You are on page 1of 18

SYNOPSIS

The recent limit state code for concrete bridges IRe: 112 have 4 sections (Sections 8 to 11) on ultimate limit states. The
sections on Ultimate limit states oflinear elements and shear including torsion have been well synthesized by T Viswanathan
in IRe Journal January-March 2014) and ING-IABSE journal Vol. 44 No.2 June 2014 respectively for the users. The author
had a good fortune to write explanatory notes on sections ULS of induced deformations and Serviceability Limit States
(SLS) in the Explanatory Handbook already brought out as IRe: SP-lOS-20lS. Thus the onus of completing the synthesis
of balance Ultimate limit state on i.e. induced deformation clearly lies with the author as a continued tradition. Most of
the provisions in IRe: 112 for ULS of induced deformations are similar to Euro code 2 which were based on documents
prepared as a part of project team for EN-1992-l-1(2004). This was later published as 'Euorocode 2 commentary' by
European concrete platform ASBL. In the detailed paper below, the most of the diagrams and background deliberations are
taken from these background documents.

1. INTRODUCTION structural members are vulnerable The word buckling is meant


to sudden failure due to instability for the "pure", hypothetical
By virtue of its high compressive
which is generally termed as buckling of an initially straight
strength, concrete was considered
buckling failure. member or structure, without
in the beginning as building
material for massive and sturdy eccentricities
structures. With the advancement
of Concrete technology and also curvature
~
very high increase in strength,
concrete structural members
are becoming leaner, slimmer planned position
position with imperfections
and slender. Unfortunately with deformed position
the increase in strength of the
concrete, proportionately modulus
of elasticity there by stiffness is
not increasing to the same extent.
As such axially loaded slender
Fig. 1 First Order Effects with Axial Load

* Senior Vice President and Member Board of Management, Gammon India Ltd., Mumbai, Email: venkat.heggade@gammonindia.com
t Written comments on this Paper are invited and will be received by the 30'h April, 2016

Journal of the Indian Roads Congress, October-December 2015


HEGGADE ON

load eccentricities or transverse First order effects or deformations for the same load, EI reduces
loading. The pure buckling is are due to transverse/lateral loads due to cracking of concrete and
not a relevant limit state in real and also include the effect of inherent non-linearity in the
structures, due to the presence imperfections, interpreted as concrete stress-strain response
of imperfections, eccentricities physical deviations in the form also increases. Thus it involves
and/or transverse loads. This of inclinations or eccentricities. both geometry and material non-
is also a reason why the word The compression resultants with linearity for reinforced concrete
"buckling" is avoided generally eccentricities and curvature elements and has to be taken in
in the Section of ultimate limit variation due to first order effects to account while choosing the
tate of induced deformation, the are shown in Fig 1. method for 2nd order analysis.
buckling is mentioned only when
a nominal buckling'load is used as Load deformation behavior and The nominal second order
a parameter in certain calculation ultimate capacity of structural moment is used in certain
methods. members and structures is simplified methods, to obtain a
significantly affected by Second total moment used for design of
order effects. cross sections to their ultimate
moment of resistance. It can be
OA - curve for end moment defined as the difference between
OB - curve for maximum the ultimate moment of resistance
column moment (ij' mid- and the first order moment. If
height) the ultimate load is governed by
column
/ Intoracllon diagram instability before reaching the
cross section resistance, then the
Axial capacity is reduced nominal second order moment
from A to B due 10 increase is greater than the true one; this
in maximum moment due to is the reason for using the word
;.rs (slenderness effects) "nominal" .
o Moment

The terms sway - non-sway is not


Fig. 2 Interaction Chart Showing Second Order Effect
used in section 11 of IRe: 112.
The words in themselves are
Second order effects are effects are calculated by second misleading, since all structures are
additional action effects caused order analysis. more or less "sway". A structure
by the interaction of axial forces that would be classified as "sway"
and deflections under load. First Majority of commercially could be just as stiff as one
classified as "non-sway". These
order deflections cause additional available structural soft wares has
terms are in fact now replaced by
moments which in turns lead the capability to carry out second
unbraced - braced.
to further deflections (Fig.2). order analysis. In this analysis,
Sometimes these effects are also in addition to the invalidity of The distinction braced - bracing
called p-~ effects as they are the principle of superposition is simple. The units or systems
the products of axial forces and the flexural rigidity of reinforced that are assumed to contribute to
deflections of the elements or concrete structures EI is not the stabilization of the structure
system. Normally second order constant. As the moment increases are bracing elements, the others

