You are on page 1of 178

SCOTT

Concrete Society Technical Report No. 43 TR.043 Published 1994 ISBN 0 946691 45 2 Published by The Concrete Society No. 3, Eatongate

Slough SL1 2JA Further copies may be obtained from Publication Sales, The Concrete Soaety
Q The Concrete Society 1994

All rights reserved, except as permitted under current legislation. No part of this work may be photocopied, stored in a retrieval system, published, performed in public, adapted, broadcast, transmitted, recorded or reproduced in any form or by any means, without the prior permission of copyright owner. Enquiries should be addressed to The Concrete Society.
Although The Concrete Society (limited by guarantee) does its best to ensure that any advice, reco-mmendations or information ii may give either in this publication or elsewhere is accurate, no liability or responsibility of any kind (including liability for negligence) howsoever and from whatsoever cause arising, is accepted in this respect by the Society, its servants or agents.

CONCRETE SOCXET'II TECHNICAL REPORT

POST-TENSIONED CONCRETE FLOORS DESIGN HANDBOOK

This Technical Report was prepared by a Working Party of the Society's Design Group which is one of the specialist technical groups within The Technical Development Cen:re.

Members of the Working Party


M E Raiis (Convenor) MA. MSc, DIC. PhD, CEng, MICE, MIStructE
G A Bell BSc, CEng, MICE

ROW& Benaim and Asmiates CCL Syslems Ltd Swift Smcrures Ltd Ove Arup and Partners

PW

Matthew

BE, MSc, MIE AusL

R T Whitde MA(Cdnmb), CEn& MICE

ACKNOWLEDGEMENT
During the drafting of this report the working party received a large number of helpful commenrrc from members of the industry. Assistance in preparation of the report by the following members of Arup Research and Development is gratefully acknowledged: Kate Benton, Ian Feltham, Jonathan F ~ n c hGeoff , Lavender, Paula Youngs.

CONTENTS

1. 1.1

Intmduction Advantages of post-tensioned floors

1 2

Sltuctural types considered

14 .
1.5

Bonded ar unbonded tendons Analytical techniques

2.1 2.2 2.3

Effecrs of pascess One-way and two-way spanning flmrs Flexure in one-way flmrs

2.4 . ..
. .. ,. .
,

- Flexure in fIat slab (hvo-way r p n g )


:

2.4.1
.

Fkt slab critehi2

2.5

~.

.shear

3.1 33
,

Column layout

:,.

... . .
,

'

: l o i h i c b and types For


Effect of reshaint to floor shortening Materials Concrete Tenendons

...

...,.

.. .

..?. :

. 33
'

4.
4.1
4.2

4.2.1
4.22 4.2.3

Smd Tendon prowtion Anchorages

Un-tensioned reinforcement

Cover requirements
The design process

1nh.oduction Design flow chart Basic analysis S m h u a l layout Loading Equivalent frame analysis Tendon profile and balanced load Prestress forces and losses Secondary effecks Flexural section design 6.10.1 6.102 6.103 6.10.4 6.10.5 6.10.6 Semceabiity Limit State sfter all losses Transfer condition Ultimate Limit State Progressive collapse Designed flexural un-tensioned reinforcement Minimum un-tensioned reinforcement

Shear seength 6.11.1 6.11.2 Beams and one-way spanning slabs Flat slabs (punching shear)

6.11.3 Openings in slabs


Anchorage bursting reinfoicement Reinforcement ktween tendon anchorages Deflection and vibration Lightweight aggrega!e concrete

Detailing Tendon distribution Tendon spacing Tendon notation Tendon supports Layout of un-tensioned reinforcement

7.5.1 At columns 7.5.2 7.5.3


Shear reinforcement At and bemeen anchorages

Penetrations and openings in flwa Conswction derails

7.7.1 Extent of pours 7.7.2 Conswction joints 7.7.3 Protection of anchorages 7.74 7.75 7.7.6
Back-propping Stressing procedure Soffit d n g

Demolition General

8.1.1 E.1.2

Structures wilh banded tendons SLmchues wilh unbonded tendons

9.

References Appendices

Appendix A:

Design examples
A.1
A.2
Solid flat slab wilh unbonded tendons One-way spanning floor wilh bonded and unbonded tendons

Appendix E:

Calculation of prestiis losses

Appendix C: Appendix D: Appendix E: Appndix F: Apziidix G : Appendix B :

CaIcuMon of tendon geometry Calculation of secondary effects using equivalent loads Calculation and detailing of anchorage bursting reinforcement Simplified Shear Check: Derivation of Figures 17 and 18 V1"mlionof post-tensioned concrete floors Advenisemen!s

Typical spanldepth ratios for a variety of section types for multi-span flwrs. Allowable average stresses in flat slabs (two-way spanning). analysed using the equivalent frame method. Tolerances on tendon positioning.

21 40

49

Grosvenor Square Car Park - Southampton New Oxford Street Typical fa slabs l Typical one-way spanning floors Wst-tensioned flat slab Post-tensioned ribbed slab Post-tensioned waffle slab Bending moment surfaces and tendon diagrams for different tendon arrangements Applied load bending moments in a solid flat slab Distribution of applied load bending moments across the width of a panel in a solid flat slab Load balancing with prestress tendons for regular column layouts Tendons geometrically banded in each direction Tendons fully banded in one direction and uniformly distributed in other directions Load balancing with prestressing tendons for irregular column layouts Preliminary selection of flwr thickness for multispan floors Preliminary shear check for slab thickness at internal column Ultimate shear check for flat slab at face of internal column Restraint l flwr shortening o Layout of unbonded tendons Layout of bonded tendons A typical anchorage for an unbonded tendon A typical anchorage for a bonded tendon Design flow chart Elastic load distribution effects Idealised tendon profile Idealised tendon profile for two spans with single cantilever Idealised tendon protile for two spans with point load Load 'dumping' at 'peaks' Practical representation of idealised lendon profile Resullant balancing forces

Hror Exchange Tower abu

1 2 2 5 6 6 6 7 9 10 10 11 12 13 14 17 18
19

23 25 26 26 27 30 32 33 35 35 35 36 36

Presnessed elemenr as part of a statically determinate shucture Reactions on a prestressed element due to secondary effects Zones of inelasticity required for failure of a continuous member Section stresses used for the calculation of un-tensioned reinforcement Bursting stresses in rectangular beam subjected to an axial symeaic fotre Bursting smss dishibution Method of notation for use on tendon layour dmwings Fiat slab fendon and support layout detailing Flat slab reinforcement layout Reinforcement arrangement a a column L Prefabricated shear reinforcement Unseessed areas between tendons requiring reinforcement Unbonded tendons diverted around an opening Intermediate anchor at a consmction joint Inffil smp for jack access Strand aimming using a disc cutter Strand aimming using purpose-made hydraulic shears Anchorages for unbonded tendons: fixed to formwork Grease-filled plastic cap t protect strand and o wedge grips Anchorage blcck sealed with mortar Stressing banded lendons at slab edges Soffit marking used to indicate tendon position

NOTATION
Area of concrete Area of prestressing tendons Area of un-tensioned reinforcement Area of s h w reinforcement Drape of tendon Width of section Breadth of member; or for T-, I- and L-beams the breadth of the n i Effective depth of tension reinforcement or tendons Depth t the cenaoid of the compression zone o Shon-term modulus of elasticity of concrete Modulus of elasticity of concrete at time of wnsfer Modulus of elasticity of prestressing tendons Modulus of elasticity of un-tensioned reinforcement Eccentricity Compressive stress in concrete at extreme fibre used to calculate serviceability un-tensioned reinforcement requisemen~ Concrete suenglh at (initial) transfer (in N/mm? Stress in concrete at the level of the tendon due to initial p r e s m s and dead load (in N/mm3 Tensile stress in concrete at extreme fibre used lo calculate serviceability untensioned reinforcement requirements Characteristic concrete cube strength (in N/mmq Tensile stress in tendons at (benm) failure (in N/mml) Effective presuess (in tendon) after d l losses (in N/mrnz) Characteristic strength of prestressing steel (in N/mm2) Maximum design principal tensile stress (=0.24Jf,. in N/mm2) Characteristic strength of bonded un-tensioned reinforcement (in N/mm3 Overall depth of section Second moment of area Effective span length Length of tendon Length of tendon affected by wedge draw-in Moment due to applied loads Moment necessary to pmduce zero stress in the concrete at the extreme tension fibre Secondary moment due to prestress Ultimate resistance moment Resuess force Slope of prestress force profile Characteristic strength of tendon (in 3 ) N Prestressing force in the tendon at the jacking end Prestressing force at distance x from jack Distance between points of contra-flexure in tendon Length of a critical shear perimeter Shear force due IO ultimate loads Ultimate shear resistance of concrete Ultimate shear resistance of a section uncracked in flexure Ultimate shear resistance of a section cracked in flexure Design effective shear force Design shear suess at cross-section Design concrete shear strength Uniformly dishiiuted load Neutral axis depth Hl the side of the prestress end block af Half the side of the prestress anchor loaded area Top section modulus

2 ,

a
a ' A

P 9 Y r 7 ,
Iii

Bottom section modulus Angle change in tendon fmm anchor to point considered (radians) Average angle change in lendon per unit length (radianslmeue) Wedge dmw-in Coefficient of fiction Creep coefficient P r i l safety factor for load ata P r i l safety factor for material strength ata Prestressed tendon 'wobble' factor (radians/metre)

INTRODUCTION
The use ofpost-tensioned concrets flwrs in buildings has been consistently growing in recent years. The greatest use of his type of consmcuon has b e n in the USA, and in California it is the primary choice for concrete floors. Post-wsioned floors have also k n used in Australia. Hong Kong, Singapore and Europe. Their use in the UK is now increasing rapidly. Typical applications have been: Offices Car parks Shopping cenms Hospitals ApamenE Indusmal buildings These are illustraed in Figures 1, 2 and 3.

Figure 1 Harbour Exchange Tower. :

Figure t Grosvenor Square Car Park. Southamptoo. :

Figure 3 New Oxford Street. : The Concrete Society has published lhree repons on this subject, Technical Report No 8"'. The Design of Post-Tensioned Concrete Flat Slabs in Buildings; Technical ReponNo 17'". Flat Slabs inPost-Tensioned Concrete with Particular Regard to theuse ofunbonded Tendons Design Recommendationu, Technical Report No 25"', Post-Tensioned Flat-Slab Design Handbwk TR17 was a revision of TR8, and TR25 amplied the recommendations of TR17. The purpose of the current repon is to update the information contained in TR17 and TR25 in line with BSgllO, 1985"', to combine these two repons into one document and to expand same of the recommendations in line with cunent practice.

This report explains the overall concept of post-tensioned concrete floor consmction as well as giving detailed design recommendations. The intention is to simpliiy the tasks of the designer and contractor enabling them to produce effective and economic structures. Post-tensioned flwrs are not complex. The lechniques, structural behaviour and design are simple and very similar to reinforced concrete smctures. The prestress tendons provide a suspension system within the slab and the simple arguments of the triangle of forces apply with the vertical component df the tendon force carrying part of the dead and live loading and the horizontal component reducing tensile stresses in the concrete. Two design examples are given in Appendix A. The report is intended to be read in conjunction with BS8110"'. Those areas not covered in BS8110'" are descrited in detail in the report with reierences given as appropriate. Tne principles laid out in the report may also be applied

to designs in accordance with E m o d e ECZ'fl, but some of the details will need to be modified. Two other Concrete Society publications give useful background information to designers of post-tensioned floors. They are: Technical Report No. 21"'. Durability of Tendons in Prestressed Concrete and Technical Report No. Z3"I. P r i l Prestressing. ata It should be noted that since the integrity of the structm depends on a relatively small number of prestressing tendons and anchorages the effect of workmanship and quality of materials can be critical. This should be understcod by all parties involved in M design and construction.

The main advantages of post-tensioned floors over conventional reinforced concrete in-situ floors, may be summarised as follows: Increased clear spans Thinner slabs Lighter structures Reduced cracking and deflections Reduced storey height Rapid construction Better watertightness
. ' .

;.

These advantages can result in significant savings in overall costs. There are also some situations where the height of the building is limited, in which the :....? . . . . .reduced storey height has allowed additional storeys to be constructed within , : the building envelope.
'

12

'~lruchY0.l opes considered


The repon is primarily concerned with suspended floors. However, the recoinmendations apply equally well'to foundation slabs except that since the -.. loads are generally upward rather than downward the lendon profies and locations of un-tensioned reinforcement are mirrored. The types of floor which can be used range from flat plates to one-way beam and slab structures. An important distinction between structural types is whether hey span one-way or two-ways. This is discussed in greater detail in Section 2.2.

Amount ofpnsbess
The amount of prestress provided is not usually sufficient to prevent tensile stresses occurring in the slab under design load conditions. The structure should therefore be considered to be partially prestressed. The amount of prestress selected affects the un-tensioned reinforcement requirements. The greater the level of prestress, the less reinforcement is likely to be required. Unlike reinforced concrete structures, a range of accepfable designs are possible for a given geomeuy and loading. The optimum solution depends on the relative costs of prestressing and un-tensioned reinforcementand on the ratio of live load to dead load.

Average prestress levels usually vary from 0.7 to 2.5N/mmz for solid slabs and occasionally up to 6N/mm1 for ribbed or waffle slabs. However, when the prestress exceeds approximately 2N/mmz or the flwr is very long, the effects of reshaint to slab shonening by supports may become imponant If the supports are stiff a significant proportion of the prestress force goes into the suppons so that the effective prestressing of the slab is reduced (see Section 3.3).

1.4

Bonded or unbonded tendom


Post-tensioned flwrs can k constructed using either bonded or unbonded tendons. The relative merits of the two techniques are subject to debate. The fouowing points may be made in favour of w h : Bonded: develops higher ultimate flexural s~ength does not depend upon the anchorage after eouting localises the effects of damage

Unbonded: provides greater available lever arm reduces friction losses simplifies prefabrication of tendons grouting not required - can be constructed faster generally cheaper

15

Analytical techniques
The design process is described in Section 6. The analytical lechniques are the same as those used for reinforced concrete structures. The structure is normally subdivided into a series of equivalent frames upon which the analysis is based. These frames can be analysed using moment distribution or other hand techniques, however it is now more common to make use of a plane frame computer analysis program. In addition to standard plane frame programs, there are available a number of programs, specifically written for the design of prestressed structures. These propams reduce the design time but are not essential for the design of post-tensioned floors. For more complicated flat slabs or for those which are be repeated many times, a grillage or finite element analysis of the flwr may be more appropriate.

2.

STRUCTURAL BEHAVIOUR
Effeects of prestress
The primary effects of presuess are a pre-compression of the flwr and an upward load within the span which balances part of the downward dead and live loads. In a reinforced concrete flwr, tensile cracking of the concrete is a necessary accompaniment to k generation of economic stress levels in the reinforcement. In post-tensioned flwrs both the presompression and the upward load in the span act to reduce the tensile stresses in the concrete. However, the level of prestress is not usually enough to prevent all tensile cracking under full design live loading at Serviceability Limit Slate. Under reduced live load much of the cracking will not be visible. The act of presUessing causes the flwr to bend, shorten, deflect and rotate. If any of these effects are restrained, secondary effects of prestress are set up. As slated above, if the level of prestress does not exceed approximately 2N/mm2 the secondary effects due to the restraint to shonening are usually neglected. However, unless the floor can be considered to be statically determinate, the displacements of the flwr sets up secondary moments which cannot be neglected. Secondary effects are diszussed in more detail in Section 6.9 and the calculation of these effects is described in Appendix D.

21 .

There are several different types of post-tensioned floor. Some of the more common layouts are given in Figures 4.5.6.7 and 8. An important distinction between types of floors is whether they are one-way or two-way spanning structures.

Solid flat slab

Solid flat slab with drop panel

Coffered fiat slab Nore:

Coffered flat slab with solid panels

Banded coffered flat slab

Figure 4: Typical flat slabs.


See Section 2.4 for limiling criteria o two-way action. f

Figure 5: Typical one-way spanning floors.

.
Figure 6: Post-tensioned flat slab.

Figure 2 Post-tensioned ribbe slab.

Figure 8: Past-tensioned coffered slab. dne-way flwrs carry the applied loading primarily in one direction and are mated as beams or plane frames. On the other hand. twc-way spanning flwrs have the ability to sustain the applied loading in two directions. However, for a structure to be considered to be two-way spanning it must meet several criteria. These criteria are discussed in Section 2.4.

23

Flexure in one-wayfloors
One-way spanning floors are usually designed Class 3 structures in accordance with BS8110"). Although cracking is allowed, it is assumed that the concrete section is uncradted and that hypothetical ensile stresses can be carried at Serviceability Limit State. The allowable stresses are discussed in Secdon 6.10.1. The behaviour of one-way flwrs at loads less than that which would cause cracking can be assumed to be linear and elastic. BS811OW1 recommends that when the tensile stresses under design permanent loads are less than the allowable stresses for Class 2 suuctms, then the deflection may be predicted using gross section properties, In other cases calculation of deflections should be based on the moment-curvature relationship for cracked sections.

2.4

Flerure inflat slobs (two-way spanning)


In the context of this report, flat slabs are those floors which can carry loads in two different directions to discrete column supports. These are defined as flat slabs in BS811OW.It must be emphasised Lhat these skucues are not the same as two-way slabs in accordance with Section 3.5 of BS8110'". Two-way slabs in BS8110L"alwaysspan on to beams or wal1s.i.e. continuous supporis, and are not considered in this report. One misconception held by some engineers is to consider a reduced load when analysing the slab in one direcuon using the equivalent h m e method. A flat slab supported on columns, rather than perimeter beams. can fail as a one-way mechanism just as a single-way slab, and it should be reinforced to resist the moment From the full load in each onhogonal direction.

Tests and applications have demonsuated thata post-tensioned flat slabbehaves ns a flat plate almost regardless of tendon arrangement (see Figure 9). The effecrs-ofthe tendons-m;of course, critical tothe behaviour as they exert loads on the slab as well as provide reinforcemenr The tendons exert equivalent vertical loads on the slab known as equivalent loads (see Section 6.7). and these loads may be considered like any other dead or live load. Since the tendon effect is opposite to the effect of gravity loads, the net load causing bending is reduced. An additional effect of the tendons is the axial precompression which counteracts flexmal tensile skesses. Therefore, at service dead load, the net downward load musing tending in the slab is normally very low and the flwr is essentially under uniform axial compression. Examination of the distribution.of moments for a flat plate in Egures 10 and 11 reveals that hogging moments across a panel are sharply peaked in the immediate vicinity of the column and that t moment at the column face is b several times the moment midway behveen columns. It should k noted that the permissible snesses given in Table 2 of Section 6.10.1 are average sbesses for the full panel. They are lower than those for one-way floors L allow for this o non-uniform disnibution of moments across L e panel. h

a) Fully banded tendons

b) Uniformly distributed tendons

C)

50%banded plus 50% evenly distributed tendons over full width

Rigure 9: Bending moment surfaces and tendon diagrams for different tendon arrangements

Experimental results

A
Span Column

Span a) Moments along section on column line

-i C

Column

b) Moments halfway between column lines

Figure 10: Applied load bending moments in a solid Rat slab.

Experimental resuitr Equivalent analysis Experimental results

+
i Column z
Panel width a) Moments on column line b) Moments halfway between column lines

Figure PI: DMbution of app!,"!jdIced beadi!zg moments ecross tbe wjdth of a panel in a solid flat slab. In contrast the sagging moments across the slab in mid-span regions m almost uniformly distributed across the panel width as shown in Figure llb.

It is helpful to the understanding of post-tensioned flat slabs to forget the arbihary column strip, middle strip and moment percenlage tables which have long been familiar to the designer of reinforced concrete floors. Instead, the mechanics of the action of the tendons will be examined first

The "load balancing" approach is an even more powerful tool for examining the behaviour of two-way spanning systems than it is for one-way spanning members. By the balanced load approach, attention is focused on the loads exerted on the floor by the tendons, perpendicular to the plane of the floor. As for one-way floors, this typically means a uniform load exerted upward along the major portion of the cenual length of a tendon span, and statically equivalent downward load exerted over the short length of reverse curvature. In order to apply an essentially uniform upward load over the entire floor panel these tendons should be uniformly dishibuted, and thedownward loads from the tendons should react against another smchual elemenr The additional element could be a beam or wall in the case of one-way floors, or columns in a twtiway system. However, a l w k ai a plan view of a flat slab (see Figure 12) reveals that columns provide an upward reaction for only a very small area. Thus, to maintain statical rationality, we must provide, perpendicular to the above tendons, a second set of tendons to provide an upward load to resist the downward load from the fusi set. Remembering that the downward load of the uniformly distributed tendons occurs over a relatively narmw width under the reverse curvatures and that the only available exterior reaction, the column, is also relatively narrow. it becomes obvious that the second set of tendons should be in narrow strips or bands passing over the columns.

Tendons uniformly spaced across floor exerting upward loads in the span and downward loads on the column lines.

Tendons concentrated on column lines exerting upward loads In the span to carry the downward loads of uniformly distributed tendons and downward loads on the columns. Figure 12: Load balancing with prestress tendons for regular column layouts.

There are two ways of accomplishing this two-part tendon system to obrain the nearly uniform upward load we desire for ease of analysis. In the first method. tendons are spaced uniformly in each of two directions and react against banded tendons along the column grid lines in each direction. This resulrs in some of the tendons in each direction being banded over the columns, and some uniformly distributed between these bands (see Figure 13). This method works well where the columns are atianged on a r'ectangular grid.

