced structural analysis (CS-ASA) is presented in this

article. The main characteristic of the program is the accomplishment of the nonlinear static and

dynamic analyses of steel plane framed structures. In these analyses, the geometric nonlinearity, or

second-order effects, can be simulated; as well as those introduced by considering semi-rigid

connections, and steel inelasticity. The influence of initial geometric imperfections (out of

plumbness and out of straightness) and residual stresses can be also considered. The introduction of

all these nonlinear effects in the numerical models and formulations defines a Structural Advanced

Analysis. This methodology for the analysis and design of steel structures makes it possible to

establish the stability and strength limit of the structural system in a direct manner, without the

necessity of separate verification of each members capacity [7]. Related studies have been
developed with the objective of examining steel structure nonlinear behavior, and the following

works can be cited: Ziemian [26], Liew [9], Barsan and Chioren [3], Chan and Chui [5], Maleck

[10], Sekulovic and Nefovska [18, 19] and Thai and Kim [23]. Structural systems with known

numerical or analytical solutions are herein analyzed to validate the nonlinear formulations adopted

and to demonstrate the applicability of the computational system proposed.

2. THE CS-ASA COMPUTATIONAL SYSTEM

Developed in Fortran 95/2003 language [12] and based on Finite Element Method (FEM), the

CS-ASA performs, as already mentioned, the nonlinear static and dynamic analysis of steel framed

structures. The three steps of the numerical simulation process (pre-processing, analysis and post-

processing) are illustrated in Figure 1. Pre-processing consists in the numerical model definition

(data files input). The data related with the model is then processed to obtain the structural results

(nodal forces, nodal displacements, among others) that will be printed in text files (post-

processing). These files are exemplified in Figure 2. The nonlinear effects simulated are also shown

in Figure 1.

Recently, Prado [14] developed an interactive graphics preprocessing tool, named CS-ASA

Preprocessor, whose primary purpose is to assist the analyst in using the CS-ASA program.

Besides the structural modeling (finite element mesh, boundary conditions, applied loading, semi-

rigid joints, among others), information regarding the type of analysis as well as the solution

strategy can be generated by preprocessor developed. As result of structural modeling and analysis

definition, data files are produced, here called neutral files, which can be used by CS-ASA users as

input information. All the necessary information to perform a structural modeling will be recorded

in the FILEIN1.d file shown in Figure 2, through the use of CS-ASA Preprocessor (Figure 3a). The

data to perform a nonlinear static analysis, or even a vibration analysis of a pre-loaded structure, are

recorded in the FILEIN2.d file, and Figure 3b shows that the dialogues present in CS-ASA

Preprocessor can help the analyst in the generation of this second data file. The FILEIN3.d input

file is necessary when a transient dynamic analysis is performed. Besides the nonlinear FE
formulation, other related data on this file are: time step number, time increment and iterative

process tolerance. The nodal displacement, velocity and acceleration initial conditions, the viscous

damping and the dynamic excitation function shall also be indicated. As the two previous files, this

third input file can also be generated using the CS-ASA Preprocessor (Figure 3c).

CS-ASA
Computational System for Advanced Structural Analysis

Steel Framed Structures

Data Input

STATIC ANALYSIS DYNAMIC

Geometric nonlinearity Geometric nonlinearity


Connection flexibility Connection flexibility
Results
Member material yielding Member material yielding

Figure 1: CS-ASA program: analysis and nonlinear effects considered [21]

Data Input (CS-ASA Preprocessor)

FILEIN1.D FILEIN2.D FILEIN3.D

ANALYSIS

Result Exit

FILEOUT1.S FILEOUT2.DAT FILEOUT3.LOG

Figure 2: Data input and results exit [21]


(a) FILEIN1.d tool boxes

(b) FILEIN2.d tool box

(c) FILEIN3.d tool box

Figure 3: CS-ASA Preprocessor [14]


This interactive graphical environment was developed using the C programming language

[15]. The IUP interface system and the CD two dimensional graphics library resources are also used

for the computational environment creation (see: http://www.tecgraf.puc-rio.br/iup and

http://www.tecgraf.puc-rio.br/cd).