Journal of the Indian Roads Congress, October-December 2015


ULTIMATE LIMIT STATE OF INDUCED DEFORMATIONS: SECTION 11 OF IRe: 112 SVNTHESISED 189

are braced. Bracing units/systems with increasing moment due • :::; 10% increase of the
should be designed so that they, to cracking. corresponding first order
all together, have the necessary effect,
stiffness and resistance to develop The above de-merits introduces
stabilization forces. The braced complexities in 2nd order analysis • :::; 10% reduction of the load
ones, by definition, do not need to as such the code gives two simple capacity, assuming a constant
resist such forces. criteria for not carrying out the 2nd eccentricity of the axial force.
order analysis.
2. CRITERIA FOR The cross sectionresistance (Fig.3)
IGNORING 2 D ORDER • Cases where second order was calculated for rectangular
EFFECTS effects are less than or equal to cross section 400 x 600 mm
10% of the first order effects. concrete C3 5, (j) = 0 .1 (total
Significant disadvantages of mechanical reinforcement ratio),
second order analysis are: • If the slenderness A..is below edge distance of reinforcement 60
a certain value A..lim,i.e., A..<
mm.
• The principle of superposition A..lim(A..limitation criteria).
is not valid in second order During the conversion process,
analysis and all actions must be 2.1 Less than or Equal to 10% the first criterion of 10% less
applied to the bridge together Criterion than or equal to first order effects
with all their respective load Before adopting this 10% criterion, was found to be suitable for the
and combination factors. two different ways of defining following reasons:
10% criterion for ignoring 2nd
• The flexural rigidity (EI) ofthe • In a column or a structure it
order effects were investigated by
reinforced concrete structure is the bending moment that
Bo Westerberg et al
is not constant. EI reduces is influenced by second order
effects.

0,5 I • The axial force is governed


MMNm by vertical loads, and is not
0,45 ......•
/ ~ 2, k-..
<, significantly affected by
0,4
/i " second order effects.
Nu;~;- X% reduction of Nand M
0,35
,'/ ~
2Y..
. "

0,3 hV" ,~
"',,
~O% increase of M • Most design methods are
based on calculating a
0,25
VI ,~
'\
bending moment, including a

0,2
/I 1. Nu: MJ1.1
,
'\
~u;t u second order moment if it is
significant.
If
"
\
2. a.g·Nu;a.g·Mu \
0,15
If \, \. The above criterion is not much of
0,1 ,, a use to practising designers as the
0,05
\
~ 2nd order analysis has to be carried
o
NMN
\,
, 1\ out to determine whether the 2nd
o 0,5 1,5 2 2,5 3 3,5 4 4,5 5 5,5 order effects are with in the 10%
of 151 order effects. In view of this
Fig. 3 Interaction Diagram for two Different 10% Criteria perhaps, the alternative criterion

Journal of the Indian Roads Congress, October-December 2015


HEGGADEON

of A limitation is more useful for After receiving the comments and


designers for ignoring 2nd order examples from Prof Hellesland M oEd = First order B.M. in design
effects and analysis. during the evolving stage in load combination in UL.S.
2001, a systematic investigation Where <P ef defined above is not
2.2 'Alimitation criterion of the slenderness limit was known, 'A' may be taken as 0.7
made, with focus on the effects of which corresponds to <Pef=2.0 that
While determining the second
reinforcement, normal force, creep would be typical of concrete loaded
order effects by simplified
and moment ratio (different end at relatively young age, such that
methods instead of non-linear
moments) and these parameters <P = 2.0 with a loading being
co

second order analysis, the


were added in Alim equation entirely quasi permanent. Using
effective length concept can be
finally. the default value of A = 0.7 is
used to determine slenderness.
reasonably conservative as the
On determination of slenderness, Higher limiting slenderness can same is in any case not sensitive
the requirement of second order be achieved where: to realistic variation of <P
ef.
analysis itself may be deduced.
The slenderness ratio is defined as • there is low creep (because the 2.2.2 The term "B" accounts
'A=1Ii where '1e 'is effective length stiffness of the concrete part for reinforcement ratio:
e
and 'i' is the radius of gyration of of the member in compression
the un cracked concrete section. is then higher). B=.Jl + 20J

there is a high percentage of Where OJ = Asfyd I(Acfcd) is the


This criterion states that second •
reinforcement (because total mechanical reinforcement ratio. If
order effects may be ignored if the
the same is not known, 'B' may
slenderness A is below a certain member stiffness is then less
be taken as 1.1, that is equivalent
value Alim, i.e., A < Alim. affected by the cracking of the
to OJ =0.1. This value would
concrete).
A"m = 20.A.B.C / Fn usually be achieved in a slender
the locations of the peak first column; however this is generous
Where, n= is the relative normal •
order is not the same as the in comparison to minimum
force reinforcement provisions in the
location of peak second order
codes.
n = NEd I «J; moment.