Figure 9 shows the bending moments derived from grillage analysis of square panels with differing arrangement of tendons. The balanced load provided by the.tendons in ea&-directionis equalro the dead load. Figure 9c gives the most uniform distribution of moments and provides a practical layout of tendons. This arrangement gives 70% of the tendons in the banded zone and the remaining 30% between the bands. It should be noted that, since the width of the banded zone is 0.4 rimes the width of the bay, this arrangement is identical to providing 50% of the tendons evenly distributed over the full width of the bay in addition t 50% concentrated in the band. However, as c n be seen from o a Figure 9 the detailed dismbution is not critical provided that suficienr tendons pass through the column zone to give adaquate protection against punching s h a - and progressive collqse.

Figure 13: Tendons geometrically banded in each direction.

Where the -column mangemern is.irregular. or where openings or other geometric considerations require if a second method may be used. In this method the uniformly distributed tendons and banded tendons may each be placed in one direction only (see Figure 14). The p w e r of the second method becomes very clear by examining afloor which has columns oEimgulx layour. For an example. with reference t Figure 15a. it is assumed that the odd o numbered grid lines are offset one half bay from the even numbered grid lines, but columns on a given letter grid are aligned.

I the 'column skip' approach illustrated in Figure 15b is retained as for conventionally reinforced floors, each span which stmed in a column strip ends in a middle strip, and tracing of load paths, a rational analysis, and proportioning reinforcemenL become difficult if not impossible and force a renun to the basic idea of balancing loads with tendons. The uniformly distributed system of tendons (parallel to letter grid lines) can be accomplished with liule regard for column location. It is only necessary to place the high points of the tendon prome (where reverse curvature and downward load occur) at the inersection of the tendons with the number grid lines. This system then reacts against the banded tendons placed on the number grid lines as shown in Figure 1%. By this pmwdute, the reaction o the gravity load balanced by the f tendons i s carried direcIly to columns, without any flexural action of the flwr. Since this balanced load is typically a large portion of the permanent load on !he flwr, errors in analysis which are due to incorrect assumptions of load path are a function of relatively small loads, and thus are small. The possible consequences of such errors can be investigated by examining the tehaviour of the floor under overloads.

Figure 14: Tendons fully banded in one direction and uniformly distributed in the other direction. Flexural cracking is initiated at column faccs and can occur at load lcvcls in the sewiccability range. Whilc these and early radial cracks remain small, h c y are unlikely to affect h e performance of the slab, Compression due to prestress delays the formation of cracks, but it is less efficient in controlling cracking than un-tensioned reinforcement placed in the top of floors, immcdialcly adjacent lo, and above the column. 2.4.1

Flat slab criteria


For a preskessed floor lo be considcrcd as a flat slab the following criteria apply: Precompression should be applied in two orlhogonal directions: Such a flwr with no, or moderate, crack formation performs as a hdmogeneous elastic plate wilh its inherent two-way bchaviour. The actual tendon location at a given point in a flwr system is not critical t thc f l w i s o two-way behaviour since prccomprcssion, which is the decisive factor, is commonly applied lo lhc flwr at its pcrimcter. The precompression at the edges of the slab is concenuated behind the anchorages, and spreads into the flwr with increasing dislance from the edge. This is uue for floors of uniform thickness as well as flwrs wilh beams in h e direction olprecompression. Flwrs wilh bandcd post-tensioning and flwrs with wide shallow beams also qualify for two-way action at regions away from h e free edges whcrc precompression is allained in bolh directions.

a) Irregular flat plate column layouL

b) Column strips and middle snips


Typical uniformly distributed tendons and loads

c) h a d balancing with bonded tendons

Figure 15: Load balancing with prestressing tendons for irregular column layouts

Past experience shows that for the precompression to be effective it should be at least 0.7 N/mrn2 in each direction. Aspect ratio (length to width) oi any panel should not be greater than 2.0: This applies lo solid flat slabs, supported on orthogonal rows of columns. For aspect ratios greater than 2.0 the middle section will Lend to acl as a one-way spanning slab.

Stiffness ratios in LWO directions: The ratio of the stiEfness of the slab in two orthogonal directions should no1 be disproportionate. This is more likely to occur with non-uniform crosssections such as ribs. For square panels this ratio should not exceed 10.0, otherwise the slab is more likely lo behave as one-way spanning.

2.5

Shear
The method given in BS81101" for calculating shear in beams and one-way spanning slabs should be used. A method for calculating shear for post-tensioned flat slabs is no[ provided in BS8llO. The method given in this manual (see 6.11.2) combines [he prestress effccts givcn in section 4 of BS8llO with L e method given for punching h shear for reinforced concrete in section 3 of BSBllO.

3.

STRUCTURAL FORM

Current experience in many countries indicalcs a minimum span of approximately 7m to make preslressing viable in a floor. However, cxamplcs are known in which prestressed flwrs have been competitive where shortcr spans have been used for architectural reasons, but presuessing was then only made viable by chwsing the right slab form. In general the ideal situation is, of course, to 'think prestressing' from the initial concept of the building and to chwse suitably longer spans. In chwsing column layouts ana spans for a prestressed flwr. several possibilities may be considered la optimise the design, which include: a) Reduce the length of the end spans or, if the architectural considerations permit, inset the columns from the building perimeter to provide small cantilevers. Consequently, end span bending moments will be reduced and a more equable bending moment configuration oblaincd. Reduce, if necessary. the stiffness of the columns lo minimise the prestress lost in overcoming the restraint offered ta flwr shortening (see Section 3.3). Where span lengths vary, adjust the tendon profiles and the number of tendoils to provide the uplift required for each span. Generally this will be a similar percentage of the dead load for each span.

b)

C)

Once the column layout has been determined. the next consideration is the type of flwr to be used. This again is determined by a number of factors such as span lengths, magnitude of loading. architectural form and use of the building, special requirements such as services, location of building, and the cost of malerials available.

F h r thkktzess m d types
The slab thickness must meet two primary functional requirements - structural strength and deflection. Vibration should also k considered where there are only a few panels. The selection of thickness or type (e.g. plate without drops. plate with drops, coffered or waffle, ribbed or even beam and slab) is also influenced by concrete strength and loading. There are likely to be several alternative solutions to the same problem and a preliminary costing exercise may be necessary in order to chwse h e most economical.

The infomadon given in Figures 16, 17 and 18 will assia h e designcr to make a preliminary choice of noor section. Figure 16 (derived from Table 1) gives typical imposed load capacities for a variety of'nal slabs and one-way floors over a range of spaddcpth ratios. These figures are based on past experience. Figure 16 is appropriate for all types of prestressed floor. Figures 17 and 18 are only appropriate for flat slabs but Figure 17 is not appropriate for coffered slabs which do not have a solid section over he column. At this slage it should b noted that the superimposed load used in Figures 16, 17 and c 18 consisls of all loading (dead and live) bar h c self-weigh~of h e section. The calculation methods used for obtaining thc graphs in Figures 17 and 18 are described in Appendix F.

Total

Note: -

Spanldepth ratio

This chart is derived from the emperical values given in Table 1 for multi-span floors. For single-span floors the depth should be increased by approximately 15%. Figure 16: Preliminary selection of flwr thickness for multi-span floors.

Slab thickness adjacent to columns

vc = 0.75 N/mm2 Column size including head = 300 x 300 m m

20 30 40 50 60 10 80 W I W 110 120 130 140 150 160

Area (rn2)
D

K::%49Z$8$:

Slab thickness adjacent

11 10

Total imposed Load (kN/m2)

'
6 5

vc = 0.75 N/rnmz

Column size including head = 500 x 500 rnm

4
3

2 1 0 20 JO 40 50 60 70 80 90 1W 110 120 130 Id0 150 160

Area (rn2)
1 6a .m n- m " m a Tm ~ '2s ,- ~ - - m m m ~
O D D D D D D D

Slab thickness adjacent to columns

iota1 imposed Load (kNlm2)

14 13 12 11 10 9

'
5
4 3

vc = 0.75 N/mm2

Column size including head = 700 x 700 rnm

1
0

24 40 50 60 70 80 90 IM H 120 130 140 150 150 O

Area (mZ)

Figure 17: Preliminary shear check for slab thickness at internal column.

8 E D 8 8 8 ~ ~
16
NNnnprrul

14
13 12 'H
'10

Slab thickness adjacent to C O I U ~ ~ S

Total 9 imposed a Load 7 (kNlm2) 6


8 4 3 2

,f = 40 Nlmrn2 Column (inc. head) 300 x 300 D.L. Factor = 1.4 L.L. Factor = 1.6

1 , , I I I I I I I I I I - I I zo 30 40 60 60 m no so iwnorzo1Joraorarsoren

Area (m2)

Figure 18: Ultimate shear check for flat slab at face of internal column.

Notes:
1. For column sizes other than 300 x 300 the slab depth should be multiplied by the factor (column perimeter 11200). 2. The maximum shear stress for f., = 40 Nlmm' and more is SN1mm2. For f., < 40 Nlmm' the maximum shear stress is 0.8 I & . For f,. = 35 Nlmm' increase slab depth by a factor of 1.06. For f,, = 30 Nlmm'increase slab depth by a factor of 1.14. 3. The value of dlh is assumed to be 0.85. 4. The ratio of V,NIV is assumed to be 1.15. 5. These curves do not takc account of elastic distribution effects. See Section 6.6.

Flat slabs e n d 10 exceed punching shear limits around columns, and often need additional shear rcinforcemcnl at these locations. The ,mphs in Figure 17 provide a prcliminary assessment as LO whcher shcar reinforcement is needed lor the scction types 1, 2, 3. 5 and 6 (all flar slabs) in Table 1. As the shear capacity 01a slab is dependent on the dimensions of the supponing columns or column heads, each graph has been derived using different column dimensions. In addition, h e shear capacity at thc face of the column should be checked. This can be done using h e graph in Figure 18. Thc graph has been dcrived lor slabs with 3W x 300mm supporting columns, and lo obtain the imposed load capacities for slabs w i h olhcr supporting column sizes, the valucs in h e graph should be multiplied by h e ratio of required column pcrimcled1200. should be followed when using Tablc 1, Figures 16, 17 and The following 18 ta obtain a slab section. a) Knowing the span and imposed loading rcquiremcnrs. Figure 16 or Table 1 can be used to choose a suitable spanldepth ratio for the section typc being considercd. Table 1 also provides a simple check for vibration effecs. If section type 1. 2.3, 5, or 6 has bcen choscn, check Lhc shear capacity of the scction. using onc of the graphs in Figurc 17 (depending on what size of column has been dccidcd upon). Obtain thc imposed load capacity lor the chosen slab section. I1 this excccds the imposcd load, then shear rc'inforccmcnt is unlikcly lo be ncccssary. If it docs not, h e n reinforcement will be rcquircd. If Lhc diffcrcncc is very largc. then a n increase in section dcpth or column size should be considcred. Check the shear capacity at the face of the column using the graph in Figure 18. If h e imposcd load capacity is cxceedcd, increasc thc slab dcpth and check again.

b)

C)

It should be noted that Tablc I and Figurc 16 are appliwblc for multi-span floors only. For single-span floors the depth should be increased by approximalely 15%. Figwes 17 and 18 arc applicable lor b o h floor types and havc b a n derived using an avcmge load faclor of 1.5 (see Appendix F . ) Figures 17 and 18 are set for internal columns. They may be used for external columns provided that (he loadcd area is doubled for edge and quadrupled for comer columns. This assumes thar the edge of the slab extends lo a1 least he'cenlre line of h e column.

Table 1: Typical spanldepth ratios for a variety of section types for multi-span floors.

Additional requirements if no vibration check to be carried out for normal office conditions: either the flwr has at least four panels and is at least 250 mm thick or the nwr has at least eighl panels and is at least 200 mm thick. either the flwr has at least four panels and is at least 400 mm Lhick or the flwr has at least eight panels and is at least 300 mm thick

Section type

Total imposed loading (kN/ma)

Spanldepth ratios 6 m 5 L 2 13 m

1. Solid flat slab

25 . 50 . 1. 00

40 36 30
A

2 Solid flat slab with drop panel .


1 0 I, I I I L-----.J

25 . 50 . 1. 00

44 40 34
A

r------I

3 Banded flat slab .


i

slab

beam

25 .

45 40 35

25 22 1 8
A

I
span16
I I

5. O
10.0

4. Coffered flat slab

.-Ji--J , L--JL..
I

2.5 5.0 10.0

25 23 20

-7 r - - 7

11

r - - 7 r1 1 11

5. Coffered flat slab with solid panels

25 . 50 . 1. 00

28
26

I
I L--

23

Table 1. Continued Section type Total imposed loading (kNlm3


J

Spanldepth ratios 6 m 5 L 5 13 rn 28 26 23

rl

6. Coffered slab with band beam

L--JL.

2.5 5.0 10.0

t
7. Ribbed slab

- 7 -

.?---,T'

2.5 5.0
I1

30

27 24

!I
tt

I/
I ,

II

10.0

!I
slab 2.5 5.0
42

8. One-way slab with narrow beam

b m 18

38
34

16 13

I I I I I I
I
I

10.0

1. All panels asumed to be square 2. 'Spanldepth ratios not affected by column head 3. t It may be possible that prestressed tendons will only be required in the banded sections and that untensioned reinforcement will suffice in the ribs. or vice versa. it The values of spanldepth ratio can vary according to the width of the beam.

33

E,iect of restraint to flwr s h o n i n g


Post-tensioned floors must be allowed to shorten to enable the presb-ess to be applied to the floor@s". Shortening occurs because of: a) b)
C)

Elastic shortening due to the prestress force. Creep shortening due to the prestress force. Shrinkage of concrete.

The elastic shortening occurs during stressing of the tendons, but the creep and shrinkage are long-term effects. The flwr will be suppaned on columns or a combination of columns and core walls. Thesesuppons offer a restraint to the shortening of the flwr. There are no frm rules which may be used to detemine when such restraint is significant. As a guide, if the presmss is less than 2NImmz the flwr is not very long and there is not more than one stiff resminr (i.e. a IiB shaft) then the e f f e c ~ restraint are usually ignored. of
A simple method of ascertaining the restraint offered by the suppons is to calculare the elastic, creep and shrinkage smins expected in the slab and then to calculate the

forces required to deflect the supports. Figure 19 shows two simple frames in which the floors have shortened and the columns have been forced to deflect. The force in each column may be calculated from b e amount it has been forced to deflect and its stiffness. The stiffness may b e calculated on the assumption h a t h e column is built-in at both ends.

..,

..

(a) Symmemcal floor supported on columns

@) F l w r supponed by columns and lifi shaft at one end

Figure IF: Restraint to noor shortening.

The calculation of elastic. creep and shrinkage strains may be based on the values given in BS8110''1. The elastic strain should be based on the modulus of elasticity at the time the tendons are stressed. -If this is at seven days after casting the modulus is approximately 80% of the modulus at 28 days. The creep strain depends on the age of the concrete when the tendons are stressed. the humidity and the effective thickness. The c w p strain would be typically 2.5 times the elastic strain. The shrinkage strain will generally be in the range 100 to 300 x 1r6, but in some circumstances it can increase to 400 x 10". Typical strains for a 300mm internal floor with a prestress of 2 N/mmz would be: Elastic smin Creep strain Shrinkage strain Total long-term s h a h ( ) , E

100 x 250 x 300 x 650 x

lo4 10' 10"

lo4

The following analysis is approximate but conservative and ignores any displacement of h e foot of the columns or rotation of the ends of the columns. A more accurate analysis may be made using a plane frame with imposed member strains. The force required to deflect each column. as shown in Figure 19. may b assumed e to be calculated as follows:

For the purposes of calculating H,, the value of EJ,for the column may be reduced by creep in the column and in some cases cracking. A reduction of at least 50% from the short-term elastic properties is normally justifiable. The total tension in the flwr due to the restraint to shortening is the sum of all the column forces to one side of the stationary point In Figure 19a. the tension is H, + H in Figure 19b. the tension is H, + H, + H,. This lension acu; as a reduction in !he ; precompression of the floor by the prestress. I the tension is small in comparison f with the prestress. it may be ignored. If the tension force is moderate. it may be necessary to subtract it from the prestress to obtain the effective precompression of the flwr. But if the resmint is so severe that flexing of the vemcal lrlembers to accommodate the shortening is not possible. other measures are required. These may include freeing the offending stiff elements during a temporary condition. However, it must be remembered Lhat creep and shrinkage will continue to occur for up to 30 years.

4.
4.4

MATERIALS
Concrere Concrete should be mixed, transported and placed in accordance with BSBIIO, Pan 1. Section 614). Choice of concrete type and grade will be influenced by durability. early strength gain requirements, material availability and basic economics. At present concrete grades of C35 and C40 are the most commonly used for post-tensioned floors. Where lightweight aggregates are used, references should be made to Lhe special requirements OF BS8110. Pan 2 Section 5l4). .

42

Tendons Strand The tendon material used for post-tensioning concrete llwrs is normally 7-wire suand. This strand should comply with Type 2 (low relaxation) as descr~hd BS in 5896, Table 6"0'.

4.1

42.2

Tendon protection Unbonded tendons Unbonded tendons are protected by a layer of grease inside a plastic sheath. An example is shown in Figure 20. These materials should comply with the recommendations given in reference 11.

Figure 20:

Layout o unbonded tendons. C

Under normal conditions, the strand is supplied direct from the manufacturer already greased and sheathed. In no circumstances should PVC be used for the plastic sheath, as it is suspected that chloride ions can be released in cerlain conditions.

Bonded tendons
Bonded tendons are placed in metal ducts which can be either circular or oval in om. An example is shown in Figure 21. The latter is used in conjunction with an anchorage which ensures lhar up IO four strands are retained in the same plane in order IO achieve maximum eccentricity.

Figure 2 : Layout of bonded tendons. i The ducts are made from either spirally wound or seam folded galvanised mela1 strip. On completion of smessing, the ducts are pumped full of cement grout which effectively bonds the strand to the structure as well as ensuring corrosion protection. Further information can be obtained from reference 12.

42.3

Anchorages
Anchorage companents should comply with BS 4447'L31. Details of these are shown i n Figures 22 and 23. In the case of unbonded anchorages corrosion protection should comply with Class A exposure as defined in reference 14. In addition, tests for unbonded anchorages should include fatigue testing consisting of cycling the prestressing force between M) and 65% of the characleristic strength of the s m d for two million cycles.

Figure 2 : A typical anchorage for an unbonded tendon. 2

Figure 23: A typical anchorage For a bonded tendon.

43

UE-tensionad reinforcenssn:

Un-tensioned reinforcement shall comply w t BS 1449''5'. ih

5.

COVER REQUIREMENTS
Nominal cover is dependent on durability requirements or fire resistance, whichever condition is the more onerous. Bonded tendons: The cover to the tendons should be in accordance with the requiremens for prestressed concrete in BS8110, P r 1, Clause 4.12.3'4' the cover ai being measured to the outside of the duct. It should be noted Lhat the cover to the centre of the tendon will be more than that to the centre of the duct, since the tendon will press againsr the wall of the duct. Unbonded tendons: There is no durability requirement for unbonded tendons protected in accordance with 4.2.2. Fire pmtec6on shall be provided in accordance with BS8110, P r 1. Clause 4.12.3.1.3") and the nominal cover to the sheath should not at be less than 25mm. The tendon is normally specified as a nominal diameter (e.g. 12.9 or 15.7mm for 7-wire super strand): 3mm should be added to the diameter to allow for the thickness of sheathing. Un-tensioned reinforcement: The cover to the un-tensioned reinforcement should be in accordance with the requirements for reinforced concrete in BS8110, P r 1, Clause at

3.3"'.
Anchorages: The cover to anchorages should be as for bonded tendons given in BS8110, P r 1, Clause 4.12.3.1(4). at Consideration should k given LO the layout of both tendons and un-tensioned reinforcement when deciding the critical cover requirements (see Section 7.5).

6.

THE DESIGN PROCESS

'Ihis section considers the various stages of the design process in more detail. As in most reinforced and prestressed concrete design work, the customary design process is of an iterative nature following the cycle: 1. 2. 3. 4. Preliminary design Check design by analysis Revise design as required Repeat steps 2 and 3 if necessary.

The analysis is normally bzsed on semi-empirical procedures such as the equivalent frame method. More rigomus analyses based on, for example, finite element methods are m d y adopted. They should only be considered for large pmjcts of unusual form where the high design costs and the inapplicability of h e empirical method justify them. ?he design is undertaken generally in accordance with BSgllO'" with additional guidance given in this report Normally the flexural capacity at Serviceability Limit State is considered firs4 and then checks on flexural and shear capaciry at Ultimare Limit State are carried out.

62

Design jlow c i w i
A typical design flow chart is shown in Figure 24.

63

Basic a ~ l y s i s
The analysis of post-tensioned flwr systems differs from a reinforced concrete design approach owing to the positive effect lhat the tendons have on the structure. In reinforced concrete the reinforcement is inidally unstressed: the suess in the reinforcementresults fmm h e deformation and cracking of the structure underapplied load. In this way the reinforcemenr may be considered lo acl passively. On the other hand, the tendons in a post-tensioned flwr are actively stressed by h e jacks so that they are loaded before the application of other loads. The force in the tendon is chosen by the designer and does not vary much with the application of Serviceability Limit State dead and live loads.

- load to be balanced

required presuess forces number of tendons

shears due to externally

Check Serviceability deflections and vibrations

Consider revisions or refinements to design

Figure 24: Design flow chart


30

The analysis of equivalent frames may be undemken by hand, using moment distribution or flexibility methods. or by computer using plane-frame analysis programs. There are also available on the market several computer programs specially written for post-tensioned flooring systems. These programs not only undenake the analysis of the frame under applied loading and loading from the tendons, but also calculate the flexural stresses. For more complex or detailed analysis, grillage or finite element methods may be used. Whichever Whnique is used for the smctural analysis it must cake into account not only the dead and live loads but also the loads which the tendons apply to the structure (see Section 6.7).