Looking to provide different analysis options of the structural system modeling, several

beam-column element formulations were adopted. The formulations implemented in CS-ASA for

the simulation of the second-order, connections and inelasticity effects, referenced in Figure 1, are

shown in Table 1. The researchers related with these formulations are also indicated in Table 1. The

SOF formulations (second-order formulations) consider the effects of geometric nonlinearities.

These effects appear due to geometrical modification of the structure during the body deformation

process. Large displacements and rotations, but small deformations are supposed. The formulations

SOF-1 and SOF-2 are based on the Bernoulli-Eulers theory of beams and follow the updated

Lagrangian approach. The SOF-3 formulation is derived from Timoshenkos theory, enabling

second-order analysis of thick structures, and the total Lagrangian reference is used.

Table 1: Formulations implemented in CS-ASA

Nonlinear effects Finite element formulations


SOF-1 SOF-2 SOF-3
Geometric nonlinearity
(Alves [2]) (Yang and Kuo [25]) (Pacoste and Eriksson [13])
SRF-1 SRF-2 SRF-3
Connection flexibility
(Chan and Chui [5]) (Chen and Lui [6]) (Sekulovic and Salatic [20])
Material yielding PHF-1 (Liew [9]) PHF-2 (Chan and Chui [5])
Combined effects AAF-1 (Liew [9]) AAF-2 (Chan and Chui [5])

Unlike the criteria currently adopted in structural design, the joints between members do not

behave as in idealized conditions: perfectly rigid or ideally pinned. The connections used in current

practice present semi-rigid behavior that can substantially influence the structural system stability

and force distribution. Here, the semi-rigid connection is modeled as a spring element linking the

structural members. The connection elements are physically attached to the ends of the beam-

column element, maintaining the equilibrium and compatibility conditions, and the joint degrees of
freedom are embedded in the beam- column stiffness tangent relationship. The springs presence

introduces relative rotations in nodes of that elements ends by modifying the equations that

describe the nonlinear behavior of the structural system. Three different procedures are used to

modify such equations that consider the effects of the connections flexibility upon programming

the SRF formulations (semi-rigid formulations).

The steel inelasticity process changes the internal force distribution in the profile cross section

and in all structure. Here, the study of the frame inelastic behavior is performed through the PHF

formulations (plastic-hinge formulations), which follow the basis of the refined plastic-hinge

method. This method monitors the process of cross section early yielding until its total

plastification, and residual tensions can also be considered. The plasticity effects associated with the

formation of plastic-hinge are captured by a strength-reducing dimensionless parameter. This

parameter changes with the level of loading by tracking the plastic region progress on the element

ends. In PHF-1 formulation, this parameter is defined from the plastic strength surface provided by

AISC [1]. In PHF-2 formulation, which follows the BS5950 [4] standard requirements, this

parameter is evaluated from the internal forces and geometrical characteristics of the section.

Finally, the AAF formulations (advanced analysis formulations), which were obtained by

modifying the PHF formulations, consider the influence of the connection flexibility in the

structures second-order inelastic analysis. Details on all these formulations and about the nonlinear

solution methodologies adopted in static and dynamic analyses are in [21].

The CS-ASA still does not have an integrated graphical post-processor for a quick visualization

of structural response. The generated program results are then stored in files of different extensions

for subsequent treatment. These files are exemplified in Figure 2. The equilibrium paths, resultant

internal forces diagrams, degree of member section plastification, modes and frequencies of

vibration, connection and section hysteretic behavior, displacements, velocities and accelerations

time histories, and element internal forces are some of the results generated by the program.
3. SOME MODELING USING THE CS-ASA

This section shows some applications made with the CS-ASA program. Initially, Figure 4a

illustrates the static analysis carried out with a classic structural system known as Lee Frame [17].

By its sensitivity to second-order effects, this slender frame is commonly used to validate the

nonlinear solution strategies and the geometric nonlinear finite element formulations. The

equilibrium paths, with load and displacement limit points, were obtained by controlling the

horizontal and vertical displacements (u and v) of the point of load application. The complete

curves were achieved using an efficient load increment and iteration strategies coupled to the

Newton-Raphson method. Several of these strategies are implemented in CS-ASA and are also

described in [21].