As the axial force 'n' becomes 2.2.3 The term "C" Accounts
These effects are accounted for by
greater, the section becomes for Bending Curvature:
the terms A, Band C respectively
more susceptible to development in the Alim equation.
of second order effects and , • C = 1.7-r, where moment ill

consequently limiting slenderness2.2.1 The Term "A" Accounts ratio r m = M 011M 02.
value become lower. for Creep as below:
• MOl & M02are the first order
In establishing the above Alim, A=lI~ + O.2¢ef) end moments at two ends of
member as calculated from
the effective creep ratio <pef <Pefis effective creep ratio.
.:« o ).--MoEqp
and the relative normal force <Pefc-Y'\OO,t the analysis of structure.
n were included as parameters MOEd
• If the end moments give
in the initial stages, however
tension on the same side , r m
the reinforcement ratio 'co' was MoEqp = First order B.M.
should be taken as positive (i.e.
not included then, since it was in quazi-permanent load
considered unpractical. combination in SLS.

Journal of the Indian Roads Congress, October-December 2015


ULTIMATE LIMIT STATE OF INDUCED DEFORMATIONS: SECTION 11 OF IRe: 112 SYNTHESrsED 191

C ::;1.7), otherwise negative Unbraced members: this equation shows that for the
(Fig. 4) (i.e. C > 1.7). theoretical case of a member with
ends built in rigidity for moment
If 'r m 'is not known, C may be Ie =1 *maxof{ (1+10. klk ·k
l 2
)
• +k2 0 (k. = k, = 0), but free to sway in
taken as 0.7 which corresponds the absence of positional restraint
to uniform moment throughout at one end, gives the effective
the member.
1+ i * 1+ 2
l+kl 1+k2 [ k)[ k)} length 10 = I

• 'C' also should be taken as Where It is the relative rigidity of


0.7 where there is transverse restraint to flexural stiffness of
loading, where first order k., k2 are the relative flexibilities
the compression member i.e.
moments are predominantly of rotational restraints at ends I
important in determining effective
due to imperfections and and 2 respectively.
length. Consequently, using the
where the members are not un cracked value of stiffness for
braced. the pier will be conservative as the
restraint will have to be relatively
stiffer to reduce the buckling
8/M= is the rotation of restraining
length to a given value. This also
members at a joint for unit bending
is in line with the definition of
momentM
radius of gyration, 'i', given in the
EI = is the bending stiffness of in IRe: 112 which are based on
compression member the un cracked section. However
it is apt that the cracking needs to
Io is the clear height of be considered in determining the
compression member between stiffness of a restraint, such as
end restraints. .reinforced concrete pier base, if
.
.~
For the unbraced members with
it significantly affects the overall
stiffness of restraint offered to the
rotational restraint at both ends, pier. It is seen that quite often
the second equation above can the overall stiffness is governed
Fig. 4 Column in Double Curvature be used. Quick inspection of by the soil stiffness rather than
2.2.4 Effective Length of
Compression Members
Regular Frames (Ie)
compression members in
regular frames, the effective
length Ie is determined in the
following way
Braced Members:

1//
I = 0 51
e . 0 (
1+ )
0.45k + k, J * (I + 2
0.45k + kz J (,ot WUlngofIMWNllp.n(btK«f) (b)~ W<:lJnQlntwl'fmodo(oriwK«f)

Fig. 5 Buckling Modes for Braced & Unbraced situation

Journal ofthe Indian Roads Congress, October-December 2015


HEGGADEON

Reinforced Cement Concrete analysis carried out by computers is found, then the buckling load is
element. Further fully rigid give reduced value of effective the axial load in the member of
restraint is rare in practice that length. interest at buckling.
minimum value of the 'k' should
The analytical method could also 2.2.5 Effective
be taken as 0.1, even if the j oint is Heights of
fully restrained. be carried out for such situations to Isolated Piers (I)e FIB model
deduce accurate effective lengths code 2010 and euro code for
Building frames where the by applying coexisting loads to all bridges give recommendations
degree of restraint can be easily columns and increasing all loads for isolated members of effective
assessed, in bridges it is difficult proportionately until a buckling lengths as given in the Fig. 6.
to assess the degree of restraint mode involving the pier of interest
as the superstructure is supported
on bearings (Fig.5) which creates
boundary conditions with in the + \
+ (,
i ,, + --. -,- )(
+ +
,, , ,
bridge system by virtue of its ,, I
\