The choice of layout and m e m k r sizing has been d i s c u s 4 in Section 3, and is probably the most important decision in the design process. Unless previous experience or ovemding factors dictate the exact form and section, seveml possibilities should be studied, although the designer should be able to limit the possible solutions by considering the various consuaints and by rough design and costing exercises. With regard to slab thickness and concrete suengths, the relationship of structural layout, slab thickness and loading has been referred to in Section 3. A determination of a trial m e m k r depth must b made at an early stage in the calculation process. e This can often be best obtained by assuming a value of about 70% of the equivalent non-prestressed member.

The loading for Serviceability Limit State should consider the dead load and post-tensioning effects acting with those combinations of live loads which result in the maximum stresses. Unless there are specific abnormal loads present, it will generally be sufficient to consider the post-tensioning effects in combination wilh the live loads as given in BS8110, P r 1, Clause 4.3.3"'. at At transfer of presuessing only the dead loads present during stressing, togelher wilh the post-tensioning effecls before losses due to creep, shrinkage and relaxation, should be considered in obtaining stresses. Where the applied loads change significantly during conshuction or phased stressing is employed, the various stages should each b checked for transfer mess limits. e At the Ultimate Limit State the load combinations shown in BS8110, Part 1, Table 2.1 and Clause 4.3.3'" shall be considered to arrive at the maximum moments and shears at any section. Secondary effects of prestressing should be included in the applied loads with a load factor of 1.0 (see Section 6.9).

6.6

Eqdvait-nt frame anaijsis


It is usual lo divide the smchlre inlo subframe elements in each direction. Each frame usually comprises one line of columns together with beamlslab elements of one bay width. The frames chosen for analysis should cover all the element types of the complete smcture. The ends of the columns remote from the sub-frame may generally be assumed to be fixed unless the assumption of a pinned end is clearly more reasonable (e.g. pad footings). The use of the equivalent frame method does not take account of the two-dimensional elastic loaddisnibution .effects automatically. It will give different support reactions from the analyses in the two orthogonal directions unless the width of slab chosen caincides with the paints of zero shear in the o h dimtion. Normally for internal bays the width of slab will be the full panel width. However for a regular layout, Lhe penultimate frame will pick up more than half the width on the side of the end bay (see Figure 25). Provided the reaction on each column is taken as the larger value from the two analyses litrle accuracy will be lost However where the size and arrangement of edge columns is different from the internal columns the width of slab should be estimated more accurately. This will ensure the correct selection of the number of presuess tendons with the profile appropriate for the frame being analysed. It should be noted that these elastic effects are automatically taken into account when the floor is analysed using grillage or fmite element methods. Irrespective of which analytical technique is used, care should be laken to ensure that the assumptions made are appropriate to the structure under consideration. In particular the prestress applied to two adjacent frames should not be very dissimilar athenvise the prestress fmm the more highly stressed frame will dissipate into the adjacent frames.
Lines of zero shear in !ongltudinal direction

@1 .. br mi Freme (a) Equivalent frame widths in transverse direction


End I~enuirfmsre Internal brami

-@
'

Lines of zero shear in trsnrvsrie direction

h
End Frame

Internal Frame

"

d k n d Frame

(b) Equivalent frame widths in longitudinal direction

Figlare 2 5 Elastic load diistributioo ere&

BS8110, P s 1, Section 3.72"' gives a clear definition on the division of a flat slab at into sub-frames or 'panels'. Other methods may also be used. It is now common to analyse smctures using plane frame computer programs. However, when longhand momencdisnibution calculations are employed, sliffness. carry-over factors and fued end moment coefficienrs must te calculated. These can be quite complicated for varying sections, column heads and drop-panels and, although often ignored in hand calculations, the effect on stiffness of the complete beam moment of inertia over the column width can be most significant, particularly for wide columns. It should also be noted that BS8110, Part 1, Section 3726" allows reduction of ...' negative moments U, the column face, which equally applies to post-tensioned members.

67 .

Tendon p m p a d balanced load


Ideally the tendon profile is one which will produce a bending moment diagram of similar shape. but opposite sign. to the moments from the applied loads. This is nor always possible because of varying loading conditions and geomemc limitations (see Section 5). It should te noted that for bonded systems the centroid of the snands will not coincide with the centroid of the duct. This is particularly true in the case of circular ducts. Further information may be available h m the manufacturer's literature. In the simplest case, for a uniformly loaded simply-supported beam, the bending moment is parabolic, as is the ideal tendon profile. The total 'sag' in the parabola is referred to as the tendon 'drape' (see Figure 26). and is limited by the section depth and minimum cover to the tendon. At the suppons the lendon has no eccentricity and hence there is no bending momenc due to the tendon forces. Tendon profiles are not always symmewic. However, the point of maximum drape is still at the centre of the points of inflection, but may not correspond to the point of maximum sag.

. .
.,...

w/unit length

f
I

drape = a

Bending moment Figure 26: Idealised tendon profile.

The upward forces applied to the concrete by a parabolic profiled tendon, as shown in Figure 26. are uniformly distributed along the tendon. At the ends of the tendon downward forces are applied to the concrete by the anchorages. I h e upward and downward forces are in equilibrium sn that no external forces occur. The set of forces applied to the member by the tendon are known as the 'equivalent' or 'balanced' loads, in that the upward forces counter-balance a proportion of the downward forces due to dead and live loads. For a parabolic profile the upward uniformly distributed load, w, can be calculated as follows:

where: s a

paw

distance between points of inflection drape of tendon measured at centre of profile between points of inflection. Note that this may not be position of maximum sag average presEessing force in tendon

Usually, in continuous members, the most effective use of a tendon in producing 'balanced loads' is achieved by having the tendon at its lowest possible point in positive moment locations, and at its highest possible point in negative moment locations. In this way the drape, and consequently the 'balanced loads', is increased to a maximum. The 'equivalent' or 'balanced' loads may be applied to the smctural frame in order to obtain the total effects of prestressing. The total effects are a combination of the Primary and Secondary effects as described in Section 6.9. It is beyond the scope of this publication to give an extensive treatise on prestressing theory or load-balancing design. Funher details may be oblained fmm reference 16.

In post-tensioned design it is common to roughly 'balance' equal proponions of the dead and applied loads in each span. Some designers set out with a preconceived idea of what load they wish to balance as a proponion of the dead or total load. Othen balance the minimum amount which will result in the final stresses due to the out-ofbalance loads k i n g as close as pssible to the maximum allowable stresses.
This latter approach is usually the most economical overall but may not always k the most suitable for deflection or congestion of un-tensioned reinforcement. Figure 27 illusuates an idealised tendon profile for a two-span memkr with a cantilever. The parabolic profiles result in the balanced loads w,, w, and w, as shown, calculated from the tendon profile and hence the 'drapes'.

L,

Span2:
8, WT

L?

-W,

L3

Cantilever Span 1:

4
WJ

h?,=-

L2
~

+ 8,
~ + e ~

Lz2 )8 =

Span 3: e , ~ (

4 '
e ~ ) = ~

Figure 2'1: I d e a l i d tendon profile for two spans with single cantilever. Figure 28 illushates a two-span member with an idealised tendon profile to provide a uniform uplift over span 1 and a concenmted uplift in span 2. The concenmed effect is useful in members transferring column or similar point loads. While te bending moments 'peak' over the supports. it is clear !hat in practice a h tendon cannot do this and some approximation must be made. Remember that the peak is where the tendon is 'dumping' the load it has picked up by its parabolic shape (Figure 29). In practice, tendon profiles are of the form shown in Figure 30. The ratio t / should generally be kept as small as possible and is usually selected 'L as 0.10. Appendix C provides information from which the parabolic tendon geometry can be calculated. The resultant balancing forces are therefore as shown in Figure 31.

Span 1: Total drape = el

I-

L,

2
81

L2

&an 2: Total drape = e,

e2 + -1-

Balanced load w, = P e t total drape x WLI2

Balanced load wz = P x total drape x 4/Lz

P is the prestressing force at ths section under consideration. Note: that the centre of gravity of the concrete and the centre of gravity of the tendon coincide at the end of the member so that no equivalent load moments are applied at the end of the member.

Figure 28: Idealised tendon profile for two spans with point load

Rigure 29: Load 'dumping' at 'peaks'. 35

Figure 3C: Practical represeabiion OF ideaiised tendon p:o:i!e.

Figure 31: Resultant balancing forces. For the reverse parabala at the support the toul load downwards:

and for the span parabala the total load upwards:

If we make L / equal to 0.1, a. suggested above, 'L


then:

= 4s,

Since the upward and downward loads must be equal, it follows that i h = 3 7 s1 s , and hence a, =

a,
4

The balancing loads upwards and downwards due to the tendons can thus be calculated.

Prestmss forces and losses


From the time that a post-tensioning tendon is stressed, to its find state many years alter stressing, various losses take place which reduce the tension irr the tendon. n e s e losses are grouped into two categories, namely:

1.

Short-term Losses, which include: a) Friction losses in ihe tendon b) Wedge set or 'draw-in' c) Elastic shortening of the structure.

These losses take place during stressing and anchoring of the tendon. 2. Lang-term Losses, which include: a) Shrinkage of lhe concrete b) C a p of the concrete under the effect of fhe prestress c) Relaxation of the steel tendon. Although lhese losses occur over a period of up lo ten or more years, the bulk occurs in the Grsr two years following stressing. Typically, losses reduce the applied preslress force by approximately 10% at transfer and 20%after aU losses. The calculation of losses is discussed in more d d l in Appendix B.

69

Secondary effects
The secondary effecrs of presmssing are sometimes called 'parasitic effects' but that implies that the effects are unwanted and harmful. This is not in fact the case. For most structures the secondary moment will be a sagging moment and will increase the moments due to applied loads at midspan but reduce the momenrs at the support. In some structures it is possible lo 'tune' the secondary effecrs by adjusting the shape of the tendon profile to obtain the optimum solution. This is more likely to be of use in the design of beams rather than slabs.

Primary prestressing forces and moments are fhe direct result of the prestress force acting at an eccentricity from the section centroid The primary moment at a section is simply the sum of the producc; of each lendon force with its eccenuicity; h e primary shear is the sum of transverse componenls of the tendon forces and the primary axial load is the sum of rhe axial components of lhe tendon forces.
When an element of a structure is prestressed its s h a p changes. It will always shorten, and will bend if the cenuoid of the prestress force does not coincide at all positions with the section centroid. (It is possible, however, to select a tendon profile which results in no rotation of the elemenl ends.) If the element is part of a statically determinate structure then lhese changes in shape will not affect the distribution of forces and momenls (Figure 32).

G isolated element Stressed = l ?


Unstressed element on supports Unstressed isolated element Stressed element still compatible with supports Figure 32: Restressed element as part of a statically determinate structure But when the element forms part of an indeterminate suucture, the changes in shape resulting from prestressing will modify the support reactions. Additional reactions are required to make h e prestressed member pass through support points and have suitable orientxion where appropriate (Figure 33). 37

Unstressed element in srructure

Reactions applied to

-t
4,

through support positions Reactions applied to make beam have compatible rotations Total secondary forces and moments for element

Unstressed isolated L-_----_l elemarrt Stressed isolated element

-V

- wp -

Figure 33: Reactions on a prestressed element due to secondary effects These secondary reactions result in secondary forses and moments in the members. These are typically consrant axial and shear forces throughout a span and uniformly varying moments. The calculation of these secondary effects can be difficult when staged construction, creep and shrinkage are considered. (Note hat secondary effects cannot develop in cantilevers as they are statically determinate.) Methods of calculating secondary effects are given in Appendix D. Equivalent loads will automatically generare'the primary and secondary effects when applied Lo the structure. Seniceability calculations do not require any separation of the primary and secondary effects, and analysis using the equivalent loads is suaightfonvard. However, at Ultimate Limit State the two effects must be separated because the secondary effects are treated as applied loads. 'The primary prestressing effects are taken into account by including the tendon force in the calculation of the ultimate section capacity. The primary prestressing forces and moments must therefore be subtracted from the equivalent load analysis to give the secondary effects. To calculate the ultimate loading oii an element, rhe secondary forces and moments are combined with the ultimate forces and moments horn dead and live loads. The Handboak to BS8llO'"', suggests that the partial load factor on secondary effects should b 1.0. The total ultimate moments can be redishibuted in accordance w t e ih BS8110. Pan 1, Section 4.2.3'''.

6.10

Flexural section design

6.10.1 Serviceability L m t State afler al! losses ii

The bending moments calculated from the critical loading conditions given in Section 6.5, including the tendon effects, provide the serviceability stresses at each section using: top fibre stress, f, = boltom fibre smss, f, = where: z ,

P+M
A,
2 ,

P - IVI
A, zb

the top section modulus

zb = the bottom section modulus M = the tola1 out-of-balance moment

e = eccentricity of tendons, taken as positive below the neutral axis MA = applied moment due to dead and live loads M, = moment from prestress secondary effects

One-way spanning floors


Bonded tendons: The maximum allowable concrete compressive and tensile stresses for floors with bonded tendons are given in BS8110, Pan 1, Section 4.3.4.2 and 4.3.4.3"' respectively. Most buildings will be satisfactory as Class 3 suuctures and thenature of the loading must be considered when deciding on a 0.1 or 0.2mm crack width (e.g, frequency and duration). Unbonded tendons: The maximum allowable concrete compressive stresses in floors with unbonded tendons are as for floors with bonded tendons and are given in BS8ll0. Part 1, Section 4.3.4.2"'. The maximum concrete tensile stresses should be taken as those given for group (b) in Table 4.2 of the Standard, with a limiting crack width of 0.lmm. These values must be adjusted for section depth as given by Table 4.3 of the Standard. If the suesses are enhanced by increasing the un-tensioned reinforcement as is allowed for bonded rendons in BS8110"'. crack widths and deflections should be rigorously chccked. All concrete tension shall be canied by untensioned reinforcemenr (see Section 6.10.5).

Flat slabs (two-way spanning)


Flat slabs may be analysed in either of two ways. The more common method is to analyse equivalent frames in each direction. In this case some account must be taken of rhe peaking of the moments at the columns, described in Section 2.4. The analysis results in moments and stresses averaged across the width of the panel. The stresses should be limired to those given in Table 2. Grillage or finite element analysis may be used, but this is normally only justified with floors of unusual configuration or where a design is f be constructed many o times, such as in a high-rise building. If such analytical techniques are used which take into account the distribution of moments and suesses across a panel, then the allowable stresses given for one-way spanning floors may be used. Particular care must be taken in modelling the columnlfloorintersection and in the interpolation of the results obtained.

Location

In Compression

In Tension with bonded reinforcement 0.454~0.45dfwithout bonded reinforcement

Support span Table 2: Note:

. -

0.29 kc"

o
0.154~

Allowable average stresses in nat slabs, (hvo-way spanning), analysed using ille equivalent frame method. Bonded reinforcement may be either bonded tendons or un-tensioned reinforcement.

In Table 2, the support zone shall be considered as any part of the span under consideration within 0 2 x L of the suppon, where L is the effective span. Outside of this zone is considered to be the span zone. Additional designed un-tensioned reinforcement is required in the suppon zone ofall flat slabs, and in the span zone oE slabs using unbonded tendons where the tensile stress exceeds 0.15Jf,. The design of this reinforcement is presented in Section 6.10.5.

6.10.2 Transfer condition


Transfer stresses should be checked for all floors. These are likely to be more onerous for floors with high imposed loads. Un-tensioned reinforcement shall be calculated in a similar manner u the , reinforcement for the Serviceability Lirnit'State (see Section 6.10.5).

One-way spanning floors


BS8110, Part 1. Clause 4.3.5.1"' gives suitable limits for one-way floor concrete compressive stresses at transfer of 0.5~"at the extreme fibre (or 0.4f, for near uniform stress distribution) where f, is the concrete strength at transfer. Clause 4.35.2 gives the limits for allowable concrete tensile stresses which, for most buildings, will be 036dfcf,,.

Flat slabs (two-way spanning)


The allowable sh'esses given in Table 2 for the Serviceability Limit State also apply lo the transfer condition for slabs analysed using the equivalent Erame method, o however, h should be substituted by f,. F r slabs analysed by the grillage or finite element methods, the allowable suesses are those given for one-way spanning floors.

6.10.3 Ultimate Lima State


An Ultimate Limit State check is necessary on all floors in addition t the n Serviceability Limit State previously covered. In this condition, the factored dead and applied loads are considered together with the secondary effects of the prestressing (see Section 6.9). The primary prestress effectr are considered as pan of the section strength. Additional un-tensioned reinforcement may be required in order to generate an adequate moment capacity.

BS 8110, P r 1, Section 4.3.7''' gives guidance on the assumptions for calculating the at concrete and un-tensioned reinforcementsuesses and the allowable design stresses for the tendons. In the above Section, equation 52 for unbonded tendons has been developed from the results of tests in which the stress in the tendons and the lenglh of the zone of inelasticity in the concrete were both determined. The floor is considered to develop both elastic and inelastic zones and the length of the inelastic zone is taken to be 10 x the neutral axis depth.
The extension of the concrete at the level of the tendons is assumed to be negligible in the elastic zones and the extension in the inelastic zone is assumed to be mken up uniformly over the length. 1, of the tendon. This is discussed further in references 29 and 30. Hence, for a simply supported flwr there is only one inelastic zone associated with the failure, but with a continuous floor the numbee of inelastic zones required for failure is more complex (see Figure 34). The length of tendon. I, in equation 52 can be modified, bearing in mind that if the tendon does not continue the full length of the continuous floor it may not include all the inelastic zones necessary for failure. It is therefore prudent to assume no more than one inelastic zone per span, and no more than two inelastic zones for the full length.

plastic hinge

Without Columns

With Columns

(a)

Ductile failure

With Columns
@)

BrilUe failure

Figure 34: Zones or inelasticity required for failure of a continuous member.

6.10.4' Progressive collopse

Where p r o ~ v e . c o l l a p e involves the use of unbonded tendons in key elements, the maximum stressin the unbanded tendon shall not exceed 0.85fW This ensures that the anchorages are not over-stressed, and protects against catenary action. In unbonded memters there is also the risk that if tendons are severed accidentally there will be a 'progression' of failure for the full length of the lendons. This is pasticularly relevant for one-way spanning members such as beams. ribs and slabs spanning onto beams or walls.

In the case of one-wa), members where horizontal progressive collapse is of concern, it is necessary to reiiiforcewiihun-tensionedsteel. This should be provided to satisfy the load case of dead load plus one third live load [DL+ (I/3)LL] with an overall load factor of 1.05, and reduced material factor in accordance with BS8110. Pan 1, Clause 2.4.3.2'" for 'effects of exceptional loads or localised damage'. Reinforcement should be in accordance with normal BS8llO limits and arrangements.
Experimental and practical evidence in the USA has established that this problem does not occur in the iniernal bays of flat slabs due to the overall 'plale' or membrane action. The possibility of horizontal progressive collapse of edge and comer panels of flat slabs must be considered These panels should be supponed for the situation where the tendons parallel to the edge have been severed. This support can typically be provided by bonded reinforcement in the panel or an edge beam. 6.10.5 Designedflexural un-tensioned reinforcement Additional un-tensioned reinforcement shall be designed to cater for the full tension force generated by the assumed flexural tensile stresses in the concrete for the Iollowing situations: All locations in one-way spanning floors using unbonded tendons. Ail locations in me-way spanning flwrs where transfer stresses exceed
0.36vfci.

Support zones in all flat slabs. Span zones in flat slabs using unbonded tendons where the tensile smss exceeds 0.15 df,. The reinforcement shall be designed, with reference to Figure 35, to act at a stress of (5/8)fy as follows: h-x = -f,, x h
f-

fa

The value off,, will be negative in tension

The reinforcementshall be designed for the soesses ar Serviceability Limit State, both afler all prestress 1osses.andat. m s f e r conditions. It shall be placed in the tensile zone, as near as practicable to the outer fibre (see Section 7.5). Under mnsfer conditions any designed reinforcement is likely t be on the opposite face to thar o required after all losses.

At Ultimare Limit State, additional un-tensioned reinforcement may also be required (see Section 6.10.3). Any reinforcement provided for the Serviceability Limit State may also be used in the calculation of the moment capacity at Ultimate Limit Stare. The designed reinforcement shall be checked against the minimum requirements given in Section 6.10.6.

Figure 35: Section stresses used for the calc~lntiw of un-tensioned reinforcemenL

6.I0.6 Minimum un-tensioned reinforcement


Where fire ratings of greater than 2 hours are required, it is recommended that antispalling reinforcement be placed in the soffit when no other reinforcement is provided.

One-way spanningfIoors
Bonded tendons: There are no minimum un-tensioned reinforcementrequirement5 for one-way spanning floors with bonded tendons. It is considered that these flwrs have sufficient tendon-to-concrete bond to distribute flexural cracking. Care should be taken to ensure sufficient reinforcement is provided to guard against cracking before stressing, if early phased suessing is not employed. Unbonded tendons: One-way spanning floors with unbonded tendons should have minimum reinforcement in accordance with BSgllO, Part I, Table 3.27, Figures 3.24 and 3.25". This reinforcement should be spread evenly across the full width of slab in accordance with the spacing rules given in BS8110, Pan 1. Section 3.12.11'4'.