The influence of static preloading on dynamic stability of a sinoidal arch is shown in Figure

4b. The geometric and connection nonlinearities are considered in this study. Nonlinear relationship

between the distributed load P and the archs lowest natural vibration frequency was verified for

different support conditions represented by the Sc parameter, which is defined as a function of arch

bending stiffness EI. The increase in support flexibility, decreasing the structural systems stiffness,

changes a lot the dynamic characteristics.

Figures 4c and 4d show the nonlinear static and dynamic analyses results of a two-story frame

with semi-rigid beam-column connections. The nonlinear behavior of these joints is simulated

through the moment-rotation curves that can be described by mathematical models, i.e.:

exponential, modified exponential, power with three or four parameters and multi-linear. Four types

of semi-rigid connections (named A, B, C and D) were considered. The initial stiffness of these

connections (Scini) is given in Figure 4c. Rigid connections (R) were also adopted. Geometric

imperfections were induced through the small intensity horizontal forces applied on the top of

floors. The static analysis, considering the material with elastic behavior, investigated the loading

capacity and stability of the structure. The load-top displacement curves show that the limit loads

are smaller for more flexible joints. In dynamic analysis, the frame with beam-column connections
of the type C is subjected to cyclic loading, and the history of the structure top horizontal

displacement is presented in Figure 4d. See that initially the displacement oscillates with variable

amplitudes characterizing the transient regime and after approximately 6s, enters into permanent

regime, which is perceived by the constant movement ranging around 5 cm. During the period of

constant response, the energy introduced into the structure by external load excitation and the

dissipated energy by the semi-rigid joints is approximately equal. This can be verified by looking at

the hysteretic cycle of joint L in Figure 4c.

Figures 4e and 4f are respectively related to static analysis where the effect of steel

inelasticity is considered, and to dynamic analysis performed with a seven-story frame with rigid

and semi-rigid joints of the types E and F. Again, see that the limit load factor increases with

connection flexibility reduction. For the frame with rigid connections, the influence of the beams

and columns plastic moment capacity in the structural behavior is evident. For the frame with semi-

rigid connection type F with nonlinear moment-rotation behavior, it is noticed that the connection

flexibility has considerable influence on the overall structural system behavior, since the connection

failure moment (101.7 kNm) is significantly less than the beams plastic moment (307 kNm).

Figure 4f shows the influence of hysteretic damping promoted by semi-rigid connections on the

frame amplitudes. No viscous damping was considered in this analysis.

Recently, a routine able to simulate the nonlinear cyclic behaviour of the material was

implemented in the CS-ASA system. Basically, this model is a natural extension of those already

implemented in CS-ASA for the static case. The methods of Newmark (integration process) and

Newton-Raphson (iteration process) are used for solving the nonlinear equations of motion. Some

examples considering the nonlinear inelastic time-history response of planar steel frames can be

found in [22]
6 120
P A1 : P h = 0.345; A 2 : P h = 0.079
0.2 L Present work Sc = 0
B1 : P h = 0.392; B2 : P h = 0.194
Present work 100 Sc = EI/L
C1 : P h = 0.521; C2 : P h = 0.474

(First natural frequency)2 (x10-3) (rad/s)2


5 u
Schweizerhof
v and Wriggers [17] 80 Sc = 5 EI/L
4 Youngs modulus: 720
Area: 6
L 60 Sc = 10 EI/L
Inertia: 2
b
Cross
3 section
h u Sc = 20 EI/L
40
Load P

L = 120
2 20
Sc
v A2 B2 A1 C2 C1
0
1 B1
P (N/mm)
-20
20 mm Sc h Sc
0
-40 Young's modulus: Cross section h = 20 mm
205 MPa
3 L = 1000 mm
Density: 78 kg/m
-1 -60
0 10 20 30 40 50 60 70 80 90 100 0.0 0.2 0.4 0.6 0.8 1.0 1.2
Displacements u, v P/h (MPa)
(a) Lee frame: static analysis (b) Shallow arch: free vibration analysis
15 24
A: Single web angle connections (Scini = 5424 kNm/rad) 80 Hysteresis M-c loops