,
,,
I

layout making the assessment of , , ,, • ,, \

•• • • I I

braced or un braced complex. I


• I
, • •,, • ,, (}
I
\
\

I
I
I
I
,,
I
•,
The value of the end stiffness ,, I
,, , \ I
,
, \
\
,
to use for piers in integral \

\ (}

construction can be determined (


\

( ( ( ( Me ---1-

from a plane frame model by


a) 10-1 b) 10 21 c) 10 8: 0,71 d) 10 -"
II
2 .) 10-' f) 112<10< 1 g) 10 > 21
deflecting the pier to give the
deflection relevant to the mode
of buckling and determining the Fig. 6 Effective Lengths as per Model & Euro Codes
moment and rotation produced in
the deck at the connection to the While arriving at the above Effective lengths to be used for
pier. recommendations, it appears that cantilever bridge piers have been
the members were assumed to bone of contention even in the
The case shown in the Fig.S (b) be infinitely rigid at bottom for earlier version of codes. While
above do not permit any rigidity the cases a) to e) and also some traditionally, the effective height
of positional restraint in sway of the peculiarities of cantilever to be used for the piers which are
cases. However, this case is bridge piers supporting decking held in position at both ends are
very hard to exist in reality and through bearings with in based on realistic end conditions
practicality. If significant lateral the bridge system were not meaning, for the slender piers
restraint is available, as might be considered as such the provisions which are fixed at both ends
the case (b) of Fig.5 where one appear to be on liberal side. On 0.70 times of actual length of
pier is very much stiffer than the other hand, the table (Fig.7) the pier, for one end fixed and
the other or carrying more axial in IRC: 112 borrowed from other end pinned slender piers
force ignoring this restraint will 1985 version of BS 5400 are on 0.85 times the length of the pier.
be very conservative as the most conservative side and warrants For the cantilever piers which
flexible piers may actually be thorough understanding of the are supporting the deck through
braced by the stiffer one. In such background deliberations for bearings, the effective lengths
cases elastic critical buckling accurate application. were said to be between 1.0 times

Journal of the Indian Roads Congress, October-December 2015


ULTIMATE LIMIT STATE OF INDUCED DEFORMATIONS: SECTION 11 OF IRe: 112 SVNTHESISED 193

Case IdNliHd colulM and Restraints in Plane 0 .Bucklina Etfeetift the actual lengths of the pier to
buckling mode lo~tion Position ROQtion length. 2.00 times. The value between
I.,
1.0 times the actual length to 2
1.
(t(U ~
~O
Top Ful None
times is based on the assumption
1.01., that the base pier is infinitely rigid
Bottom Ful None for restraint while at the top of
~. the cantilever pier, the degree of
Top Ful Ful· restraint varies depending upon
~

{l Bottom Ful Ful


0.101., the type of bearings used from
various degrees of restraint to no

~u
.s. restraint at all.
Top Full Non.
0.851., The piers which are not in regular
Bottom Ful Ful· frame and supporting decks
through bearings, finding out the
~.
Top Non.- Non.- extent of restraint at top is very
1.31., complex as such there will be
r.~':""'" Ota~
Bottom Ful Fur always a tendency to advocate and
adopt the highly conservative and

~[r
e.
Top None None uneconomical effective lengths
1.41., of 2.30 10 as proposed under the
Bottom Ful Fur case 7 of Fig.7. In fact, after
much deliberations, the said table
o.
,,//i, /" Top None Fur was adopted in the code from

7.
b[l B-ottom Ful Fur
1.51.,
BS:5400 part 4 which was based
on a scientific investigation as
brought out by P.A. Jackson in his
technical report no. 561 for cement
and concrete association viz. 'The

r.ItTII
Top None Non.
2.31., buckling of slender bridge piers
Bottom Ful Fur and the effective height provisions
ofBS: 5400:Part- 4'.
Fig. 7 Effective Heights of isolated Piers
It is very essential for us to
understand the idealization of
slidIng roiling deforming
functions (Fig.8) of translation
displace-
¢=> ¢=> ¢=> and rotation in bearings to assess
ment
I I
0 I
V V
I
the buckling effects in slender
piers. In guided POT -PTFE,
spherical and disc bearings,
rotation
§&i ~ ~
translation as well as rotation
are enabled and in fixed bearings
translation is restrained. In
Fig. 8 Translation and Rotation Function in Bearings pot and disc bearings rotation