Flat slabs (two-way spanning)


A 1 flat slabs shall have minimum un-tensioned reinforcement at column positions to 1 dismbute cracking. The cross-sectional area of such reinforcement shall be at least 0.075% of the gross concrete cross-section (O.WO75 x AJ, and shall be concentrated between lines that are 1.5 times the slab depth either side OF the widlh of the column. The reinforcement shall be placed as near as practical to the top of the nwr, with due regard for cover and tendon location, and shall exrend at least 0.2 x L into the span or as far as necessary by calculation (see Section 6.10.1 and 6.10.2). The maximum pitch of the reinforcement should be 3Wmm.

In the span zone, there are no minimum requirements. However, when unbonded tendons are used it would normally be necessary to provide designed un-tensioned reinforcement in the hotlam of the slab (see Section 6.10.1). This reinforcement should extend at leas1 to within a distance of 0 2 x L, measured from the centre of the support It should be placed at a spacing of 3 x slab thickness or 500mm, whichever is the lesser.

Slab edges
Un-tensioned reinforcement should be placed along edges of all slabs. This should include U-bars laced with at least two longitudinal bars top and bottom, as shown in Figure 38. See also Section 6.12. Reinforcement should k provided in the triangular unstressed area between anchorages. See Section 6.13.

6.33

Shew strength

6.11.1 Benms and one-wny spanning slabs


The method in BS8110, Part 1. Section 4'" should be used. Where unbonded tendons are used, the value of v, in equation 55 of BS8110'41should be reduced by a factor of 0.9 as recommended by Regan"".

6.112 Flat slabs (punching shear)


BS8110idJdoes not provide specific guidance for checking punching shear for prestressed flat slabs. The working party considered a number of-different methods while preparing this handbook, with a view IO satisfying the following aims:

Design capacities to be in line with other international standards. Increased punching shenr capacity for bonded tendons. Increased punching shear capacity when tendons are concenhated in the vicinity of the column. A design method which complements BS 8110'4' as far as possible. A design method which allows a smooth transition from reinforced concrete to prestressed concrete and allows for situations where the slab is prestressed in one direction only.

The following method achieves these aims and is recommended. Calculate !he effective shear force, V,,, in accordance with BSBIIO, Clause 3.7.6. The shear resistance, V. is obtained by adding together Lhe contributions from each of the sides of the critical shear perimeter as given in BS8110, Clause 3.7.7"'. The shear resislance of each side of the critical perimeter should be calculated in accordance with BS8110, Clauses 4.3.8 and 4.4'4J as modified below.

Fiat slabs are generally not heavily prestressed and will therefore be governed by the design for "sections cracked in flexure", using equation 55 (BS8110, clause 4.3.8.5"3. Equation 55 does not, however, provide a smooth transition from reinforced to prestressed concrete because of the term:

For lighrly presuessed structures the inclusion of this term in equation 55 can lead to a shear capacity less than that which would be calculated for the same slab but without prestress. This is obviously incorrect The British Cement Association @I1 has recently compared various forms of shear calculation with published test results and concluded that equation 55 would be more consistent with the test results if the above term were omitted. It is thaefore recommended that the shear resistance of each side of the critical perimeter be calculated from equation 55 moditied'as follows:

where v,, b. and d are fhe values for the relevant side of the critical perimeter. The value of v, should be calculated taking into account both A, and A,, for bonded tendons in accordance with BS8110 Clause 4.3.8.1. However the presence of unbonded tendons should be neglected in this calculation of v,. No further reduction is considered necessary (e.g. as suggested in reference 17, page 98). The de-compression moment, M,,, should be calculated for the width of the side of the critical perimeter under consideration. It should be noted that the axial effects of prestress, P/A,, are uniformly distributed over the width of the slab whereas the prestress moment effects (P,+ M,) are concentrated at the lccation of the tendons at the critical perimeter. Hence the two contributions to M, have to k calculated separately as follows (for a hogging moment region):

where: 0.8 P = a safety factor on presuess (BS8110, Clause 4.3.8"1


= =

the tntal prestress force, over the full panel width, after all losses the concrete section area across the full panel width

A,
'

= section modulus for the top fibre over the width of the side of the critical perimeter = = the total prestress force for all tendons passing through the side of the critical perimeter the eccenhcity of the prestress force, P', at the critical perimeter, measured positive below the centroid

P '
e'

The value of V/M must be calculated for the load case under consideration, normally that which generates the largest. , V/h4 should strictly be calculated at the location V of the critical perimeter but may be calculated conservatively at the column centreline. For a typical in~ernal column V/M will vary from 5.5& to 6.0/L, depending on h e ratio of dead load to live load, where L is the span length.

6.11.3 Openings in slabs


Tendons should be continuous and displaced horizontally to avoid small openings. If tendons are terminated at the edges of large openings, such as at stairwells, an analysis should be made to ensure sufficient strength and proper behaviour. Edges amund openings may be reinforced similarly to conventionally reinforced slabs: in the case of large openings, supplementary post-tensioning tendons may be used to strengthen the edges around openings.

6.12

Anchorage bursting reinforcement


Reinforcement is usually required to resist the tensile suesses caused by the concenlrauon of the forces applied at the anchors. At some distance from the edge of the floor (or the anchorages) it can be assumed Ihat the distribution of stresses is the classic linear distribution and depends only on the magnitude and position of the resultant of the forces applied to the edge of the flwr. Betwen the edge and the above plane the lines of force are curved and give rise to transverse tensile suesses in both directions perpendicular to the applied force direction. Figures 36 and 37, adapted from reference 18, illusbate the varying proportions of the presoessing force manifesting itself as a splitting tensile force of magnitude depending on the anchorage and floor relative geometries.
bursting stress

I-

Anchoraoe Zone

---

Figure 3 6 Bursting stresses in rectangular beam subjected to an axial symmetric force.

mrnprerslve

F i p r e 3 : Bursting strtresc dstrib@tioe. 7


46

Where a group of anchorages exist, as is often the case for 'banded' slab tendons. the bursting stress zones for both the individual and collective anchorages should be considered, and reinforcement placed accordingly. Care should also be taken to ensure that the phasing of the application of prestress to anchorage groups does not create a bursting condition which may be critical. If this condition is unavoidable, reinforcement should be added accordingly. BS8110, Part 1. Section 4.11"' gives design bursting tensile forces of a similar nature to Figure 35 and limits the steel stress to 2M)N/mm1 at Serviceability L i i t State. It is suggested that ban with f, = 4 6 0 ~ / m mare used for this reinforcement ~ Alternatively the bursting forces and distribution may b calculated in a more rigorous e method, such as suggested by GuyonCm'. some cases it may be shown that the In concrete is capable of withstanding bursting without the addition of reinforcement. At Ultimate Limit State for unbonded tendons only, reinforcement requirements should be checked in accordance with BS8110, Clause 4.11.3. This Ultimate Limit State check is unlikely to be governing. Where anchorages are grouped, or where the distribution of anchorages does not reflect the distribution of concrete in the cross-section, it may be necessary L include o 'equilibrium' reinforcement to prevent splitting between anchorages. Also when anchorages occur within the plan area of the floor rather than at the perimeter, it may be necessary u, include 'following' reinforcement This reinforcement runs parallel to the tendon past the anchorage to limit cracking adjacent to the anchorage. These effects are discussed in CIRIA Guide No. 1"". Post-tensioning system suppliers often test their anchorage systems in concrete prisms. reinforced in a similar manner t that encountered in practice and using a orism size o similar to the common on-site member thichess, ek. Such tests may be deemed under BS8110, Part 1. Section 2.6'" to satisfactorily model the on-site conditions and the reinforcement may be considered adequate provided suitable safety factors are observedu4'. Two examples showing the calculation of, and the detailing of, bursting reinforcement are given in Appendix E.

6-13

Reinforcement between lendon anchomges


Figure 43 shows an area of slab between tendon anchorages which require reinforcement to span the unstressed zones. Any presuessed tendons which pass through this zone, parallel to the slab edge, may be included with the relevant rcinforccmcnt, provided it is in the local tension zone. Thc area of lension reinforcement (ancl/orpprestressed tendons) provided parallel to the slab edgc should resist bending momenls fmm the ultimate vertical loads calculaled for a continuous slab spanning I,. This reinforcement should be evenly disnibuled across a width equal to 0.71,. and should be continuous along the edge. The area of reinforcement placed perpendicular to the slab edge should be the greater of 0.13%bh, or a quarter of the reinforcement provided parallel lo the edge. It should be placed evenly between anchorages, and extend the greater of I, or 0.7L plus a full anchorage length into Ule slab.

Deflection
This is a Serviceability Limit State relating to the complete structure. The deflections of a structure, or of any pans of a structure, should not adversely affect appearance or performance. The final calculated deflection (including the effects of tempmure, creep and shrinkage, and camber), measured below the line between the supports of the flwr and roof, should not in general exceed s p d 5 0 . In addition.where -internal partitions. cladding and finishes can be affected by deflection. the deflections should be limited in accordance with BS8110, Clause
3.2.1.2(4'.

As a guide for a prestressed solid slab, continuous over two or more spans in each direction, ihe span/depth ratio should not generally exceed 42 for floors and 48 for m f s . These limits may be increased to 48 and 52 respectively, if &tailed calculations show acceptable behaviour with regard to short- and long-term deflections, camber and vibration. Lower span/depth ratios will often apply to slabs with high liveldead load ratio& The span/depth ratios for waffle slabs should not generally exceed 35.

Vibration
Prestressed flwrs are usually thinner or span funher than unpresuessed floors. They herefore lend to have lower natural frequencies and greater consideration must be given to their dynamic performance. The Steel Construction Institute has published a design guide on the vibration of floorsm'. This guide covers sources of vibration excitation in buildings. human reaction to vibration, evaluation of natural frequencies, response of floors and design procedures. Although it was written primarily for checking the acceptability of lightweight concrete composite floors on steel beams, most of the guide is relevant for any floor system. Appendix G gives a procedure, based on the guide for checking presnessed flwrs with a rectangular grid. Vibration should not k a problem for general office buildings if the total slab depth is greater or equal to the values given in Table 1. For more sensitive locations, or for slabs shallower than the above criteria. an aswsment of the dynamic response of the floors should be made.

6.15

Lightweight aggregate concrete


Additional considerations on the use of lightweight aggregate concrete are given in BS8110, Part 2, Section 5'4', and the Guide to the Structural Use of Lightweight Aggregate Concretem1. The allowable tensile stresses given in Sections 6.10.1 and 2 should be reduced by a factor of 0.8: however, the allowable compressive stresses need not be reduced. In BS8110, Part 2, Clause 5.4"], the maximum allowable shear smss is given as Lhe lesser of 0.63Jf- or 4N/mm2. These values are 0.X times the values given for normal weight concrete. However. this limit is related lo compressive strut failure, not to tensile failure. In the view of h e Working P r y h e Limitations for normal weight at concrete, the lesser of 0.8dfWand M/mm2, may also be used for lightweight concrete.

(Reference should also be made lo 'Standard method of delailing slruclurai concrete'*")

7.4

Tendon &tributiQn
Various methods for dishibuting the tendons can be used. These are discussed in Section 2.4. In situations of ovaload, tendons passing through the colurnn/floor intersection are more effective than lendons elsewhere. It is therefore recommended that a minimum of two tendons should pass through this section. For ribbed slabs or beams, the dishibution of tendons is dictated by the spacing of members.

7.2

Tendon s@g
The maximum spacing of uniformly distributed tendons should not exceed six times the slab depth for unbonded tendons or eight times the slab depth for bonded tendons. Unbonded tendons may be placed in groups if required. It is recommended that grouped tendons are laid side by side and do not exceed four lendons per group. The minimum horizontal distance between ducts or groups of tendons shouid be the greater of 75mm or the gmuplduct width. Should it be necessary to arrange the tendons in vertical layers in beams or ribs, then it is recommended that the gap between the layers should be at least the venical dimension of the lendon or duct In the case of bonded tendons where oval metal ducts are used, it is recommended that their positions are staggered to ease the placing of concrete. If tolerances on tendon pasitions are not staled, the values in Table 3 should be adopted.

Slab thickness h c 200mm h>Wmm

Tolerances Vertically f lu +l Horizontally

* 20mm
5 20mm

t 5mm

Table 3: Tolerances on tendon positioning

73

Tendon notation
The accepted standard notation or tendons on drawings is shown in Figure 38. It is recommended that this legend Figure is included on all tendon layout drawings.

Tendon quantiry, length, colour.code, elongsrlon

Placing sequence lwhen requiredl


7Dimenrionr

from reference line to centreline of rendon group

One strand
T w o rtrandr

Three rtrandr
Four strands
Y

Five raandr

1x10 6 7 laddedl

2.40

+ Oead end
Add rrrandr

RedA = 75

3x24 50 ithrul
Blue A = 1 6 5

Lf

Will be marked with one of ,he above rymbolr

Note: When more than one symbol appears on a tendon group the number o f rtrandr equal the rum o f the symbol derignarionr.

Edgeof slab

Figure 38: Method of notation for use on tendon layout drawings. Figure 39a shows an example using the legend showing groups of tendons and anchorages types, together with the tendon sequence, detailed. This Figure is based upon reference 24 modified along Lines recommended in this document. Tendon profiies in the longitudinal and transverse directions are shown using an exaggerated scale for the vertical dimensions. These are usually given from the soffit of the slab to the centreline of the ducvsheath and are plotted at intervals of lm. Closer centres may be necessary for sharp vertical curves. For ease of placement on site, shop drawings are detailed giving the vertical tendon position from soffit m underside of tendon.

.me profile of the tendons is critical to the floor performance. It is therefore recommended that the support centres do not exceed lm. For ribbed stabs or beams, support bars can be adequately held by f wire ties. Spot welding can be used but m this makes any adjustment difficult. Figure 39b shows a typical support bar layout. The actual layout may be modified by the contractor depending on the support system adopted, so that the specified tendon profiles are attained and adequate suppon is provided.

Placing sequence not shown

Section A.A (a) Flat slab tendon layout

Note: 1. Height given is from soffit of slab to underside 2. Diameter of support bar is 10mm.
@) Typical tendon pmfie and support bar layout for a flat slab

'tendon

Figure 39: Flat slab tendon and support layout detailing.

ET12

-.14 12 - 81 +

Floor plan

Section A.A showing reinforcement details Figure 4 : Flat slab reinforcement layout 0
75

Layout of un-tcmswned reinforcemni


Figure 40 shows an example of the reinforcement that is always required at edges and in the top of flat slabs at columns. It also shows the reinforcement needed in the boaom of the slab at midspan for some design applications. S e Section 6.10 for details.

7.5.1

At columns
Reinforcement should be placed in the Lop of the slab over columns. The design of such reinforcement is described in Section 6.10.5 with minimum requirements given in Section 6.10.6. Figure 41 shows a typical arrangement of tendons and un-tensioned reinforcement amund a column.

Figure 41: Reinforcement arrangement a t a column. Shear reinforcement Shear reinforcement in flat slabs. if required, is usually in the form of links or hairpins, although prefabricated shear reinforcementis available. Fabricated steel shear heads may also be used. See Figures 41 and'42 and Section 6.11 for details.

Figure 42: Prefabricated shear reinforcement.

At a n d between anchorages
An adequate amount of reinforcement should be placed at anchorage end blocks to avoid splitting of the concrete. A sample calculation m determine the amount of this reinforcement is given in Appendix F.
Reinforcement should be provided in the 45" wedge area between the anchorages (Figure 43).

Figure 43: Unstressed areas between tendons requiring reinforcement,

7.6

Pem~atiom openings in frwm ond


Unbonded tendons may be diverted around the openings as they are relatively flexible (see Figure 44). The change of direction of the tendon should occur away from the opening, and trimmer bars should be provided to avoid possible cracking at the comers.

Figure 44: Unbonded tendons diverted around an opening. The oval sheathing used in bonded tendons is very rigid in the transverse direction, and cannot be bent around openings. In this instance, openings should t confined to e the areas between tendons. h e cutting of penemions in finished slabs is not a problem in ribbed slabs where the tendon positions are, in effect, defied. Gmuted tendons, providing the gmut is effective, can be cut without significant loss of prestress. However, when unbonded tendons have been used, care must be taken to lotate the tendons before concrete removal. Tendons can be cut and reinstated but it is recommended that this work be carried out by a specialisr

7.7

Construction d@IaiLr

7.7.1

Extent of pours
With bonded tendons, friction losses usually restrict fhe length of single end stressed tendons to 25m, and double end stressed to 5Om. The lower friction values for unbonded tendons extend these values to 35m and 70m respectively. Longer lengths are achievable but the friction losses should be carefully considered. These limitations usually determine the extent of pours. Prestressing tendons may be continuous thmugh construction joints allowing larger areas without any permanent joints. Allowances should be made in accordance with good practice to accommodate temperature variations by the provision of expansion joints on larger slabs.

7.7.2

Consmction joints
Generally consmction joints should be made in the vicinity of quaner and third points of the span from supports.

Shear provision in accordance with good practice should be made by the introduction of expanded mesh, by roughening the previously poured surface or by the introduction of a shear key.
In long slabs, intermediate anchorages may be introduced which allow the stressing to be continuous through the construction joint (see Figure 45). Alternatively infill skips can be used, but it should be noted that fhese will nor be prestressed. These suips are cast after the stressing of the adjacent sections is complete (see Figure 46). This operation should be delayed for as long a period as is reasonable to reduce the effects of creep and shrinkage.

Figure 45: Intermediate anchor at a construction joint.

55

Figure 46: Infill @rip for jack access. In assessing the movement of slabs at expansion or confraction joints from the time of polning concrete, a strain of 650 x 10"should be. considered as normal. The drying out effect of air conditioning can increase this to 1WO x lo4.
lo the direction of stressing should be avoided as this The use of dowels .could prevent the stress being transferred to the slab. Generally dowels should be avoided in slabs saessed in two directions.

7.7.3

Protection qf anchomges
Tendons are normally anchored within the middle third of the slab to ensure adequate edge cover to the anchorage. Pocket fonners at anchorages should be large enough lo allow adequate trimming of the tendons after stressing, thus ensuring good end cover to the sirand. Trimming should be. carried out using a disc cutler or hydraulic shears (see Figures 47 and 48). In no circumstances should the tendon be trimmed by flame cutting. Pocket formers are normally proprietary plastic or polystyreneinits which make up part of the anchorage fuings. Anchorages f i e d lo formwork are shown in Figure 49. It is recommended for unbonded tendons that, after trimming the strands, the wedges and the s m d end are coated with grease of similar specification to that used in the tendon and that a watertight cap be applied over the coated area (see Figure 50). The minimum end cover to !his cap should be 25 mm.

Figure 4 2 Strand trimming using a disc cutter.

Figure 48: Strand trimming using purpose-made hydraulic shears.

Figure 49: Anchorages for unhonded tendons: fued to formwork.

57

Figure 50: Grease-filled plastic cap to protect strand and wedge grips. and The pockets for anchorage are particularly vulnerable to ingress of mois~ure it is therefore essential that they be properly filled with a non-shrink m o m as s w n as possible afler snessing is complele (see Figure 51). Before installing the pocket m o m , the concrete surfaces should be coaled with a suitable bonding agent. In no circumstances should the monar conlain chlorides or other malerials which could be harmful to the presaessing steel.

Figure 51: Anchorage block sealed with mortar.

7.7.4

Back-propping
Back-propping may be required to ensure that the consvuction loads can be safely carried by the earlier consauction stages, and this must be considercd by the designer in a similar manner lo normal reinforced concrete construction.

7.7.5

Stressing procedure
The stressing forces, extensions and sequence should be specified on the drawings. This has to be planned in such a way that the prestress is applied as uniformly as possible, and that no overloading of the formwork occurs. The banded [endons are usually stressed f i t to ensure this is the case (see Figure 52). Wherever possible the use of different forces for tendons of the same size should be avoided.

Figure 52: Stresing banded tendons at slab edges. In members where early stressing is desired to reduce the risk of early shrinkage cracking, it is common to stress the tendons in two stages. The first stage is usually 50% of the final prestress force. and is carried out as soon as the concrete has obtained adequate strength for the anchorage being used. This concrete strength is usudly between 12 and 15 N/mmz.

7.7.6

SofJit marking
Tendon positions in flat slabs ate not apparent on completion of concreting. Recent practice has been to introduce soffit marking, where the cover to the tendon is less than the penetration of ceiling and service fittings. An iiluslration of typical marking is shown in Figure 53. Unpainted zones indicate no tendons. Dark zones indicate tendons near the soffit and white zones indicate tendons near the top of the slab.

Figure 53: Sofit marking used to indicate tendon position.

Special precautions are required for the demolition of presaessed concrete structures, and it is recommended that the advice of a presbessing specialist is obtained before planning the demolition. Two references giving useful information are the FIP guide to good practiceN' and the P T publication on the resent demolition of a past-tensioned I

8.1.1

Structures with bonded tendons


These can normally be demolished using recognised methods of demolition. However, it is of fundamend importance that, during the initial staEes, it is ascertained that the grouring is effective.

8.1.2

Structures with unbonded tendons


The energy introduced into the tendons in this instance is only secured by the anchorages. Release of this energy will occur over the complete length of the tendon no mallet where it is cut The sequence of releasing the tendons must be planned in detail to take into account the structure's ability lo carry dead loads without prestressing and the introduction of temporary supports where necessary. Safety precautions should be taken near the anchorages although recent experimental work has shown that most of the energy is dissipated by friction, dislodging the wedges and breaking the concrete cover'n'.

90 .

REFERENCES
1.

THE CONCRETE m . "The design of post-tensioned 'oncrete flat slabs S in buildings." Technical repon No. 8. Publication No. 53.028. 1974.
T H E CONCRETE SOCIETY. "Rat slabs in post-tensioned concrete with

2.

particular regard to the use of unbonded tendons -design recommendations." Technical repon No. 17. Publication No. 51.079. 1979.