Moment M (kNm)
14 (connection L)
B: Top and seat angle connections (Scini = 10759.8 kNm/rad)
13 20 40
C: Flush end plate connections (Scini = 12430 kNm/rad)
12 D: Extended end plate connections (Scini = 34804 kNm/rad) F(kN)
16 0
11 R: Rigid connections 7.5
P P
Displacement u (cm)

10 t(s) -40
F u 12 -7.5
Load P (MN)

9
R Sc W360x72 Sc 1s 1s
W310x143

8 -80
D 8 -8 -6 -4 -2 0 2 4 6 8
3.66 m
7 P P Rotation c (x10-3 rad)

6 Sc Sc
2F
4
L W360x72
5 C
W310x143

Young's modulus:
4 200 GPa 3.66 m 0
B
3 F = 0.001P
A
2 -4
Present work 6.10 m
1
Chan and Chui [5]
0 -8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 0 2 4 6 8 10 12 14 16
Displacement u (cm) Time (s)
(c) Two-story frame: static analysis (d) Two-story frame: dynamic analysis
1.4 100
Present work: linear connections Rigid connections
1.3
Present work: nonlinear connections Linear connections type F
1.2 Sekulovic and Nefovska [19] 80
50.0 kN/m Nonlinear connections type F
1.1 F1 u
F2(kN)
1.0 2 67.5 kN/m 60
F2 24
Displacement u (m)

0.9
Load factor ()

2 67.5 kN/m t(s)


0.8
F2
40 1s
F F 2 67.5 kN/m
0.7 F2
7 x4 = 28 m

R
0.6 1 67.5 kN/m 20
E F2
0.5 Young's modulus: 205 GPa
1 67.5 kN/m
0.4 Residual stress: 235 MPa F
2 0
F2 = 2 F1 = 24 kN
0.3 1 67.5 kN/m
R: rigid connections F2
0.2 Semi-rigid connections type E: Sc 1: IPB600 S c
Scini = 200 MNm rad; Mu = 450 kNm -20
1
2: IPB450
0.1 Semi-rigid connections type F: Beams: IPE400
Scini = 12.43 MNm rad; Mu = 101.7 kNm 8m
0.0 -40
0 10 20 30 40 50 60 70 80 90 100 110 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Displacement u (cm) Time (s)
(e) Seven-story frame: inelastic static analysis (f) Seven-story frame: elastic dynamic analysis
Figure 4: CS-ASA structural applications.
4. FINAL REMARKS

A computational program developed for nonlinear static and dynamic analyses of steel framed

structures has just been presented. The numerical results obtained with this software demonstrate

the validity of formulations implemented in CS-ASA. Accuracy and efficiency in nonlinear

behavior assessment of the steel structures were also verified. The graphical tool developed, the CS-

ASA Preprocessor, with its friendly interface, assisted the papers authors in the structural models

editing/visualization, and made the numerical study more efficient. It should be mentioned that a

graphical post-processor for the CS-ASA is now being developed and implemented.

Acknowledgments

The authors express their gratitude to CAPES, CNPq and FAPEMIG for the financial support

received in the development of this research. Special thanks go to Harriet Reis for her editorial

reviews.

REFERENCES

[1] AISC. Manual of Steel Construction, Load and Resistance Factor Design Specification for

Structural Steel Buildings. American Institute of Steel Construction, AISC, 1st edn., Chicago,

IL, 2003.

[2] R.V. Alves, Instabilidade no-linear elstica de estruturas reticuladas espaciais. Rio de

Janeiro: PEC COPPE/UFRJ, 1995. (Ph.D. Dissertation).

[3] G.M. Barsan and C.G. Chiorean, Computer program for large deflection elastoplastic analysis

of semi-rigid steel frameworks. Computers & Structures, 72(6): p. 669-711, 1999.

[4] BS5950: Structural Use of Steelwork in Buildings. Part 1. British Standards Institution,

London, England, 1990.

[5] S.L. Chan and P.P.T. Chui, Nonlinear Static and Cyclic Analysis of Steel Frames with Semi-

Rigid Connections. Elsevier, Kidlington, UK, 2000.