Journal of the Indian Roads Congress, October-December 2015


HEGGADE ON

is enabled by deformation of that the pier deflects at the top a schematically idealized bearing
elastomer confined within the with the line of axial load moving as shown, the effective height of
pot, while in spherical bearings sideways with the pier top being the bearing is 2.0 10 for an infinitely
rotation takes place by sliding parallel to it as shown in the rigid base.
of surfaces on cylindrical or
Fig. 9. When the friction resistance However as shown in the Fig.1 0,
spherical surfaces.
offered to translation as well if the rotation is above the
Normally for a cantilever pier in as rotation is zero and also the translation in a schematically
a classical sense it is assumed translation is above the rotation in idealized bearing, the axial line

,
\ ,

Llo~01 thru.t "I ~


axial toad
(romalns p.:ar;r;Ucl
I
wHo pJor top but
l;onLlnuo$ to or;t
through contr0'41nO'
of bailrtng at deck
I
I

lovel) •

I
I
•.....•_.......,..-kl_,..,. _01.....,,__ ....0!1,..,
y, i trl ••

(a) With Inlinn. baSI) I'OStI':lint (b) With zero


I
b.it$$ T$$U8;nt
_ T«htIiuI rtpO!f $IS!, C & CA. P.AJ.!c.\s<)t!, JuM fNj
Source: Technical report 561, C & CA, P.A.Jachson, June 1985
Fig. 9 Left: Buckling Mode Normally Assumed
for cantilever. Right: Effect of Bearing Friction Fig. 10 Alternative Buckling Modes for Cant
Piers
of thrust no longer can move with it but
has to rotate along the top of the pier

: ; * :
~~ ~~ ~~ ~~
which is because there cannot be shear
along the rolling/sliding surface. In his
paper under reference (3), Paul Jackson Case (a) : Fixed @ P1 and Guided @A 1, P2 & A2
proves mathematically that for a fully
restraint base, the effective height is ~\.•.o~ ~~ 6,j1>~ 6,j1>o~

1.12 10, If the restraint considered at


base is 4EIIlo' the same works out to be
1.410,

As can be seen from the Fig. 11, there


: ~* :
Case (b) : Fixed @A1 and Gulded@

'(,~.
••o~(
.,eIi"
At
P1, P2 & A2

.,0
~t
",,0.'"
o~(

could be different combinations of


bearing layout with in the bridge system ~",,\y
~
>,.\(4

1"'· I~- "'~~


c)ot
~
-" ot
~

6ri,
~(4
~v

depending upon the response required for


the extemalloadings/actions. Under the
@ le) : Fixed ®P1 & P2 and ®
case Guided
@ @A1 & A2
@
case (a) of the same Fig. 11, there could Case (d) : Elastomer bearings @ A1, P1, P2 & A2
be fixed bearing at P 1 which could be Pot,
Disc and Spherical bearing which allows Fig. 11 Bearing Layout with in Bridge System
only rotation but not the translation,

Journal of the Indian Roads Congress, October-December 2015


ULTIMATE LIMIT STATE OF INDUCED DEFORMATIONS: SECTION 11 OF IRe: 112 SVNTHESISED 195

holding the deck in position.


Other piers and abutments may
have Pot-Ptfe, Disc, Spherical or
elastomeric bearings which allow
both translation and rotation. •• •• p
X- -" -.p
Only the difference in case (b) is i X 1--':''*.''".
~ OfplOr
Rotation
rRHultzlnt
-Y1' \ Une 01 thrullt

that fixed bearing is at abutment \


Al instead of pier PI in case (b) pIfII' fOP ~d by Y1 (e) PI," fOP rol.r.ct by n'
r--c.ntr-.lin. of
(a). For this bridge system when i x undetl_t.-<t pier
· ,.
there is an external action due
to Temp/Shrinkage/Creep, the .----f'---- I
"-;t

deck expansion or contraction is


assumed to be originated at fixed
tp •
I
••p
X i--c.ntro
x"
RGt
~p
(M'I-Y'
• tlnoof
location and by virtue of friction t-v,-i undeflected pI.,
(.) Undeflecred (b) PifN top dflflfK:ted by Y1 {f:} Pier top rot/d.d by Yr
at sliding surfaces, the piers which
are not fixed deflect. Even under Fig. 13 Top Behavior of Roller Bearing. Bottom:
the external actions of earthquake/ Behavior of Elastomeric Bearing
braking, the fixed pier deflects in
the direction of force on other than freedom for free piers are In fact, the above effective
fixed piers sliding is enabled and represented by roller bearings in lengths are without taking into
by virtue of friction to the extent top of Fig. 13 for POT-PTFE, disc. consideration the relief provided
of resistance provided by friction, and spherical bearings whereas to buckling force due to frictional
piers deflect. The buckling mode the same is depicted at bottom force as illustrated in Fig .9 (Right).
for the cantilever piers having Fig. for elastomeric bearings. In In reality the bearings cannot
sliding/elastomeric bearings is fact, boundary conditions derived be frictionless but the friction
depicted in Fig. 12. for both are same though the always opposes the motion which
translation is achieved through means that it always opposes the
The boundary conditions for different behavior. As can be seen
buckling. As can be seen from
rotational and translational from the figures, when pier top
the fig., the pier can only fail by
moves in pure translation for a bending to right. This involves
distance of Y' the roller moves
the translational movement across
by a distance 'Inwhile elastomer
, the bearing that will be resisted
deforms by the same distance.
, I by friction. Thus there will be
I ' When the pier rotates by an angle
I ,I
Centre line of ~ frictional force opposite to the
undeflectlon pier Y 'I' the roller is not squeezed out
direction of buckling force. In
due to the frictional forces FI and
all the derivations of effective
F2 developed that stop squeezing
Line of thrust of axlal Ioad ~~' heights, the opposing frictional
out. As has been mathematically
(rotates and moves half as far
as pier top)
I