3.
4. 5.

THE CONCRETE SOUETY. "Post-tensioned flat slab design handbook." Technical repon No. 25. Publication No. 53.044. 1984. B ~ S STANDARDS H INSTITUTION. BS8110:1985 Pam 1 and 2. "Structural use of concrete." BSI, London. EURoCoDE NO. 2. "Design of concrete suuctures. Part 1: General ~ l e and s rules for buildings." Published in draft only. THE CONCRETE m T Y . "Durability of tendons in prestressed concrete." S Technical report No. 21. Publication No. 53.037. 1982. THECONCRETE SOCIETY. "Partial prestressing." Technical report No. 23. Publication No. 53.040. 1983. AALAMI, B.O. AND BARTH, FG. "Restraint cracks and their mitigation in unbonded post-tensioned building structures." Post Tensioning Institute. USA, 1988. MA~EW,P.W. "Practicalconsiderations in design and construction of posttensioned slabs." Sociely of Engineers Journal. Vol. LXXIX No. 2, June 1988. B ~ H - S T A N D A RID S T O N . BS5896:1980. "Specification for high Nm tensile steel wire and suand for the prestressing of concrete." BSI, London. FEDERATION ~ V A T I O N A LLA P R E C O N T R A ~ . P~ DE E "Recommendations for the corrosion prolection of unbonded tendons." Institution of Structural Engineers, Landon. 1986.
FEDERATIONINTERTATIONALEDELA PRECOhWTE. "Grouting of tendons in prestressed concrete." Thomas Telford. 1990.

6.
7.
8.

9.

10. 11.

12. 13.

BRITISH STANDARDS INm710i-f. BS4447:1973. "Specification for the performanceof prestressing anchorages for post-tensioned construction." BSI, London. FEDERATION I ~ A ~ O N A L EPRE~ONTRAINTE. DE LA "Recommendations for the acceptance and application of post-tensioning systems." Institution of Structural Engineers, London. 1981. BRITISH STANDARDS INSTITUTION. BS4449:1988. "Specification for czbon stml bars for the reinforcement of concrete." BSI, London.

14.

15. 16.

Lw. T.Y. "Load balancing method for the design and analysis of prestressed
concrete structures." lournal of the American Concrete Institute Prffieedings Vol. 60,No. 6. June 1963.

17.

B m s H STANDARDS m U n O 5 . "Handbwk to BSBIIO, structural use of IN concrete." Viewpoinl Publications No. 14.015. 1987.

SAUNDERS, BREW, J. AND DUNCAN. "Strength and behaviour of D.H. R.R. closely spaced post-tensioned monosuand anchorages." Post Tensioning Institute. OCL 1987. REGAN, Punching: current practice sheet 106. Concrete, Issue No 12, P.E. Dec. 23 -24, 1985. GUYON,Y. "Limit srate design of prestressed concrete." John Wiley and Son, New York and London. 1972. CONSTUC~~ONINDUSTRY RESEARCH INFORMATION AND ASSOCIATION. Guide No. 1. "A guide to the design of anchor blccks for post-tensioned presuessed concrete." CIRIA, London. May 1976. WYATT.T.A. "JJesign guide on the vibration of floors.'' Publication No. 076. Steel Conshuction Institute, 1989 I N m O F STRUCTURALENGINEERSTHE CONCRETE AND SOCIETY. "Guide to the srmctural use of lightweight aggregate concrele." 1987. THE CONCRETE SOCIETY AND bXilTUTE OF S T R U ~ R A L ENGINEERS. "Standard method of detailing srmctural concrete". August 1989.
IhTERNATIONALE DE LA PRECONTRAINTE. "Guide t0 good FEDERATION practice - demolition of reinforced and prestressed concrete structures." Institution of Slructural Engineers, London. 1982.

BARlH.F.G. AND AALAMI.B.0. "Conuolled demolilion of an unbonded posttensioned slab." Post Tensioning Institute. USA. 1989. WAIDRON,P. w 1 u ~ s . M . S ."Movement of unbonded post-tensioning AM) tendons during demolition." Proceedings of the Institute of Civil Engineers, Part 2. London. June 1989. CONSTRUCIION INDUSTRY RESEARCH AND INFORMATION ASSOCIATION. Report No. 74. "Prestressed concrele - friction losses during stressing." CIRIA, London. 1978. P ~ LF a.. "The ultimate moment of resistance of unbonded prestressed concrete beams." Magazine OF Concrete Research. Vo1.21, No.66. March 1969. pp.43-54 PANNEL,FN. ANDTAM, A. "The ultimate moment of ~sistance unbonded of panialy prestressed reinforced concrete beams." Magazine of Concrete Research. Vo1.28. No. 97. December 1976. pp.203-208.

APPENDICES

POST-TENSIONED CONCRETE FLOORS DESIGN HANDBOOK

Appendix A:

Design examples
A1 A2

Solid flat slab with unbonded tendons One-way spanning floor with bonded and unbonded tendons

A p p n d i B: Appendix C: Appendix D: Appendix E: Appendix F: Appendix G:

Calculation of prestress losses Calculation of tendon geometry Calculation of secondary effects using equivalent loads Calculation and detailing of anchorage bursring reinforcement Numerical baris For fiat slab capacities given in Figures 17 and 18 Vibration of pst-tensioned concrete floors

NOTE: In some of the calculations given, there are instances where "rounded figures have been shown in the equation but the result has been based on the "unrounded" figure.

APPENDIX A: Design Examples

Two examples are given. The first is based on the use of unbonded tendons in a flat slab. The second on both bonded and unbonded tendons in a slab and beam arrangement spanning one way. The examples show differences which arise due t variations in smctural form and o tendon type, and illustrate the different design methods which can be adopted. The procedures used in the examples are widely adopted in the design of such structures today.

Page Example 1: Solid flat slab with unbonded tendons


68

Serviceability Limit State Al.l.l Transverse direction

72
72 87

A1.1.2 Longitudinal direction A1.1.3 Serviceability un-tensioned reinforcement ~ a l c ~ l a t i o n ~

98 99 100 101 103 104 104 105 105

Ultimate Limit Stale A1.2.1 Transverse direction

A1.2.2 Longitudinal direction Minimum un-tensioned reinforcement requirements Summary of presuess and un-tensioned reinforcement requirements A1.4.1 Presuess summary A1.4.2 Un-tensioned reinforcement summary F'pnching shear check Example 2: One-way spanning floor with bonded and unbonded tendons.

Serviceability Limit State A2.1.1 Transverse direction A2.1.2 Longitudinal direction A2.1.3 Serviceability un-tensioned reinforcement calculations

Ultimate Limit State A22.1 Transverse direction A2.2.2 Longitudinal direction A2.2.3 Ultimate un-tensioned reinfolrement calculations

A2.3 A2.4

Minimum reinforcement requirements Summary of prestress and un-tensioned reinforcement requirements

A2.4.1 Presrress summary A2.4.2 Un-tensioned reinforcement summary

Al

Emmpk 1 SoEdflat s with unbonded tendons : M


Equivalent frames are used to model the slab in each direction. Maximum design moments are obtained by a combination of live and dead load, with the equivalent prestress load from the tendons. Where tendon anchorages are away fmm the neutral axis, or are inclined to the neutral axis of the slab (either at the ends of the slab or within the span where tendons are 'stopped ofF), their effect can be included in this method of analysis by the inucduction of moments or point loads.

A floor plan of the building is shown in Figure A1 together with a typical Pdnsverse s u b h e . This example analyses subframes on gridlines 5 and B. Calculations are carried out for full
bay width. The smcture is checked both at Serviceability and Ultimate Limit States. These checks are m out at transfer, during wnsrmction (where typically, when two weeks old, the slab may e d be required to carry its own weighL plus the weight of the floor above at concreting, plus associated consmction loads), and under working load conditions.

concrete: (strength at bmsfer) (elastic modulus at 2 days) 8 (elastic modulus at rransfer from BS8110,P r 2. Section 7"3 at bonded reinforcement

presuessing steel:

1 . mm diameter superstrand with high-density polythene or polypropylene sheath and with 29 lubrication/corrosion protection as detailed in Section 4.2.2.
pt

= 186 kN
= 100 mm2

A,
$,

E,
Loading

= PJA, = 1860 N/mm2 = 195 kN/mm2

(characteristic strength of tendon) (area of tendon) (characteristic strength of prestressing steel) (elastic modulus)

Imposed loading:

finishes:

partitions %red services in floor zone ceiling typical offim building

live load:

Total imposed loading

From Figure 16, a slab depth of 210 mm would be adequate. However, to reduce shear reinforcement requirements (see Figure 17), a depth of 225 mm is chosen. Self-weight (using density of 24 kN/m3) Total dead load Total live load Check at temporary conshuction stage construction load self-weight of slab under construction above additional construction load share this load between two lower floors by propping

=
=

8.6kN/m2
4.0 kN/m2

I
I

i
load per floor self-weight of flwr under design Total construction load per floor worst loading = dead load + live load situation
=

12.6 -kN/m2

At this stage in the calculation, it is recommended that the amount of load to be balanced is considered. The designer's experience can simplify this opration. In this example a balanced load consisting of all the dead load is chosen. (Balanced loads are discussed in more derail in Section 6.7 of this repon)

!
I

t slab (next floor)


Ail dimensions in mm

Section A-A Design subframe (transverse)

Figure Al: F l w r plan and subframe for Example 1


Tendon profiles

Cover requirements in accordance with Section 5 of this report for adequate cover against corrosion for 1.5 hours fire resistance
Take nominal cover to be

25mm

The dimensions from the top surface of the slab to the tendons and reinforcing steel are shown in Figure A2. The positioning of the reinforcement must be considered at this stage, so as to obtain a practical mangemenl of the steel at internal supports.

~firnm cover

tendon and sheath


I

(a) Transverse direction

transverse tendon '

25 mrn cover

I
I
I

itendon and sheath s mmgf


(12.9+3)

(b) Longitudinal direction

Figure A2: Tendon and reinforcing steel positioning

Preliminary shear check


Taking a slab depth of 225 mm, check the punching shear capacity, and the shear capacity at the face of the column, for both internal and external columns. Internal Columns (500x 500) The load From the slab on lo the internal columns will be Fearer than that due to half the span because of elastic disnibution (see 6.6). For this example, a reasonable estimate for the increase in equivalent flwr area is a factor of 12. Equivalent floor area = 12 ( . + 2.25) x 7 . 35

= 48.3 m3

From Figure 17b for a total imposed load of 7.2 kN/m2 and equivalent floor area of 48.3 m2. some shear reinforcement will l required for a slab of depth 225 mm. x From Figure 1 the check for maximum shear stress (0.8df, or 5 N/mm2) is tine. 8 Edge Columns (300x 300) Figures 17 and 1 are set up for internal columns. In order to use the figures for an edge 8 column, the equivalent loaded area is doubled.

Flwr area = 2 x 3.5 x 3 5

24.5 m2

From Figure 17a for a totnl imposed load of 7.2 kNlm2 and equivalent flwr area of 24.5 m2, no shear reinforcement is required for a slab depth of 225 mm. From Figure 18 the check for maximum shear stress (0.8df, or 5 N/mm2) is fine.

M.IJ

Transverse direcnbn

After deciding the limiting tendon eccentricities (Figure A2) and the pasitions of the poinrs of inflection - 0.1 times the span, from the cenke of suppons - the tendon profile can be cdculared: see Appendix C.

Figure A . Transverse tendon profile 3

Calculation of maximum drape Assume that the maximum drape occurs at midspan. Using the equation of a parabola'

from the tendon prome calculation (Appendix C), we know !hat

also
when

s = 3M)O mm
x = 1800 mm

y = 87.16 mm

At this stage losses are assumed as follows: At m f e r 10% of the jacking load At service 20% of the jacking load.
A thorough check will he carried out after the stress calculations to check that these initial assumptions of 10% and 20% are within reason. If they are not, another estimare should be made and the procedure repeated.

Initial prestress
The initial prestress force, i.e, the jacking force, has been taken to be 70% of the characteristic strength (see BSXIIO, Pm 1. C1 4.7). For the transverse direction, the tendons will k saessed along gridline A only.

, Calculation of P
Jacking force = 0.7 x 186 Prestress force at uansfer (10% losses) Prestress force at service (20% losses)

130.20 Wtendon 117.18 kN/tendon 104.16 kNItendon

Next the value of ptesuess force required in each span is calculated. This is done using the chosen balanced load of 8.6 kN/mZ(the dead load). the disrance between points of inflection. s, and the drape a, as shown in Figure A4.
C
a = 87.16

P
,P ,
= ws2
8a

a = 87.16

~ i ~ uA4: Drapes Tor load balancing re The prestress force is obtained from the following equation, which assumes aparabalic profile.

For span CB,

, P

= 8 . 6 ~ 7 ~ 3 6 0 0= 1 8 x 87.16 x lo00

1119kN 10.7;
try 11 tendons

Therefore number of tendons For span BA.

=1 1 19
104.16
= 8.6 x 7 x 56001

per panel

, , P

=
=

2707 kN 25.99; try 26 tendons p panel a

8 X 87.16 X loo0 Therefore number of tendons

= 2707

104.16

As the longer span requires more tendons than the shorter span, 15 of the rendons will be stopped off at the paint of inflection in span CB.next to support B. When accurate losses are calcula~ed, different force pmfiie of these shorter tendons must be taken i l account. the no

The effect of the tendons on the slab is modelled by means of equivalent loads, as shown below. Equivalent loads are discussed in more detail in Appendix D and Section 6.9 of this repori It should be noted that the portions of the cable from the edges of the slab to gridlines A and C are horizonral and so do not contniute to the equivalent loads.

Figure A5: Calculation of equivalent loads due to tendon The equivalent load, w, between any two poins of inflection for the chosen number of tendons is given by:

where:

n a s

P ,

is the number of tendons is the drape at the point considered is as shown in Figure A5 is the average force provided by each tendon.

11 Equivalent load. a1 uansfer

1 Full length tendons (n = 11)

11 Short tendons (n = 15)


I
I
232.6

11
I I
-69.4 900 439.6 761.9 1400 181.7 314.9 5600 -39.1 -67.7 1400 131.1 227.2

a (mm)
W

Wm)

Total w (kNIm)

I
Equivalent loads at m f e r

I(Full length tendons (n = 11)


~ ~~

11
18.3
900

a (mm)

-87.2
3600

25.3 900 286.5

25.3 1400 118.4

-87.2 5600 -25.5

18.3 1400 85.4

s (mm)
W

mum)

206.8

-61.6

Short tendons (n = 15) n x P,@W

/ I I
-61.6

1562.4

1562.4

1562.4

1562.4

a (mm)
w (kN/m)

900 390.7 383.3

1400 161.5 158.4

5600 -34.7 -60.2

1400 116.5 202.0

I
206.8

Total w (kN/m)

Table Al:

Calculations of equivalent loads due to transverse tendons, at transfer and after all losses.

When tendons are anchored within the span, as in this example, additional equivalent loads may be generated by the end condition. These must be included in the Erame analysis when obtaining the bending moments and shear force diagmms. The forces consist of a vertical and horizmtal component of the tendon farce applied at the anchor. Figure A6 below shows the effect of an anchorage in terms of additional equivalent loads on the slab.

PI

.Psin rr
tendon centroid

Figure A6: Equivalent loads at anchorages. The vertical component of the tendon force is easily calculated, and should be applied to the slab as a vertical point load at the point where the tendon is anchored. The horizontal component forms a positive moment about the centroid, owing to its eccentricity from the centroid of the section, and should be applied in chis form to the slab. It should be noted chat the position of the tendon at the anchorage can be arranged so that the tendon is'both horizontal (no vertical force) and at the centroid of the section (no eccentric moment). In this example the anchorages at the ends of the full-length tendons fulfil this requirement and no additional loads are generated. Vertical force = Psina Eccentric moment at the point of inflection = Pcosa x e For a parabolic tendon

&=
dx

s2

and so sina = Therefore:

2a
S

sina =

2 x 25.32 450

= 0.1125

Eccentricity of tendon =

-112.5 + 25(cover) + 16(diieter of un-tensioned reinforcement) + 8(half the tendon diameter) + 25.32(drape)

At kansfer

Psina =

197.82 ItN

After all losses

P =
Psina =

1562.4 kN 175.84 kN

The equivalent loads from the tendons, the anchors and the superimposed loads are then used to calculate design momenE and shears by any convenient method of structural analysis. This is normally done using an appropriate computer program.

At Seviceabiliiy Limit State, an elastic method of analysis should be used for analysing posttensioned flat slabs, and patterned loading should be used in multi-span situations (see BS8110, P r 1, Section 4.4'')). at

Summary of prestress equivalent loads

loads after all

268 0.

-61.6

677.3

279.9

60.2

202.0

Table A2: Summary of uniformly distributed equivalent loads from transverse tendons.

Vertical force
At transfer At service

Eccenhic moment

197.8 kN 175.8 kN

66.7 lcNm 5 . kNm 93

Table A3:

Summary of additional equivalent loads due to internal anchorages.

Summnry o applied bending momenls f

(a) SeK-weight Only

@) Service Load Envelope

At Transfer , (c) Due to equivalent presuess loads

At Service
203.7

At Transfer (d) Total Applied Load Figure A7: Applied bending moment diagrams

At Service

Check that Prestress Loads tolal zero


Upward londs

= 61.6 x 3.6 + 60.2 x 5.6

+ 175.84

Downward loads = =

(206.8 + 677.3) x 0.45 + (279.9


735.18 kN

+ 202.0) x 0.7

The smatl difference between rhese values is due to earlier approximations. The equivalent loads were altered to total zero at !his point to enable consisrent calculation of secondary moments.

Calculation o stresses f

A.

7 x 0.225 X lo6 = 1.575 x 10' mm'

As the section being considered is rectangular and symmetrical about the centroid, ,z and z , are equal.

As this example is a flat slab, analysed by the equivalent frame method. the allowable stresses s are a detailed in Table 2 (Section 6.10.1) of this report. To increase ease of construction, untensioned reinforcement ha. been deliberately omitted from the spans by keeping tensile stresses below 0.15Jf, (transfer) and 0.15Jf, (service). All tensile stresses are negative.

Table A4: Stresses at transfer for the transverse direction

* These values include prestress secandary effecrr.

Table A5:

Shesses arter aU losses for the transverse direction.

no Hogging and sagging values are given where they are both in one zone. Each span is split i t to L. three zones, from the end to & from Z to and from
10 10 10 10

In this example, the construction load is smaller than the load at service and larger than that at hansfer. This means that the construction case is not likely to be a governing situation and so the stresses are not calculated.

Loss c a l c u h ~ n s
At this stage the losses should be calculated accurately to check that the initial assumptions of 10% at transfer and 20% at senrice were reasonable. The method for calculating the various types of loss is given in Appendix B.

Full-length tendons Shorl-ferm losses


a) Losses due to friction

Table B1 gives recommended values for the coefficients p and a:


p = 0.06

and

m = 0.05 radslm
= 16 x total dram

deviated angle per rneBe, a'

L2
total drape (the same for both spans)
Span CB
a'

= 18.27 t 25.32 + 87.16


2 = 16 x 108.96 x 10" 4.5'

108.96 mm

0.086 rad/m

Span BA

a'
Jacking force

= 16 x 108.96 x 10"
72 = 130.2 kN

0.036 rad/rn

Forces after friction losses (see Figure A8) are:

b)

Losses due to wedge set

force loss at anchorage, 6Pw= 2p'I' where: p' = slope of force profile 1' = length of tendon effected by draw-in

lake wedge draw-in, A = 6mm Eps = 195 IrN/mm2 Aps = 100 m ' m

therefore,

I'

= d((6 x 1 0 3 x 195 x 100)/0.79) = 12.14 m

1 '

As I' is grealer than the length of the tendon,

SP, at suessing
anchorage and

= (A x q,x &)/I + (P' x 1)

S , at dead end P
therefore,

= (A x E, x %)/I - @' x 1)

6. at stressing P
anchorage

= 16 x 195 x 1001 + 0.79 x 1 1 5 11.5 x 10' = L x 195 x 100) - 0.79 x 11.5 6 11.5 X 10'

SP, at dead end

Forces after friction losses and wedge see (see Figure A8)

C )

Elastic losses

where:

E,

= 0.5 x f,

E.1
f, is the stress in the concrere adjacent r the tendon. Since this is unlikely to be critical, the n stress is calculated at a representative point and will be taken as uniform over the whole tendon length.

Prestress at tran$er
Prestress force at A = 110.9 - 0.89 Prestress force at B = 117.4 - 0.89 Prestress force at C = 120.0 - 0.89

Long-term losses
a)

Relaxation of steel SP, = 1000-hour relaxation value x relaxalion factor x preskess force at transfer

From Table B2 in Appendix B values are taken for an initial jacldng'force equal to 70% of the characteristic strength. loss due to relaxation = 2.5% relaxation factor = 1.5

b)

Shrinkage of concrete

SPA

= EBX

P$.

from BS 8110, Part 1, Clause 4.8.4'"

C)

Creep of concrete

where $ = creep coefficient (see Appendix B) !

Prestress after all losses


Presrress force at A Prestress force at B Premss force at C
= 110.0 - 4.13 - 5.85 - 3.57 = 116.5 - 4.37 - 5.85 3.57 = 119.1 4.47 - 5.85 - 3.57

119.1
105.3

PL1.,
FORCE A T TRANSFER

STRESSING FORCE

//!
I 126.6

130.2

116.6

--------102.7

110.0 96.5

FORCE AFTER A L L LOSSES

STRESSING END

Figure AS:

Force profiles for full-length tendons

- Short tendons
Friction losses and elastic losses are the same as for the full-length tendons, as are the longterm losses. The effect of wedge set is different as the tendon length is different and must be recalculated.