[6] W.F. Chen and E.M. Lui, Stability Design of Steel Frames. CRC Press, Boca Raton, Flrida,

1991.

[7] W.F. Chen and S. Toma, Advanced Analysis of Steel Frames. CRC Press, Boca Raton,

Florida, 1994.

[8] J. Fish and T. Belytschko, A first course in finite elements, Wiley, 2007.

[9] J.Y.R. Liew, Advanced Analysis for Frame Design, Purdue University, West Lafayette, 1992.

(Ph.D. Dissertation).

[10] A.E. Maleck, Second-order inelastic and modified elastic analysis and design evaluation of

planar steel frames, Atlanta, GA: School of Civil and Environmental Engineering, Georgia

Institute of Technology, 2001. (Ph.D. Dissertation).

[11] E. Madenci and I. Guven, The Finite Element Method in Engineering Using Ansys, Springer,

2005.

[12] M. Metcalf, J. Reid, and M. Cohen, Fortran 95/2003 Explained. Oxford University Press,

USA, 3rd edn, 2004.

[13] C. Pacoste and A. Eriksson, Beam elements in instability problems. Computer Methods in

Applied Mechanics and Engineering, 144: p. 163-197, 1997.

[14] I.M. Prado, CS-ASA Preprocessor: programa grfico interativo de pr-processamento para

anlise avanada de estruturas, Ouro Preto: Programa de Ps-graduao em Engenharia

Civil, DECIV/Escola de Minas/UFOP, 2012. (Masters Thesis).

[15] J.L.M. Rangel Netto, R.F.G. Cerqueira, and W. Celes Filho, Introduo a Estrutura de

Dados: com Tcnicas de Programao em C. Editora Campus, Rio de Janeiro, 2004.

[16] SAP 2000 Advanced Integrated Software for Structural Analysis & Design, Version 10.0.1,

Technical Reference Manual, Computers & Structures Inc., Berkeley, 2005

[17] K.H. Schweizerhof and P. Wriggers, Consistent linearization for path following methods in

nonlinear FE analysis. Computer Methods in Applied Mechanics and Engineering, 59: p. 269-

279, 1986.
[18] M. Sekulovic and M. Nefovska-Danilovic, Static inelastic analysis of steel frames with

flexible connections. Journal of Theoretical and Applied Mechanics, 31(2): p. 101-134, 2004.

[19] M. Sekulovic and M. Nefovska-Danilovic, Contribution to transient analysis of inelastic steel

frames with semi-rigid connections. Engineering Structures, 30: p. 976989, 2008.

[20] M. Sekulovic and R. Salatic, Nonlinear analysis of frames with flexible connections.

Computers & Structures, 79(11), p. 1097-1107, 2001.

[21] A.R.D. Silva, Sistema computacional para a anlise avanada esttica e dinmica de

estruturas metlicas, Ouro Preto: Programa de Ps-graduao em Engenharia Civil,

DECIV/Escola de Minas/UFOP, 2009. (Ph.D. Dissertation).

[22] A.R.D. Silva, W.L. Fernandes, R.A.M. Silveira, and P.B. Gonalves, Elastoplastic analysis of

plane frames under dynamics loading. In: Oate E. and Owen D.R.J., editors. XI International

Conference on Computational Plasticity: Fundamentals and Applications (COMPLAS XI),

Barcelona, Espanha, p. 1-12, 2011.

[23] H.T. Thai and S.E. Kim, Pratical advanced analysis software for nonlinear inelastic analysis

of space steel structures. Advances in Engineering Software, 40: p. 786-797, 2009.

[24] M.B. Wong, Plastic Analysis and Design of Steel Structures. Elsevier, 2009.

[25] Y.B. Yang and S.B. Kuo, Theory & Analysis of Nonlinear Framed Structures. Prentice Hall,

1994.

[26] R.D. Ziemian, Advanced methods of inelastic analysis in the limit states design of steel

structures, Ithaca, NY: School of Civil and Environmental Engineering, Cornell University,

1990. (Ph.D. Dissertation).

Das könnte Ihnen auch gefallen