I derived by P A Jackson under


force is neglected as such the
,
reference (3), for the above effective heights proposed are on
I highly conservative side.
boundary conditions the effective
""".J""",J height is 1.010 for fully restraint
When bearings are fixed at PI
Fig. 12 Buckling Mode with base and for restraint of 4EIIlo' the
and P2 location as indicated in
SlidinglElastomeric same is 1.310,

Journal of the Indian Roads Congress, October-December 2015


HEGGADE ON

case (C) and decks are supported 3. ANALYSIS FOR 2ND ~ Equal strains in reinforcement
by elastomer bearings at all ORDER EFFECTS and concrete at the same level.
locations as shown in case (d)
under the action of seismic! There are three methods of analysis ~ Given stress-strain relationships
braking the piers sway in the for 2nd order effects deliberated in for concrete and steel.
direction of force. Under such codes.
un braced situation for the case These assumptions are "classical"
General Method. and are known to give realistic
( c ) and ( b ) in Fig. 5 for the 1.
fixed piers, the effective lengths 2. Nominal stiffness method. results. In the stress-strain
relationships, a tensile strength
may be taken as 2.310 as shown 3. Nominal curvature method.
in the Table 1 under the case 7 of the concrete can be included,
since the buckling mode will or disregarded. If a tensile
While the first method of analysis
be analogues to the depiction strength is included, the effect of
'General method' is a non-linear
in Fig. 3. However, for the case tension between cracks (tension
method of analysis more suited
(d) where elastomeric bearings stiffening) can also be taken into
for computers, the 2nd and yd
are provided for unbraced account. However, usually no
are simplified methods meant
situation, as buckling mode will effects of concrete tension are
for manual calculations. For
be analogues to depiction in included. This is more or less
some reasons, 'nominal stiffness
Fig. 4, the effective height conservative, "less" rather than
method' is not recognised in
remains to be 1.3/0, "more", and furthermore, it is a
IRC: 112.
common principle to disregard the
From the above deliberations, it 3.1 General Method of Non- direct effect of tension in concrete
is clear that the effective length linear Analysis in ultimate limit state design.
provisions in Fig. 7 of IRC: 112
Non-linear analysis of slender "General" here refers to the fact
is on conservative side as in the
compression members includes that the method can be used for
derivation of the same, low base
material and geometric any type of cross section, any
restraint is considered and friction
nonlinearity (second order variation of cross section, axial
force opposing buckling is not
effects). The method used rests on load and first order moment, any
taken in to consideration. While
the following basic assumptions: boundary conditions, any stress
for all the other situations of
cantilever piers where bearings ~ Linear strain distribution. strain relations, uniaxial or biaxial
are supporting the deck, the bending etc. The limiting factor
effective heights of 1.3 10 can be assumed is the capability of the
assumed
computer program.
variation of
safely taken, for the unbraced curvature.
e.~. Iinear. __
situation as shown under case constant or
II varianon of
curvature. e.g.
similar to thaI
of the first
other order moment In the accurate method
(c) of Fig. 11, for the fixed piers
cross section(s) as can be seen from Fig.
PI and P2 the effective height of ~ for wjuch equi .•
librium aud __ 14 (Left), conditions
the pier can be considered to be i-'i>-i: --- defonn..uion
_______ comparibifiry of equilibrium and
2.310 or can be worked out from \ are considered
deformation compati-
the expression given in the code
bility are satisfied in a
for unbraced members where the
numberofcrosssections,
restraint at bearing location may
and the deflection is
be considered to be infinity in the Fig.l4 Accurate (Left) & Simplified (Right) calculated by double
expression. version of General method.