STRESSING FORCE

I I

FORCE AT TRANSFER
A

I
-

FORCE AFTER ALL LOSSES

108.5

99.8

J
95.0

Figure A9: Force profile for short tendons


n") Force at dead end = 125.58e~aC5"0~~am6+

= 125.14 kN

6P. at stressing anchorage

Forces after friction losses and wedge set:

Prestress at transfer
Resmss force at A = 109.4 0.89 Reseess force at B = 114.3 - 0.89

Prestress afrer all losses


R e s k s s force at A = 108.5 4.07 5.85 3.57 Preseess force at B = 113.4 4.25 - 5.85 3.57

= 95.0 kN

= 99.8 kN

Short- and long-term losses


Check that the losses assumed initially are reasonable by calculating accurately. Average short-term loss for span CB (full-length tendons only)

Average shm-term loss for span BA


@oth tendon lengths)

Average overall short-term loss Average long-term loss for span CB (full-length tendons only)

Average long-term loss for span BA (both tendon lengths)

Average overall long-term loss

Although the assumed losses of 10% and 20%, respectively, have been exceeded in span BA. recalculation is not considered necessary, as it will not cause an increase in the number of tendons. Also the calculation OF stresses for the correct losses are unlikely to exceed the allowable values.

AI.I.2

L o n g i h i d i ~ direction l

Analysis of longitudinal frame along gridline B.

Figure A10: Longitudinal tendon profile The method of calculation follows that given for the msverse section. Losses as assumed before, 10% of the jacking load at transfer 20% of the jacking load at service Initial prestress forces as before Prestress force at transfer, P, Prestress force at service, P," The elastic reaction on the internal columns along grid line B at working load was calculated for the transverse direction,

The total uniformly distributed load Hence the effective width of slab =

593
12.6 x 7

Assume a balanced load equal to 100% dead load as for the transverse direction.

Where w
S

= 8.6 kN/m2 = 5mmm = 81.54 mm (for span 1 lo 3 and 9 k 11) a 101.6 mm (for all other spans)

The length of slab is greater than 30m and so stressing is assumed to take place from both ends. This produces symmetry about gridline 6 and so the tendon requirements on either side of this will be the same. Spanl-3: , , P

8.6 x 5600' 8 X 81.54 X loo0

= 413.44 kN/m width

Chwse number of tendons for this load balance:


Span 1 3 : No. of tendons

= 413.44 x 6.72

= 26.7, say 27
= 21.4, say 22

104.16 Span 3 5,5 7 : No. of tendons = 331.81 x 6.72 104.16

I
Equivalent loads at tmnsfer

(1 Full-length tendons (n = 22)

Equivalent loads after all losses

11 Full-length tendons (n = 221

11

1) Short tendons (n = 5)

w WIm)

37.7

-10.8 -58.5

48.9 264.1

54.0

I
-59.4

I
237.6

I
-59.4

- Total w OcN/m)

203.6

291.6

Table A6: Calculations of equivalent loads due to longitudinal tendons, at transfer and after all losses.

Effect of tendons anchored within the span sina cosu


=

700
= 0.997

At transfer

After all losses

Summnry of prestress equivalent load analysis

I
Equivalent loads at transfer Wlm) Equivalent loads after all losses (kN/m) Table A7:

I
291 2. 203.6 -65.8 -85 5. 297.1 241 6. 328.0 216 9. -66.8 -94 5.

I
267.3 237.6 -66.8 -59.4

Summary or uniformly distributed equivalent loads from longitudinal tendons.

Vertical force At transfer At service Table AIl:

Eccenhic moment

4 . kN 25 3 . kN 78

1. W m 29 1 . kNm 15

Summary of additional forces due to internal anchorages

Summary o applied bending moments f

(a) Self-weight Only

@) Service Load Envelope

At Transfer At Service (c) Due to Equivalent Prestress Loads

At Transfer

At Service
(d) Tolal Applied Load

Figure A l l : Applied bending moment diagrams

Calculation of stresses

1
Zone
I

S m s e s are calculated as far the transverse direction.


S t r w due to prestress * (Nlmm2)
DP

Stress due to self-weight (N/mma)

Combined stress (NJmm3

Allowable

stress
(NJrnrn3

46 .1 02 .8 -.6 13 6.25 3.32 15 .7

-1.39 13 .9 2.00 -2.00 -0.36 03 .6

32 .2 16 .7 0.M 4.25 29 .6 1.93

6.00 -2.25 -0.75 8.25 82 .5 -0.75

1
bottom
DP

1-3(hogging)

1-3 bgging)

bottom
DP

bottom

bottom
DP

-1.16 23 .9 1.60 -1.15

13 .9 -0.12 0.12 15 .3 -1.53

02 .3 22 .7 17 .2 0.38 36 .1

-2.25 8.25 -0.75 -0.75 82 .5

5-7( = g g k )
bottom

5-7 (hogging)
bottom Table A9:

5.14

Stresses at transfer for the longitudinal direction

Table A10: Stresses after all losses for the longitudinal direction

Loss calculations
Lasses are now calculated as before to ensure that the initial assumptions are reasonable.

- Full-length tendons
Short-term losses
a)

losses due to friction Span 1-3: Total drape

= 101.92 m m

Span 3-5:

Total drape

= 127 m m

Span 5-6:

Total drape

= 127 m m

Forces after friction losses are:

PI
p 3

= 0.7 x 186

= - 1303~'7 n0smm3~.m)
= 125.72~'7 W0alfl.m X

p 5

p s
b) losses due to wedge set

- 120.gge'35 x 0 . W W l M ~

I' is less than the tendon length, therefore

Forces after friction loss and wedge set

C )

elastic losses

Prestress at tranrfer

(See Figure A12)


Presness force at 1 Presness force at 3 Prestress force at 5 Prestress force at 6

= 112.6 1.20 = 117.4 1.20

= 121.0 - 1.20

= 118.7 - 1.20

Lung-lerm losses
a)

relaxation of steel

as before 3.754b

b)

shrinkage of concrete

as before .g5kN
C ) creep of concrete

Prestress force after all losses


Prestress force at 1 Prestress force at 3 Prestress force at 5 Presuess force at 6
= 111.5 = 116.2 = 119.8 = 117.5

- 4.18 - 5.85 - 4.8 - 4.36 - 5.85 - 4.8 - 4.49 - 5.85 - 4.8 - 4.41 - 5.85 - 4.8

0
13.33 rn
L

&

130.2

STRESSING FORCE

FORCE AT TRANSFER
111.5
96.6

FORCE AFTER ALL LOSSES

Figure A 2 Force profile for full-length longitudinal tendons. 1:

FORCE AT TRANSFER

108.9
94.1

1 1

FORCE AFTER ALL LOSSES

I I
I

7700

Figure A13: Force profile for short longitudinal tendons.

- Short tendom
Recalculate wedge set for the shorter tendons, all orher losses as before.
+

Force at dead end

= 125.72e4.'

x"06@a41

. " ) '

125.25 kN

6 , at dead end P

= 6 x 195 x 1004/700 - 1130.2 125.251 x 7700


7700

GP, at stressing anchorage

= 6 x 195 x 100/7700 + j130.2 - 125.25) x 7700


7700

20.14 kN

Forces after friction losses and wedge set:

Prestress force at transjer


Prestress force at 1 Prestress force at 3

= 110.1 - 1.2 = 113.4 - 1.2

Prestress force after all losses


Prestress force at 1 Prestress force at 3
= 108.87 - 4.08 - 5.85 - 4.8 = 113.4 - 4.25 - 5.85 - 4.8

94.1 kN 98.5 kN

Short- and long-term loss values


Average short-term loss for span 1-3 (both full-length and shon tendons)

Average short-term loss for spans 3-5 and 5-7 (full-length tendons only)

Average overall short-term loss Average long-term loss for span 1-3 (both full-length and shon tendons)

Average long-term loss for spans 3-5 and 5-7 (full-length tendons only)

Average long-term loss Initid essumptions E e fine.

22.xLpb

A1.13

Designed serviceobil&v un-tensioned reinforcement calculations

Un-tensioned reinforcement is required in the span if tensile stresses exceed 0.15.lf, (see Table 2).

Transverse direction
No tensile stre'esr, al C, therefore no designed un-tensioned reinforcement is required at serviceability. All stresses in the spans have been kept below 0.15Jf, so no un-tensioned reinforcement is required in any of the spans.
At support B, there is a tensile stress and so designed un-lensioned reinforcement is required.

steel required tensile force

4
F,

0.625fr
=

f-(h-xlb 2

h-x = _f, x h - fm+ f= At B, worst slress situation is when f , = -1.727 N/mm2


fa

= 5.165 N/mm2

h = 225 mm

b = 7WOmm

use

= 1210mm2

Longitudinal direction
Tensile stresses at suppons 3 and 5. Support 3 Support 5

h-x

= 45.99 mm

h-x

= 43.61 mm

A,
Use

= 842 mm2 = 1005mm2

A,
Use 3T16

= 595 mm2
= 603mmz

The equivalent frame analysis at Ultimate Limit State may be carried out in accordance with Clause 3.7 of BS8110, P r 1, using the simplification of load arrangements given in Clause at 3.53.3. Section analysis may be carried out in accordance with Clause 4.3.7 of BS8110, P r 1. at From BS8110, Pan 1, Equations 52 and 53'"'.

where: f,

= characteristic strength of prestressing tendon

where: $ ,

= design effective prekvkss in the teqdons after all loses.

The value of I is taken as the full length of the tendons. This may be considered to be conservative (see Section 6.10.3). First calculare f, and x. then calculate h.L, and compare with MA (the applied moment).

AI.2 J Transverse direction

Applied Momenh
Figure A14 shows the moment envelope for the factored dead and live loads on bolh spans with 20% redistribution.

Figure A14: Applied moment envelope

Moments of Resistance
The simplified method given in C14.3.7.2 of BS8110, Part 1 is used to calculate M,, at each critical section.

1 = 11500 mm

Table A l l :

Ultimate capacity due to prestressing tendons over the full panel width (7.0 m).

Having calculated the ultimate moment capaciries above, a comparison wiih the total applied moment at supporn and midspans must be made t fmd out whether further un-tensioned o

reinforcement is required.

The total applied moment is the sum of the moment From the Ultimate Limit State and ihe secondary moment Secondary moments arc discussed in more detail in Section 6.9, and methods of calculation are shown in Appendix D. The values given in Table A12 are for the full panel width (7m).

M,
secondary moment Wm) Table A1Z:

B
.

B
40.7

28.7

1. 08

83.7

Secondary moments for the transverse direction using method B

The values shown in the above table have ba - calculated from method B given in Appendix D.

Table A13:

Comparison of applied transverse moments and of resistance a t Ultimate Limit State

No un-tensioned reinforcement is requited

A1.2.2 LongiMinal direction


Applied Moments
Figure A15 shows h e moment envelope for the factored dead and live loads on both spans with 20% redistribution.

Figure A15: Applied moment envelope

Table Ald:

Ultimate capacity due to prestressing tendons over the full panel width (6.72111).

Secondary moments and the applied momenrs are calculated as before.

M .
secondary moment Wm) Table A15:

1
134.9

55.0

58.0

69.2

66.6
2

Secondary moments for the longitudinal direction using method B.

Table Alh:

Comparison of applied longitudinal momenh and of resisfance at Ultimate Limit State.

No un-tensioned reinforcement is required.

AI3

Minimum un-tensioned reinforcement requirements


(see also A1.1.3)

Transverse direction
Supports C. B and A reinforcement required = 0.075% x

4 (see Section 6.10.5)


=
1.575 x 106 mm"

= hb = 225

70M)

Ue s

= 1206 mm2

Steel positioning Along external edges 50% of the above values are used and positioned as shown in Figures A16 and A17.

j
I

Figure A16: Transverse reinforcement positioning for internal columns The reinforcement should extend into the span by 0 2 x span measured fmm the centreline of the column. The width of strip is the column breadth plus three times ihe slab depth.

Figure AD: Reinforcement positioning for external columns The width of strip is the column breadth plus 1.5 Limes the slab depth.

tongitudiml direction
Suppons along grid line B. reinforcement required Use

= 0.075 x 225 x 6720


100

= 1206 mm2

Along the external edges 50% of the above value is used. The reinforcement is arranged in the same manner as for the hZnSVetSe direction.

M.4

Swnl~rargof prestress and un-tensioned mmlaforcement requimmcnfr

A1.4.1 Prestress surnmory


Prestress tendons are unbonded 12.9m diameter superstrand with a jacking force of 0 7 $ = 1302 kN/tendon. . ,

Transverse direction

Span

Tendons per panel 11

Prestress force

CB
BA

163.7W 386.9 kN

26

Note: Prestress force given per metre widlh of slab. for the service condition. Summary of prestress requirements in the transverse direction.

Table A17:

Longitudiml direction

Span

Tendons per panel

Prestress force

1-3
3-5 5-7

27 22 22

4 9 1 kN 8. 398.5 kN 3 8 5 kN 9.

Nore:

Restress force given per metre widlh of slab, for the service condition. Summary oi presiress requirements in the longitudinal direction.

Table A18:

h ~ g i f u d idirection ~l

Pane1 width Effective depth, d,, = 225 - 25 - 8 Length of critical perimeter perpendicular to longihldiml direction, 4, = 500 + 3 x 192 Area of reinforcement provided (6T16s) Width over which reinforcement provided Area of reinforcement provided in width b, , = 1205 X 107611175 100 U , d , , = 100 x 1104/(192 X 1076) ~ Hence v,, = (0.7911.25) x ( 0 . 5 3 ) ' ~(400/192)*x (4005)" Effect of Prestress e,, = -22512 + 65 + (25 say) For Span 3:5 M V = ((331.9 - 355.9) + 18.44 X 6.72 X 7 X 3.5)/7 V , : = 0.76 x 1076 x 19211000 + 41.95 x 430.3D31.9 For Span 5:7 V,,, = 0.76
X

1076 X 19211QQQ 41.95 +

433.7D37.7

Total Shear Resistance, V, = 214 + 172 + 21 1 + 21 1


Applied Shear Force Transverse direction provides worst condition v = 499 + 353 M = 448 - 297.8

V,, = V(l + 1.5 W x ) see BSgllO, Equation 25 where x = b, V,, = 852 (1 + 1.5 x 150/(1.028 x 852)) Vm,> V, hence shear reinforcement required.

& required = 1000(1071 - 743)/(0.87 x 460)


Using T6 single leg links Number required = (819 x 4)/(1r x 36) Maximum spacing permilled = 1.5d = 1.5 x 176 Place 12 No on a perimeter of 0.5d from the column face Place 24 No on a perimeter of 2.25d from the column face

A2

E m @ 2: One-wny spanningflaor with bonded and unbonded tendons


The design philosophy used in this example is similar to that used in the first. Since it is a one-way system a preliminary shear check is considered unnecessary. The floor plan and subframe for the sbucture are shown in Figure A18. It can be seen that the sbucture is a one-way system, with a ribbed slab spanning onu, band beams. The band beams run along the column lines in the longitudinal direction. Unbonded tendons are chosen for the ribbed slabs, and bonded lffldons for the band barns.

SECTION A-A

Figure BIS: Flwr plan and subframe for Example 2

Properties
Concrere and bonded reinforcement as for the first example. Restressing steel: 15.7mm diameter supersh-and is used for both bonded and unbonded tendons, with a highdensity polyhene or polypropylene sheath surrounding the unbonded tendons, as detailed in Section 4.2.2.

P = 265.5 kN ,
= 150 mm2 ,f = 1770 N/mm2 E, = 195 kN/mm2

(characteristic force of tendon) (area of tendon) (characteristic strength of prestressing steel) (elastic modulus)

Imposed loading for a typical office building: finishes: live load: total imposed load In this example the loading from finishes will be considered as a live load. This will rake account of moveable panitions. Total live load = 5.0 kN/m2 Using h e values of span/depth ratio given in Table 1, h e section dimensions are as shown in Figure A19. For analysis purposes, the nibed section will be treated as one large T-section spanning in the eansverse dimtion as shown in Figure Am.

(a) Section though rib

1200

1
1
I

bj
1

7,

EFFECTIVE FLANGE WIDTH = 2508mm

-760

[
1
750
1

(bl Section through beam

Figure k19: Section details

Wgure A20: Taeetian w d for analyses

- transverse direction
self-weight of T-section
C'
,

= (0.11 x 7.2 x 24) + (024 x 1.8 x 24)

/'

29.38 W/m

Although the T-section becomes a rectangular section when it meets the band beam, for this example the T-section will kconsidered t run the full distance between the suppons. o An additional load wl be applied t the T-section at the support ends, to represent this il o additional self-weight additional weight of concrete. at supports due to band beam

- longitudinal direction
self-weight of band beam

Balonced load

- transverse direction
balanced load = 1.5 x 29.38

,, -

For the ttansverse direction lake a balanced load of 1.5 times the self-weight.

- longitudinal direction
Tendon profiles

For the longitudinal direction take a balanced load equal to the self -weight

Cover requirements as for the fust example.

At this point the practical arrangement of the tendons and un-~nsionedreinforcement must be considered, espxially es links will k required in bath the ribs and the band beams. (See Figure A18). Steel mesh wl be placed in the rop of the ribbed slab as a provision against il

cracking, and l reinforce the mncrere for a loading situation between ribs. Continuity berween o slab and beam will b maintained by swaight bars which pass from the beam into the slab, and e which are then tied to the mesh. It should be noted that the cover to tendons in ducts must not be less than 50 mm.

(a) Transverse direction

5 mm 0 mesh at 150 mm spacing

/
I
TI6 nominal

I
reinforcement unbonded tendon
V'

b
(cl Section through rib mid span

1 Omm 0 open link

4 _ 1 ,25rnm
T

Figure All: Tendon and reinforcement positions

111

A2.Z.Z Transverse direction


The kndon profiie is calcuialed as before and the resulting profile is shown in Figure ,422.

9000

9000

Figure A22: Transverse tendon profile


Calculation of maximum drape From Figure A22 and Appendix C y = k(s-x) k = 1.40 x 1C5 and s = 72W mm therefore, when x = 3600 mm, y = 181.70mm Losses are assumed to be 10% and 20% of the jacking loadas before.

r:

Initial prestress
'I

The jacking force is taken as 70% of the characteristic seength. w..:

Seessing of the tendons will take place along gridline A only for the hansverse direction.
ladring force in tendon = 0.7 x 265.5 -. Prestress force in tendon at uansfer, P , Prestress force in tendon at service. P ,

148.68 kN $GI.(.>
YO

Calculate the prestress force required in each span using the balanced load previously chosen and the drapes shown in Figure A23.

Figure A23: Drapes for load balancing


Force required is the same in each span {

fiv?l' ,j*.*s',-L

lbLf

p'J

Tnerefore, number of tendons required = 1 7 . 8 516 148.68

As there are six ribs in the section, N o tendons are chosen For each rib giving a total of 12 tendons.

P, (at wnsfer) P (after all losses) ,

= =

12 x 167.27 12 x 148.68

2W7.2.W 1784.16 irN

Equivalent loads are now calculated as for the f s example. ut

MP..
s2

where

w is the equivalent load


n is the number of tendons a is the maximum drape of tendons for zone considered (mm)
s is the distance between tendon inflexion points (mm)

I Equivalent loads at m s f e r (n = 12)

I(Equivalent load.; alter all losses (n = 12)

Table AS1: Calculations of equivalent loads due to transverse tendons, at transfer and after all losses, for the full slab width. Using an appropriate computer program, the bending moments and shear forces for each case can be obtained and then the snesses calculated. As for the transverse direction in example 1 patterned loading for the SLS is used.

It should be noted that, as the section being considered is not rectangular, z, and z are not , e u a l and I is not bd3/12.
By calculation it can be shown that I = 1.14 x 10" m4 and that h e cennoid of the section is 2330101 above the soffit Therefore:

z, = 97.4 x lo6 mm3 z = 48.9 x lo6 mm3 ,

Sunamary of results

I
Equivalent loads at m s f e r OcN/m) Equivalent loads after dl losses (kNIm) 202.6 180.1 -56.3 -50.0

I
247.8 220.2 -56.3 -50.0 202.6 180.1

Table A22:

Summary of equivalent loads from transverse tendons for the fuli slab width.

CaIculnh'on of stresses

The shwes can now be calculated. (Owing to symmeuy about gridline B, the stresses at A and C will be the same, as will those in spans AB and BC.)

Area of section

= (7.2 x 0.11) + (0.24 x 1.8)

1224 rn2

/"

Stresses at msfer: 'The maximum allowable tensile and compressive shesses for one-way spanning structures at m s f e r are the same for both span and support locations (BS8110, P r 1: Clause 4 . 3 3 at max. compressive shess = 0.5 t, = 0.5 x 25 max. tensile s e a s = 0.36dfd = 0.36 425

12.5N/mmz 1.8 N/mml

Support region (un-tensioned reinforcement

Table A23 Stresses at transfer for the transverse direction.

_, Stresses after all losses: 7

C;:->J

..

/\

The maximum allowable compressive and tensile stresses for one-way spanning structures after all losses are shown below, using Tables 4 2 and 4.3 of BS8110, Pr 1, Clause 4.3.4.2 and at Clause 4.3.4.3'''.

Take the crack width limitation to be 0.lmm. mar tensile stress = -4.1 x 1.025

, ' (

Ltc.>A
r

h.0

r;

407 -

1 . 4 -

.-

-42N/mm2

ma*. compressive stresses:

- span locations
mar compressive stress = 0 3 f .3,

.f! ,

~S>,,@O

13.2 N/mma

max. compressive stress = 0.4f,

c.i+ ~0 \ , :

16.0 N/mm2

-tensioned reinforcement

Table A24:

Stresses after all losses Tor the transverse direction.