Journal of the Indian Roads Congress, October-December 2015


ULTIMATE LIMIT STATE OF INDUCED DEFORMATIONS: SECTION 11 OF IRe: 112 SVNTHESISED 197

integration of the curvature, For concrete, 'Yce = 1.5 for based on the final value of the
having an assumed vanation strength takes into account not creep coefficient and reduced
between the selected sections. In only strength variation, but also with regard to the relative
the simplified version (Right in geometrical deviations in the cross effect of long-term load in a
Fig.14), only one cross section (or section. Assuming a factor 1.1 for load combination.
certain critical sections) is studied, these deviations, and considering
and the curvature is pre-assumed the relationship between strength 2. The shrinkage of concrete has
to have a certain variation in other and E-modulus, a reasonable been neglected.
parts of the member. This enables value of the factor for E c is 'Yce = 3.
The strength increase of
simpler computer programs and 1.1*(1.5/1.1)1/3 =1.2.
concrete with time has been
faster calculation compromising a
For steel, 'Y = 1.15 includes a factor neglected.
bit on accuracy. s
of about 1.05 for geometrical
3.1.1 Treatment of Time
Any stress-strain relations can be deviations. Thus, a design value
Dependent Parameter Creep
used. A continuous curve with a Esd=Esm/1.05 would be logical,
Creep can be taken into account in
descending branch is considered considering that variations in
different ways. The most accurate
to be the most realistic alternative the E-modulus are negligible.
model would be to increase load
for the concrete while the same is However, a factor 1.0 has been
and time in steps, for each step
also convenient for computational chosen as simplification.
taking the stresses, strains (and
reasons.
In all the methods of analysis the corresponding deflections) from
The safety format is particularly following additional assumptions, the previous step as starting values
important in second order related to the time-dependent for the next increment. For each
analysis, where the absolute properties of concrete, have been step, strains would be calculated
magnitude of deformations has a made: taking into account their time-
direct influence on the ultimate dependence. A simplified model
load. This has been taken care of 1. The effect of concrete creep is to multiply all strain values in
by satisfying 2 basic criteria: has been taken into account by the concrete stress-strain function
extending the concrete stress with the factor (1 + <Pef), as in
• By using the same set of strain curve according to Fig Fig.15, where <Pef is an effective
material parameters in all 15, i.e. all strain values are creep ratio relevant for the load
parts of the member, in order multiplied by (1+<.Def). <.Defis a considered. With this model, the
to avoid discontinuities and so called effective creep ratio, analysis can be made either in
computational problems. steps for loads of
different duration
• By adopting a format or directly for
compatible with the general the design load
design format based on partial combination III
safety factors. one step.

Both these criteria have been Through the


satisfied in IRC: 112 by use of Fig.16 the relevance
design values directly in ultimate of the effective
analysis. creep ratio for
Fig. 15 Simple way of taking creep in analysis slender columns

Journal ofthe Indian Roads Congress, October-December 2015


198 HEGGADE ON
1

er
creep ratio <D (line AD in the
Fig.16). For the definition of <Def
there are two main options:

a. based on first order moments


f
I
I
I
;-
,,,',,'.,.'--'''1
.,.
: ,
.C
I
MOL and MOD' i.e. <Def = <D.MOL I
MOD
I
I
f
I'
I.,.,
1::"....
, ",'
I:
••
, •

I
I
I ~I : II b. based on total moments ML
Tint! , ,.
! :
f I and MD' including 2nd order
moments, i.e. <Def = <D. M/
~ t'-_Y",!~;JI:R_.":"Yl'--"'::Y.J..1 --i Yz

The
MD

relevant deformation
Fig. 16 The Concept of Effective Creep Ratio (<Def)
parameter in second order analysis
is curvature, which depends
are examined. A slender column (1 + <D). This corresponds to line
primarily on bending moment.
behaves in a non-linear way, due AC, and the calculation is then
Therefore, the axial load should
to both material and geometrical reduced to two steps. The last step
not be included in the definition of
non-linearity. can be calculated in further two
effective creep ratio. Alternative
alternative ways:
b. is the most realistic one, since
The load history can be divided
creep deformations will mainly be
into three steps: 1. After calculating point C, the
governed by total moments. With
additional load QD - QL is added,
this alternative, however, iteration
1. For the application of long- with deformation starting from y..
is inevitable since second order
term load QL, immediate See line CD in Fig.16.
moments depend on stiffness,
deformation is Y!' for <Def = 0 which depends on effective
2. After calculating point C, the
creep ratio, which depends on
2. For the long-term load total load QD is applied "from
total moments etc. Therefore,
QL during time t-to' total scratch", but with Yo = Y2 - Y1 as
alternative a. will be the normal
deformation is Y 2' for <Def = <D an initial deflection added to other choice in practical design.
first order effects. See line ED in Alternative a. is always more or
3. For the load increase up to Fig.16.
less always the safe side. As the
design load QD, additional
second order moment is a non-
deformation is Y3 - Y2' for <Def Alternative 2 can is used in two- linear function of the axial load.
=0 step calculations in the following
Therefore, the moment increase
way. The distribution of Yo along due to second order effects will
The calculation representing the the column should in principle
be greater under design load than
above load history involves three be the same as the distribution
under long-term load, and the ratio
steps, including the relevant first of Y2 - Y1• For a pin-ended M L1MD will be lower if second
order moments or eccentricities column, however, a sinusoidal order moments are included.
for each step. For simplification, or parabolic distribution shall
teps 1 and 2 can be combined into be adopted as a simplification. A Apart from creep, other parameters
one, using a stress-strain diagram further simplification is a one-step have a fundamental effect on the
with the strains multiplied by calculation, using an effective ultimate capacity of a slender