Loss calculations

- Short-term losses
a)

losses due to friction

P ,

= P,x e.P**+q

.
a are,

Recommended values for the values of p and

p
total dmpe

= 0.06 and n = 0.05 rads/m = 40.88 c 49.99 2

+ 181.7

deviated angle per mem. a', the same for both spans is

a'
p* ps

16 x 227.14 x 10"
g2

jacking force 185.85 x e'9 176 56


c 4 ! .
X

PC

~ + b m w RBI *

'o O 5 + 0.W mW .

b)

losses due t wedge set o

a',
where: I'

2pl'

d ( ( ~ $ x P$3/p') x

and
take therefore: and

A
I'

6mm 13.20 m

S, P

26.66 kN at the stressing anchor

Forces after friction lases and wedge set: (see Figure ,424)

c)

elastic losses

= ,

%X%XA,

from s t m s calculations, f, = 2.183 N/mm2 therefore:

.cl,

and

1.47 kN

As this is l u s than 1% of the initial force, elastic losses will be ignored.

Prestress at transfer
Presmess force at A Presbess force at B Prestress force at C

a)

Relaxation of steel
SP, = 3.75% as for the k t example

b)

Shrinkage of concrete

C)

Cresp of concrete

Prestress after all losses


Restress force at A = 159.19 - 5.97 8.78 - 4.71 Presaess force at B = 168.22 6.31 8.78 - 4.71

P r e s h ~ ~ at C = 167.73 - 6.29 - 8.78 - 4.71 For~e


7,

13.20

185.9'
159.2 139.7

176.d

.368.2

148.4,

_- I% + ,
[Ca
,

1172.4

167.7

-2-

147.9

Figure

Force profiies
4 1

,;: aq.3+' , ,.'

j\?

Short- and brig-term losses


average shorr-term loss average long-tern loss

Tq ,,.++l96

= (14.i%+ 9.5% c 9.7%)/3 = (24.8%+ 20.1%

Although the initial values of 10% and 20% have been exceeded, the calculations will not be revised. This is because the stresses at transfer are not close to exceeding any of the allowable hits, and the number of tendons would not be affected if the losses were increased by 1-2%.

A2.12 Longitudinal direchn


The effective width of slab was found fmm the transverse direction shear force diagram for full unfactored load.

Self-weight =

Weight of slab

Weight of beam

w' @-=
27.14

-.

r'

48

134.12 I

I '

--

184

00

-I
2331
1

I
3429 2000

2000

j
720

L
250 720' 7200 7200

!
e
1

Figure A25: Longitudinal tendon profile

In this example the centmid of the tendons was taken t be the same as the centroid of the o duct, as with this arrangement (20mm oval duct with 15.7mm tendons) the difference in eccennicity is negligible. (See section 6 7 of this manual.) . Calculation of maximum drape in each span. Span 1-2 Fmm Figure A25 and Appendix C

k = 1 7 x 10" and .4

s = 57M)mm

Maximum drape assumed to be at x = W2 = 2880 Maximum drape, y=144mm A detailed calculation would show the maximum drape a r m s at x = 2331 mm and is 139 mm as shown i n Figure h25. Span 2-3 lviaximum drape

Losses assumed to be:


15% of the jacking load at transfer 25% of the jacking load after all losses

These are higher than far the transverse direction because losses tend to be higher for bonded
systems in beams.

Preslress forw at hansfer, P , PtesWss force at savice, P , ,


Calculate presDess force required for each span

Balanced Icad, w = 48.2 kNhn


Spans 1-2 and 4-5

Spans 2-3 and 3-4

For a bonded shuctme, construction is easier if the number of tendons in each span is e q d Although ten tendons w.dd be just sufficient from this approximate dculation, the choice of 11will ensure no overstresing in the outer spans even if the calculation of final losses exceeds 25%. For this analysis it is considered reasanable to calculate the effective flange width for stifmess and stresses according to BS8110 Clause 3.4.1.5, ie. web width plus 0.7 times p 4 5 . The lcading width is, of come determined by the analysis in the transverse direction.
As the beam is a T-section, z and 2 are not equal. By calculation it can be shown that , , I = 6.79 x llT3 m and that the centroid of the section is 196 mm above the soffit ' A = 0.6359 mZ.

Therefore:

z,
Zb

44.1 x I@ mm3 34.6 x 1@ mm3

Equivalent loads at transfer (n = 11) nxP , o a (mm)

13. 776 3. 08

1737.6 -144.0 5760 -60.3

1737.6 4. 12 1440 276.2

1737.6 46.00 1440 308.4

1737.6 -184.0 5760 -71 7.

13. 776 46.00 1440 308.4

s (mm)
W (IrNb)

1440
265 0.

Equivalent loads after all losses (n = 11) n x P-W a (mm)


a (mm)
W W/m)

1533.3 3. 08 1440 182.2

13. 533 -144.0 5760 -532

1533.3 41.2 1440 243.7

13. 533 46.00 1440 272.1

1533.3 -184.0 5760 -68.0

1533.3 46.00 1440 272.1

Table A25: Calculntions of equivalent loads due to longitudinal tendons, at transfer and after all losses, for the full slab width.

Summnry of equivalent loads

I
Equivalent loads at nansfer W m ) Equivalent laads after all losses (IrNIm)

I
206.5 182.2

I
308.4 272.1 -77.1 -80 6. 384 0. 221 7.

60.3
-532

276.2 237 4.

Table A26: Summary of equivalent loads from longitudinal tendons for the beam.

Cakularion o stresses f

Table A27:

Stresses at transfer for the longitudinal direction.

Allowable stress calculations as for the a s v e r s e direction.

Table A28: Stresses at transfer after all losses for the longitudinal direction. Allowable skess calculations as for h e m s v e m direction.

Loss calculations

- Short-tenn losses
a) Lasses due to friction take recommended values oE spans 1-2 and 4-5 total drape
p = 0.2 El = 0.0085 radslm

= 167.65mm

spans 2-3 and 3-4 Iota1 drape

= 230mm

jacking force = 185.85 kN

(See Figure A26)

PI p 2
p 3

jacking force
185 85

169 37e62x7.XMJJ71 a w l +

e4-2X7.&W%+0mhn

b)

Losses due m wedge set

Take A = 6mm

therefore SP, C)

40.81 kN

Elastic losses

Elastic losses are less than 1% and are ignored in the calculations. Prestress a transfer l
Prestress force at 1 Prestress force at 2 Prestress force ar 3

- Long-term losses
a).

Relaxation of steel
3.75% as before SP,
Spa

SP,

= = =

5.44 kN 6.09 kN 5 6 kN .6

b)

Shrinkage of concrete
SP,

8.78 kN as bsfore

c)

Creep of concrete
SP,

h XE,

Prestress ajfter all losses


Prestress force at 1 = 145.06 - 5.44 - 8.78 - 3.41 Prestress force at 2 = 162.47 - 6.09 - 8.78 - 3.41 Prestress force at 3 = 151.70 - 5.66 - 8.78 - 3.41

Figure AZ6: Force profile.

Short- and long-term losses


average short-term loss average long-tern loss
= =

(21.9%+ 12.6%+ 18.8%)/3

(31.4% + 22.4%+ 28.0%)/3

1.% 78
27.3%

Initial assumptions of 15%and 25% losses respectively are reasonably accurate and no further calculations are considered necessary.
M.13 ServiceabiNry un-tensioned reinforcement calcuIntions
Transverse direction

Bottom steel from tensile stresses at transfer.

h
therefore:
X

350mm and b = 18Wmm

59.15rnm

Top steel from tensile stresses after all losses.

Span AB

Bottom steel from tensile stress at service.

As the ribs will contain links, two bars will be needed in the lop and bottom of each rib although the actual amount of reinforcement was calculated to be much less.

LongitUdi~l direction
No un-tensioned ~einforcement required at serviceability in this d i i d o n . is

A2.2.I Transverse direction


Applied moments: The simpiified method of analysis is assumed using the single load case given in BS8110 Pt.1 clause 3.5.2.3 with appropriate moment redistribution.

M.
where:

d.
X

fm
f,

fm

%,
I
at suppom,

1800 mm

(lotd rib width) (flange width)

in spans,

72W mm

Table A29:

Ultimate capacity due to preshessing tendons over full panel width (7.2m).

Calculate the secondary moments using method B to obtain the applied moment at Ultimate Limit Slate.

Table A30:

Secondary moments for the transverse direction.

Compare the moment of resistance of the tendons with the total applied moment to see wheffier un-tensioned reinforcement is required

Moment from ULS OtNm)

Secondary moment

Wm)
115.7 136.6 157.5

Applied moment OtNm)

Redistributed M OrNm)

MU
(kNm)

A
AB

-315.2 439.6 -858.0

-199.5 576.2 -703.4

-159.6 691.4 -560.4

-386.0 590.1 -556.1

Table A31: Comparison of applied and ultimate moments. Reinforcement is required in span and at B as M has been exceeded. ,

A2.22 Longitudinal direction

$.
f-

$.
1

b
%'$.

= 1770 N/mmz = 40 N/mm2 = 929.27 N/mmz = 1650 mm' = 28800 mm = 1500 mm at supports = 2508 mm at midspan = 0.525

d (mm)
1

!i & n

Td b ,

'~b

(NImm3

x (mm)

&
(mm)

MU @Nm)

196 290 290

0.25 0.10 0.17

13Z.30 1539.90 1493.70

9210 63.80 92.80

41.50 287.00 41.80

376 3. 639 6. 611.6

1-2 2

Table A32:

Ultimate moment capacity due to prestressing tendons over the full panel width (15m).

The value of M for support 3 is the same as for suppon 2,and the value of M, for span 2-3 , is the same as that for span 1-2.This is because the values of d and b do not vary at these

pit. ons

1
secondary moment Wm) Table A33:

2 4. 76

118.8

62.6

73.8

Secondary moments for the longitudinal direction


lo

Compare the applied moment reinforcement is required.

the moment of resistance

lo

check whetha un-tensioned

Table A34:

Comparison of applied and ultimate moments.

. Un-tensioned reinforcement required at suppons 1 and 2,and therefore 4 and 5


No moment redistribution has been applied.

A223

UUimntc un-tensioned reinforcement calculorions

Transverse direction.
Span AB

As k c k,compression reinfmement not required. '


z
d
z

= =

d(0.5

+ $0.25 - 0.028fl.9))

5 0.95d

290 mm 290 x 0.97 mm

Therefore
z

0.95d = 2 7 5 5 mm

681.1 mm2

Therefore

20% redistribution

As k c k, compression reinforcemenL not required. '

A,

42.1 mm'

Therefore

A,

= .

42.1
6

7.02 mm2 per rib

Longitudinal direction.
Support I

As k > k, compression reinforcement is required. '


z d
z
=

= = =

d(0.5 + d(0.25 - k'l0.9)) 196 mm and d'= 35 mm 152.3 mm


@ k7f bd' 0,87fy(d d')

4'

4'

394 mm'

Support 2

As k . k', compression reinforcement not required. :

A23

Mhi~num un-tensioned reinporceme~i requirements Transverse direction (see section 6.10.6)

at For unbonded tendons, minimum reinforcement is required in accordance with BS8110, P r 1, Table 3.27, Figures 3.24 and 3.2514'.
From Table 3.27 At supports: flange in tension

A,

0.26%b.h

Therefore, A,=

1638

= 273 mm2 oer rib

6 In spans: Flange in compression

Therefore, A,

0.18% x b, x h

Therefore, A, = 1134 6

Longihrdinal direelion

For bonded tendons, no minimum un-tensioned reinforcement is required sez Section 6.10.6.

A24

S u n a m of prestress and un-tensioned reinfommenl nquirements

A2.4.1 Prestress Summary


Jacking force 185.5 irN per tendon.
Transverse direction

Two tendons required in each rib


LangiNdiml direction 11 tendons required lhroughout band beam

A2.42 Un-tensioned Reinforcement Summary

Table A35: Transverse direction. Areas and bars given per rib.

Table A36: Longitudinal direction. Areas and bars given per beam. For calculation of anchorage bursting reinforcement, see Appendix E, example 2.

APPENDIX B: Calculation of Prestress Losses

Friction Losses in the Tendon


Friction losses can be calculated in accordance with B.58110"'. However, iris the view of the Working P r y that the foilowing calculations an more realistic. line losses are due t the at o friction resulting Fmm the change in angle of the tendon and unintentional 'wobble' in the tendon. Both effecfecrs an considered in the common formula for Friction, viz.

where: P ,

Po P a
tis

force at distance x from stressing end = stressing force (at anchor) = friction caefficient = angle change in tendon from anchor to point considered (radians) = 'wobble' factor (radiis/m)

This is equivalent to equations 58 and 59 in BS8110, P r 1, Clause 4.9". at


The values of the Friction coefficient will depend on the preshessing system chosen and. in the case of bonded systems, the state of the suand in terms of rust filmm. In the absence of detailed information on friction coefficient and wobble factor from the prestress system supplier, it is recommended that the factors in Table B1 are used:

Unbonded tendons friction coefficient p wobble rn (radlm) 0.06 0.05

Bonded tendons 0.20

0.0085 .

Table B1: Typical friction coefficients and wobble factors. For slab type structures with unbonded tendons it is normally reasonable m assume a uniform angle change per unit length. This angle change can be obtained by calculating the total angle turned thmugh over the full length of the tendon and dividing by the full tendon length. Alternatively a simple method based on the typical drape and span can be used. Figure B1 illushates the geometry for a typical pmbolic tendon with a reverse parabola at the support The tangent to the curve at the point of inflection extends through poinrs 'c' and 'a'.

Figure B1:

Typical geometry of tendon prolile for internal span. =


tan" ((2 x total drape)/A)

Thus the slope 9.

Similarly, using points 'b' and 'c', slope 8, can bs obtained Over the span L the total deviated angle = 2(6,+eJ The average deviated angle per unit length, a',is therefore

On the assumption that paint 'c' is at the centre of the span, this may be simplified to:

a = '

1 x total d r a ~ e 6

L1
In such cases, the friction formula may be written as:
p
=

pOe'Wd-)

The prestress force profile after friction losses can now be drawn.

Wedge set or draw-in


Most post-tensioning systems used in buildings depend on a wedge-based system for anchoring. In order For the wedges to grip, there must be a small movement of the strand into the anchorages. This inward movement reduces the prestress and the amount of movement depends on the pnrticular prestressing system employed; a typical value is 6 mm. The draw-in effect is as shown in Figure B2.

Figure 82: Loss of prestrw due to wedge draw-in 'Ihe force loss is calculated as follows:

where: A

S. P

= wedge draw-in = force loss

EP AP
1'

= modulus of elasticity of tendon = area of tendon = length of tendon affected by draw-in

If it can be assumed that the tendon has a uniform angle change per unit length, then the force profile is appmximately linear. Consequently, if 1' is less than the length of the tendons, then:

where: p' = slope of the force profile and

S. at anchorage = 2 x p' x I' P


The force loss, within the length 1'. is then given by:

If the wedge draw-in affects the whole length of the tendon, then:

SP, S, P

at s w i n g = (A x E, x A&+ anchorage
at

p'x 1

dead end =

(A x

E .x +/I

- p):

Elastic Shorlening of the Structure


As strands are tensioned, the structure will shorten elastically. In most building flwrs, this shortening is insignificant in terms of losses, but it may be significantin highly stressed beams. The force loss is given by:

where: &

= 0.5xf,

%I
f,

E, = ,

the stress in the concrete adjacent to the tendon after uansfer. the modulus of elasticity of the concrete at the time of uansfer.

In Ihe formula for E, given above, the factor of 0.5 rakes account of the averaging effect of several tendons stressed sequentially (BS8110, P r 1, Clause 4.8.3"'). If this is not the case. at this factor may have to be modified.

Shrinkage of the Concrete


BS8llO'" covers this subject extensively in Section 4.8.4 of P r 1 and Section 7.4 of P r 2. at at Special care should be taken in thin members (e.g, slabs) subjected to low humidity (such as in some.healed buildings) when shrinkages of more than 400 x lod can occur. The force loss is given by:

SP* =
where: &

E S h XEp X

A,

shrinkage sbain of concrete

Creep of Concrete
Creep loss is based on the stress in the concrete at the level of the tendons. These losses are extensively covered by BS8110, Part 1,Section 4.8.5 and Part2,Section 7.3"'. They can have a very large effect in highly suessed thinner members. The force loss is given by:

where:

E ,

f, X $ E, .
the creep cefficient (BS8110, P r 2, Figure 7.1")). at

For ribbed shuctwes, an effective thickness should be obtained from the ratio of volume to surface area

ReInxation of lhe tendons


The stress in the tendons reduces with time because of the relaxation of the steel. The amount of relaxation depends on the type of strand and the initial smss. Figure B3 illustrates typical relaxation curves for various types of svand and load levels. A

- -2.6
0 0
1
I

-2.8 -2.4 -2.2


-2.0

/'

I-

I
0

-1.8 -1.4

a? - ----- -1.0
1.2
k

LL -1.6

,'
/

. '

. '

---#(

-0.8 0.6

- 0.4

------- -----

-/'

/ . '

Relation between 1000 hr relaxation and initial load at 20'C.

-15

-14 -13

rslaxstlon and lnltlal load ar

4 @ ~ temperatures

___---1
55

fj0

60

65

70

75

80

85

I N I T I A L LOAD ( % O F ACTUAL BREAKING LOAD1

Figure B3:

Relaxation curves lor types of strand a t various load levels

The force loss is given by:

SP, = 1000-hour relaxation value x relaxation factor


x the pnslress force at m s f e r
The 1000-hour relaxation value is given in BS5896(lq and the relaxation factor in BS8110, Paq 1. Clause 4.8.2'". Class 2, low relaxation s%l. is normally used in buildings. Data for the relaxation of this type of steel are given in Table B2. 138

Force at transfer as a % of characteristic strength of tendon


80% 70%

1000-hour relaxation
4.5% 2.5% 1.0%

Relaxation factor
1.5 1.5 1.5

Force loss as a 5 of % force at transfer


6.75% 3.75% 1.50%

60%

Table BZ: Relaxation for CIass 2 low-relaxation steel Note:


1. 2.

Characteristic sneogth of tendon = f, x

The 1000-how relaxation values from BS 5896@"given in Table B2 above can be replaced with the manufacturer's values if available.

APPENDIX C: Calculation of Tendon Geometry

Figure C1: Tendon geometry

Consider the Uurz parabolas AB, BCD and DE.


Parabola

AB

=
=

k,xz
k 2

BCD DE
For parabola AB,

kSZ

- a,
Similarly for DE,

~IP?

- a.
Let

=
=

k ( pJZ z-

QI

ql-42

and

QZ

e-%

Then for parabola BCD,

and

The slop-, of the parabolas a any point is dyldx, and the pafabolas are tangential at B and D. t

For parabola AB,

similarly for DE,

As the parabolas are tangential at B and D. the slopes of h e two parabolas which meet at each of these points will be equal. For parabola BCD,

and

Using these equations it is possible to obtain expressions for k, and k, in terms of k.

and

Substitute the values of k, and k, into the original equations for parabolas AB and DE. Therefore:
and

al

=
=

k l - PI) pv
kp,(L-L'-pJ

Substitute the values of c, and c,into !he original equations for parabola BCD. Therefore: Q, - kp~(L' PI) and

kW - PI)'

Solving for k in each case

and

These equivalents rationalise U, give the quadratic:

where:

with the solution

L '

- m k Jfrn2-4111)
21

' Once L has been calculated, a, and a, can k found using:

and

Figure C2: Solution for the transverse direction of Example A l .

For the case shown:

APPENP)IX D: Calculation of Secondary Effects Using Equivalent Laads


Equivalent loads can be used to represent the forces from prestress. These will automatically generate the combined primary and secondary effects when applied to the structure. Figure Dl shows the commonly occurring equivalent loads for typical prestress situations.

Parabolic drape

Figure D : Commonly occurring equivalent loads l One method of separating the secondary fmm the primary effects is t use a frame analysis n with the equivalent prestress load acting alone. The resultant moment and shear diagrams include both the primary and secondary effects. In order to obtain the secondary effects, it is only necessary to consider the moments and forces at h e supports and subhact the primary effects from them. The secondary moments along each span vary linearly from end to end. This method will be known as method A. To illustrate m e w A. the U t m t Limit State for the uansverse direction in Example A1 liae of Appendix A is used and the secondary effects obtained as follows: 1 . Calculate the equivalent prestress loads in the spms using a load factor of 1.0.

Figure DZ: Equivalent balanced loads

143

2.

Analyse the structure and obtain the bending moment diagram.

a) slab

b) columns

Figure D3: Moments due to primary and secondary effects


3.

Calculate the primary moments due to prestress (Pe) in the slab at each support. There are no primary moments in the columns. At support C, At support B(C), At support B(A), At support A, Pe = 0 Pe = -172 irNm Pe = -172 kNm Pe = 0

4.

Subtract the primary moments fmm step (2). At this srage it should be noted that the moments and reactions in the columns from the b e analysis are due entirely to secondary offecls.

Figure D4: Bending moment diagram due to s c o n d a r y effects

Figure D5: Shear force diagram due to secondary effects

An alternative method of calculating secondary effects is detailed below. This will k known as method B.
As there are no primary presness forces in the columns, the column moments and reactions are entirely due to secondary effects. So the secondary effects in the slab can be easily obtained by applying these column reactions and moments to the slab as shown in Figure D6.

Figure D : Column reactions and moments due to secondary forces 6 This results in the secandary moments and shears in the slab as shown in Figures D and D5. 4

APPENDIX E: Calculation and Detailing of Anchorage Bursting Reinforcement


In this Appendix two examples requiring bursting reinforcement are considered. For each the un-tensioned reinforcement requirement is calculated and its position in the slab derailed. Note: The first example refm to lendons in Design Example 1 of Appendix A and the second example refers to the beam tendons in Design Example 2 of that appendix.

Example 1
Depending on the tendon layout chosen from the calculations of Design Example 1 in Appendix A, anchorages will be in groups of 1.2. 3 or 4. The following example is for a p u p of 4 tendons of 12.9mm snands (unbonded) in a 225rnm Uiiik slab, as shown in Figure El.
A group of unbonded anchorages for four 12.9mm strands in a 225mm deep slab, as show

in Figure El.

Figure El: Anchorage layout for Example 1. Characteristic strength of the tendon
= 186 kN

In the y-y direction,

therefore,

y,,jym =

058

Fmm Table 47" .'

a)

Figute E4 shows how the end block can be divided into individual end blocks or prisms for each anchorage. These must be rectangular and symmehical.

Prisms for anchorages A and B are 125 deep x 700 wide. The prism for anchorage C is 125 deep x 1500 wide. Anchorages A and B

x-x direction

The jacking force per saand is 185.85 kN Hence P , = 4 x 185.85 = 743.4 WII From BS 8110"),Table 47 .

Hence F, = 0.203 X 743.4 = 150.9 kN and the reinforcement required at the allowable stress of 200 N/mm2 is
A, = 150.9 x 1@/200 = 7 4 5 mmz 5. positioned between 0.2% to 2.0% i.e. 70 lo 700mm from anchor.
;

Similar calculations for the y-y direction of anchorages A and B and both directions of anchorages C (three shands only) yield the following reinforcemenrs: Anchorage A and B y-y direction

Anchorage C x-x direction

y-y direction

b) Overall stability in y-y d i r e c t i ~ n ' ~ '

flange 110

MOMENTS = 3.215 N/mrn2

SHEAR

Figure E4: End block moments and forces: y-y direction.


Because anchorages forces are increased evenly as explained above, and the anchorages are IocaIed on the c.g.c., the stress block behind the anchorages is uniform and equal to

(0.6359 x 1'@is the area of the section fmm Example A2)


The moments can be calculated thus:

Hence, for the maximum moment of 57.68 kNm


and a lever arm of 'h block length (= 175 mm),

steel requited A, =

$7.86

0.175

0.200

dishibuted over distance of 175mm IO 35Chnm ftom the anchorage faces. Fmm reference 21, minimum steel

= 0.3% x 1500 x 350

1575 mm', which is satisfactory.

Similarly in the r-x direction

wMA= M ,
=

MOMENTS

SHEAR

~21.5 mm2 *N

Figure E5: End block moments and forces: x-x direction.


3.215 x 110 x 504 x (400 + 504D) X +3.215~350~4001x%x10"

lo6
206.2 kNm

Hence,

Minimum steel = 0.3% x 350 x 1500 dishibuted over distance of 750 to 1500mm born the anchorage faces

1575 mm2

Note: The above moments are slightly oversrated since the anchorage force h been assumed s (conservatively) to be a point load Flow of mess into flangem' Load in flange = 3.215 x 2508 X 110 X 10"
Width of web

= 1500 mm

From BS 8110'" Ffi0 = 0.14

therefore F = 124 kN , Required a e of steel = 620 m d diihibuted over distance 250 lo 2500 mm from the ra anchorage faces. Check on horizontal shear capacity. From Figure E4 , maximum shear force giving a shear s m s of 935.2 x 1@

= 935.2 kN = 1.78 N/mm2


? 2.25 N/mmZ therefore fme.

1503 X 350
Check on vertical hear capacity. Fmm Figure E5, maximum shear force giving a shear stress of 572.1 x 1 C ' = 5721kN = 10 N/rnmz .9 ? 225 NImm2 therefore fine.

1503 x 350
In the flange area, maximum shear force giving a shear stress of 1782 x 10.' 1500 x 110
=

178.2 kN = 1.08 N/mmz therefore me.

The reinforcement layout given in Figure E6 satisfies a l l the preceding bursting and end-blcck stability requirements.

Anchorages

50-

6 T 10 Legs at 150

- --

Section A-A
Fiwe E : Layout OF end block reinforcement. 6

APPENDIX F: Simplified Shear Check: Derivation of Figures 17 and 18

Assumptions:
1.

Charts are drawn for internal columns

2.

v,

vc::uxd/1000inkN where v, is shear resistance of lhe concrete (N/mml) u is lhe lengh of the critical perimeter (mm) d is the effective depth (mm)
= 0.9h where h is lhe deplh of slab

3.
4.

Columns are square of dimension c

6.

Loading is uniformly dislributed.

U t Load = 1.5 x (Char. Dead Load+Char. Total Imp. Load) l


where Char. Total Imp. Load, Q, = Live LoadcFinishes 7. 8. Concrete density = 24 icN/m3 Applied shear force V = 1.5 A (24W100a + Q,) in kN where A is ihe appropriate area of floor in m2

Check atfirst critical perimeter (Figure 17)

v
Therelore

1.5(24h/lO + Q,) x A < 3.6hv,(c + 2.7h)IlWO Q, 5 12v&(c + 2.7h)/50aOA - 24h11000 kN/m2

SV'

Check at face of column (Figure 18)


Assume fa = 40 N/mm2 Maximum design shear strength

= 5 N/mm2 or 0.8 df, N/mm2 = 5 N/mmz

1.5A (24h/lOW + Q ) <_ l R c h / l ~ ,

APPENDIX G: Vibration of Post-tensioned Concrete Floors


The SCI publication 'Design Guide on the Vibration of FloorscNwas specifically written for composite steeVconcrete flows. Post-tensioned flat slabs usually have greater mass and a lower fundamental natural frequency than slabs on profiled metal decking. Although increasing the mass of the floor reduces the dynamic response, floors of lower natural frequency ate excited by much larger components of the walking force. It is therefore likely that dynamic response will control the thickness of some flat slabs. Most of the SCI guide is equally relevant to other forms of Boor construction. The main difficulty in using the guide for flat slabs is estimating their effective widths and spans. The values for these dimensions are given in rerms of beam and slab stiffnesses. The values cannot easily be modified for the various configurations of flat slabs. The design procedure given in the SCI guide only considers the fundamental frequency of the floor. However only pan of the mass of the floor is considered as mkhg part in the response. This area is defined by the effective width and span. An alternative approach is to consider all the natural frequencies and the total mass of the floor. In practice, only a limited number of natural frequencies need t be considered because the contribution to the total response from a the higher Frequencies is minimal. The procedure given below uses this second approach. The dynamic response of the fundamental frequency is calculated. This is then multiplied by a factor to give the total dynamic response. There are problems wirh the SCI guide in the uansition between high- and low-frequency floors. There can sometimes bz a jump in the calculated response of a factor of ten. The pracedure below gives functions which lead to a stepless transition. It also gives expressions for estimating the fundamental natural frequency of flat slabs, including the caes where there q e no perimeter beams. The dynamic responses of floors have been checked to the deflection spanldepth ratios determined by deflection considerations. For normal office space, the responses fall within the acceptance criteria, provided additional requirements on the minimum number of bays and slab thicknesses are met. Table 1 includes these additional requirements.

Design procedure for checking vibmtion response of past-tensioned concreteflaom.

n
I

4
I

-3
r I

Figure G : I Two families of vibration modes are considered. This is conservative because some of the higher harmonics of the N o families are in common

(a) x family Figure G2: Families or vibration modes


x family

y family Effective aspect ratio = 74

Effective aspect ratio = ,?

EI,is flexural stiffness of slab spanning in the x direction per unit width (Nm%)
EL, is flexural stiffness of slab spanning in the y direction par unit width (Nrnz/m)

k, and k, depend on the effmtive aspect ratio. m is mass per unit area (r/m2) (dead load + 10% live load)
Let
"I" represent

"x" or "y" as appropriate

1 k = I + ,

for solid or coffered slabs for ribbed slabs

)2

k,=

I + -1

For slabs wirh perimeter b a s : fr = $ For slabs without perimeter beams: for solid or coffered slabs

Note: fb is natural frequency of a single panel supported at its four corners

f , = f:

- (f: - f,)

(lin= linv)

(3)

8 = li(0.5

+O.linolt,

for solid or coffered slabs for ribbed slabs

= 1+(0.65 + O . l h ( ) A ,

5 is critical damping ratio = 0.02 for open plan offices

C, = 30

20Hz S f ,

Where R -< 4 Special office R -< 8 General ofice R 5 12 Busy office

APPENDIX H: Advertisements

Page No.
Arup Oasys
160

CCL Systems Limited


PSC Freyssinet Limited

161
162

ADPRSL is part o f the Oaryr ADC suite o f computer pmgrams for the analysis and design o f reinforced and prestressed concrete elemenn ADPRSL addresses pat-tensioned flat slabs with bnded M unbmula? t&lr %program was wi to follow rhe requirements o f BS 81 10 Part 1 and Pan 2, Gmcrere Sonety im Technical Repom No 17 and 25, and the draft version of a new Concrete Sociery D e e Manual on Post-tenshd C o n m e FZwr Design

% manual of continuous slabs inpost-tensioned concrete can involve hm of arduous e, f fto~determine an efficient tmdon orofile and rn , prestrksingfme ADPRSL automates pmcess to*odxe higm that are economic and efficient

K i s

Numenius options have been mcorparated into the @gram to meet the needs of a wid^ variety of design problems Following analysis the user may explore orherprestress and balanced load combinatim quickly.

ADPRSL will also pmvide information on quantities of matm'als used so aumuing comparirim w'd~ fm mnmuclion orher of
%program ir currentlyfor singkwaY annlysis and and the urer must consider bod1 frame direnim of the sirware separately. ADPRSL uses the 'Loadh h m ' n g Metho* rn carry out the design of rectangular slabs and Teesections, for oneway and t w ~ w a spanning floors y

Up m 10 spans including md cantileup~s be considered and rectanghr may drops at columns can be simulated
Columns are considered m fired inpsition and in direction m their remore ends. The s m m e is analysed w'ng d~ Equivalent h e Method In this method rhe muaure is analysed as a frame in me directin& with the slab considerzd to be a continuous mip o f width equal to the panel width
e

lhe mr of equivaht lo& automatically t& into acrmmr the serumby u e @arasitic)e i f e m the stmcme. ~

% initial input infaMrirm required consists o f the slab and c o h n geomehy and design loads An uIrimnfe analysis is then cam'ed our without redistribution After data checking ADPRSL requests tendon data, global loss faaol~,deflection and limits, mareiialpmpememes balanced load h a i L and then cam'es our the prestressed analysis and a'esignwherever relevant dejault values are pwuided Serviceability and Ulzimare Limit State ( U moments can be reduced to the U) column face $ requird and deud and live loadperenrages setfor the tr&er condition it is ako possible to individualizeprestress fmes in any spans and set a minimwn premess levef

160

w s && Hmrrq Ail . n %u N m m l e LIpl T y e NEl ZEB opl.:Si i c i g Tr~7MnrO91-2Sim9 .0

.w

CCL SystemsLimited- In we wayday worldwide

Quality and Service

CCL Systems are recognised worldwide as originators in the field of prestressed concrete construction. The company's wealth of experience in this highly specialised area - covering more than 30 years - is fully reflected in CCL Spanforce, the state-of-the-art bonded and unbonded systems for post-tensioned buildings. Major advantages of the CCL Spanforce system include simplicity of installation and uncongested detailing, leading to easier and faster concreting. To obtain the maximum benefits from the system, however, it is necessary for Spanforce to be incorporated into building detail of initial design stage. CCL Systems is a BSI registered company, BS5750 Part 1 (IS0 9001 - EN 29001). Our quality assurance system is based on these international standards and is designed to ensure that all operations within the company are controlled, ensuring that total confidence in our ability to satisfy customer needs is maintained. For further details on Spanforce, or for assistance with any of our other product ranges, please contact: CCL Systems Limited, Cabco House, 201 Elland.*... . . Leeds-LSll 8BH, We& Yorkshire. Road. . . . ... . .

'

From design to completion a total service from

PSC FREUSSINET
post-tensioned systems for buildings
Britain's leading authority on the design and construction of post-tensioned suspended, ground (including piled) and roof slabs.
PSC Freyssinet specialise in providing designers and contractors with a fully comprehensive service - literally from design to finished installation. PSC Freyssinet's service includes:
e

Preliminary scheme design and cost estimate. We provide a free scheme desigdbudget cost service which can be invaluable in comparing the relative costs of post-tensioning versus traditional R.C and steel frame options. Final design and shoplworking drawings. We can undertake the production of final design calculations and working drawings. Alternatively we can produce shop drawings from consultants' design sketches.

Supply and installation of post-tensioning and associated reinforcement. Stressing. Grouting for bonded tendons. Concreting of high tolerance post-tensioned floor slabs.

e
e

THE CONCRETE SOCIETY


Founded in 1966 Thc Concrete Society brings together all those with a n interest in concrete to promote excellence in its design, construclion and appearance, to encourage new ideas and imovations and to cxchange knowledge and experience across all disciplines.

TECHNICAL DEVELOPMENT CENTRE


The Concrele Society is a ccnlrc of cxccllencc for technical development O F concrctc, producing smtc-or-lhc-arl reports, recommendalions and pntctical guidcs, and iniliat ing and undcrraking R&D whcrc appropriate. The work of the Centre is under lhc dircction of lhe Technical Development Board, which includcs senior personnel from all sides of the indushy. The Technical Management Committee and its supporting specialist p o u p s and working parties carry out the work of [he Cenue, with ihe help of the Technical Manager and his staff. The Centrc collaborates with other overseas Concrcte Strictics in mutually beneficial progrnmmcs of lechnic;tl tlc\~clupmcnl. C O N F E R E N C E S AND EXHIRITIONS The Society organises nadonul and intcmutianal conrcrcnces and exhibitions including DTI-supporterl loin1 Venture cxhibitions at m.djor international cvents wnrldwide. The Society's annual 'Concrcte Day' has become a mdor national event in thc calcndnr of the construction industry.

CONCRETE ADVISORY SERVICE


Thc Concrctc Advisory Service provides prompt impartial lcchnical advicc on concrete and related mauers lo subscribing Members of any discipline. The regionally based advisory staff are all Chartered Engineers and have wide experience in all aspecls of the use of concrele and of providing appropriate advice. Any member of slaff of a subscribing group Member can tap into this vast reservoir of information on materials, technology and practice. In addition L telephone advice, o visits lo Member's offices or sites to discuss technical mattcrs or solve problems can also be arranged. Reports can bc providcd where applicable. AWARDS Awards which ensure that excellence in concrete is publicly acknowledged are made lor completed suuctures in building, civil engineering and mature structures. MEMBERSHIP Thcre nre two kinds ofMcmbcrship Group Membership for firms, partnerships, government depnrtments, k~cal authorilies, cducutional establishments elc. Pertionol Membership giving Concrete Society benefits to individuuls. E N T I T L E M E N T T O USE DESIGNATORY L E T T E R S Mcmbcrs who havc appropriats knowledge and experience may apply for tho qualifying grades 'Member' (MCS) and 'Fellow' (FCS).

REGIONS B;

cLuns

The Sociely is organised into 22 regions and clubs which each arrange a comprehensive prngrnmme of technical and social events in their areas. As wcll as all thc lcchnical hcncfirs, the Sanely provides M ideal sncittl forum for tncmhcrs tu mnke valuable business and pcnonal cunldcls. CONCRETE Members in all categories arc scnt a copy of each issue ol CONCRETE, the Sociely's journal, providing up-to-date inlormalion on all aspects of concrete including dcsign, mnterials, construction techniqucs, quality control, cquipmont, mainrenance and repair.

R E G I O N A L ADVISORY ENGINEERS N O R T H OF E N G L A N D &SCOTLAND Deryli Sinrpsc~n Bolton - Tel: ((I 1912) R 15516 Fax: (01942) 842533 doltn Plimrner York- Tel: (01937) 834827 S O U T H OF E N G L A N D , WALES & N O R T H E R N IRELAND Maldywn Enoch Cardill - Tel: (01443) 237210 Fax: (01443) 237271 George nurnbrook Romscy - Tel: (01791) 524455 Fax: (01794) 321325 Dick Rr~berts Guildford - Tcl: (01428) 725269 Fax: (01428) 727891

APPEhTP)M E: Calculation and Detailing of Anchorage Bursting Reinforcement


In this Appendix two examples requiing bursting reinforcement are considered. For each the un-tensioned reinforcement requirement is calculated and its position in the slab detailed. Note: The first example refers to tendons in Design Example 1 of Appendix A and the second example refers to the beam tendons in Design Example 2 of that appendix.

Depending on the tendon layout chosen from the calculations of Design Example 1 in Appendix A, anchorages will be in gmups of 1.2. 3 or 4. The following example is for a gmup of 4 tendons of 12.9mm sbands (unbonded) in a 225rnm thiik slab, as shown in Figure El. A group of unbonded anchorages for four 12.9mm strands in a 225mm deep slab. as shown in Figure El.

Figure El: Anchorage layout for Example 1 . Characteristic strength of the tendon = 186 kN

In the y-y direction,

y. therefore,

= 22512 0.58

= 112.5 rnrn

y a m=

From Table 4.7'"

Therefore, at Serviceability Limit State with reinforcement acting at a suess of 200 N/mmz,

At Ultimate Limit State, A,

= 186 x 0.146 x 10' 0.87 x 4M)

68 mma/anchorage (use 2T10)

This un-tensioned reinforcement should be between 22.5 and 225 mm from h e anchor-bearing
face.

In the x-x direction,

x .
Therefore, From Table 4.7'6

= unlimited (say 15M) mm)

= 120kN
Therefore.

600 mm2(use 2 x 4T12)

This un-tensioned reinforcement should be between 150 and 1500 mm from the anchorage bearing face. Figure F2 shows the practical detailing of these requirements.

Section A A

Figure E2: Bursting reinforcement distribution for Example 1. Comment It is not usually required to do an equilibrium study for flat plates w t regularly ih spaced tendons, provided they are stressed in such a sequence as t avoid problems at coners. o

The ribs in Example 2 of Appendix A have unbonded anchorages and their bursting design will be similar to Example 1 above. The design of the anchorage bursting reinfonement for the banded tendons in the beams is outlined below. The design requires eleven 15.7 mm banded strands in the beam. It is decided to use two tendons of four strands each and one of h w shands, with the anchorages arranged as in Figure E3.

u
1500

prism 3 125 x 1500

Figure E3: Anchorage layout for Example 2.


f i e anchorages are positioned so that the c.g.s of the tendons corresponds to the cenm of gmvity of the concrete (c.g.c.); from the calculations in Example 2, this is 196mm above the beam soffit Hence, assuming the anangement in Figure 3,

hence

A = 105 mm.

until all For the stressing sequence, it is assumed that one strand in each tendon is sir& strands in aIl three tendons have been stressed (i.e. if the strands are numbered 1-11 as shown in Figure 3 the stressing sequence would be 1, 5 , 9 , 4 , 8, 11, 2.6, 10, 3 and 7). In this way, there is no need to consider in~ermediitestages and it is likely to give the least amount of bursting reinforcement In order t check the end block fully, two individual checks are required, namely: o a) Single anchorage bursting b) End block stability

a)

Figure E4 shows how the end block can be divided into individual end blocks or prisns for each anchorage. These must &z reztanguiar and symmerrical.

Prisms for anchorages A and B are 125 deep x 700 wide. The prism for anchorage C is 125 deep x 1500 wide. Anchorages A and B

x-x direction

The jacking force per m m d is 185.85 kN Hence P , = 4 x 185.85 = 743.4 kN From BS 8110", Table 4.7

Hence , F = 0.203 x 743.4 = 150.9 kN and the reinforcement required at the allowable seess of 200 N h m * is
A, = 1 0 9 x 1ff/200 = 7 4 5 mml 5. 5. positioned between 0.2% to 2.0% i.e. 70 w 700mm from anchor. ,

Similar calculations for the y-y direction of anchorages A and B and both directions of anchorages C (three shands only) yield the following reinforcements: Anchorage A and B y-y direction

Anchorage C x-x direction

y-y direction

b Overall stability in y-y dire~tion'~' )

flange 110

MOMENTS

SHEAR

= 3.215 N/mm2

Figure E4: End block moments and forces: y-y direction. Because anchorages forces are increased evenly as explained above, and the anchorages are located on the c.g.c., the swss block behind the anchorages is uniform and equal t o

(0.6359 x 1D6 is the area of the section from Example A2)


'Ihe moments can be calculated thus:

Hence, for the maximum moment of 57.68 W m

and a lever a m of 'h block length (= 175 mm), r


s1 required A, =

57.86
0.175
X

0.200

dislribuled over distance of 175mm KI 3 5 h m from the anchorage faces. From reference 21, minimum see1

1575 mm', which is satisfactory.

Similarly in the x-x direction

w-

MOMENTS = 1.215 *N mm2

SHEAR

Figure E5: End block moments and forces: x.x direction.

Hence,

A,

= 234.9/(0.75

0.200)

Minimum steel = 0.3% x 350 x 1500


distributed over distance of 750 to 1500mm Erom the anchorage faces

1575 mm2

Note: The above moments are slightly overstated since the anchorage force has been assumed (conservatively) lo be a point load Flow of stress into flange"" h a d in flange = 3.215 x 2508 x 110 X Width of web
=

lom3

1500 mm

From BS 8 1101" F a = 0.14