Journal of the Indian Roads Congress, October-December 2015


ULTIMATE LIMIT STATE OF INDUCED DEFORMATIONS: SECTION 11 OF IRe: 112 SYNTHESISED 199

compression member with a given Slender compressive members in ultimate limit state design this
cross section, which are are seldom restrained in such a can be ignored.
way that it could give any effect
When there are second order
• Amount and configuration of of shrinkage on the member as
longitudinal reinforcement effects, in which case other
a whole; furthermore, if there
time dependent effects like
were such a restraint, it would
• Slenderness creep and shrinkage become
rather be a favourable effect
important, the problem is
since it would reduce the axial
• Boundary conditions different. In the beginning of
force. This may be a reason why
the service life, there is little
• Magnitude and distribution shrinkage is never discussed as a
strength increase but, on the
of first order moment (or factor of importance for the load
capacity of slender columns. other hand, deflections are not
eccentricity of axial load)
However, shrinkage will have yet much influenced by creep
• In case of biaxial bending: an effect on the internal stress (and shrinkage). Therefore,
distribution, which might the critical design condition
vi' another first order moment, in turn have an effect on is normally assumed to occur
independent of the first deflections and hence on the at the end of the service life,
one ultimate load. Shrinkage leads when deflections are at their
to a transfer of compression maximum due to the creep and
vi' proportions of cross from concrete to reinforcement, to some extent shrinkage. On
section much like the effect of creep the other hand, at that time we
but independent of stress. This also have the maximum effect
3.1.2 Effect of Shrinkage
can be both favourable and of strength increase, and it is
on Analysis of Compression
unfavourable, depending on no longer self-evident that the
Member Shrinkage is a material
the concrete and reinforcement most critical design condition
property of concrete which is as
compressive stresses, cracking is always to be found at the end
inevitable as creep. Taking into
etc. It is difficult to say whether of the service life; in principle
account the effect of shrinkage
favourable or unfavourable it may occur at any time within
is self-evident in many design
effects will dominate. the service life. Whatever
situations, e.g. with regard
possible unfavourable effect
to cracking in members with 3.1.3 Effect of Increase
on the load capacity due to
restraint to shortening, time- in Strength on Analysis of
neglecting of shrinkage can be
dependent losses of prestress Compression Members It is well
considered to be neutralised
and in accurate calculations established that the strength of
by the favourable effects that
of deflections. These design concrete, at least the compressive
could have occurred due to
situations represent serviceability strength, increases with time, due
limit states. However, in to the continued hydration of the strength increase.
ultimate limit states the effect cement. The tensile strength may
3.1.4 Simplified Method of
of shrinkage, like many other not increase to the same extent, it
Analysis for 2nd Order Effects In
types of imposed deformations, may even decrease due to tension simplified methods the difference
is usually disregarded. This is in the cement paste when the between cross section resistance
justified in most cases, due to the aggregates resist the shrinkage, and first order moment, Mu - Mo
possibility for redistribution of but as long as the tensile strength is used as a nominal second order
stresses and moments. is not relied upon for direct tension moment. When this moment is

Journal ofthe Indian Roads Congress, October-December 2015


HEGGADE ON

added to the first order moment,


a design moment is obtained
for which the cross section can
In the stiffness method 11r is
be designed with regard to its
ultimate resistance. expressed in terms of an estimated
nominal flexural stiffness, i.e. 11
There are two principal methods r=M/EI, while in curvature method
to calculate this nominal second curvature is estimated directly
order moment: on the basis of assuming yield
• Estimation of the flexural strain in tensile and compressive
reinforcement i.e.
stiffness EI to be used in a Fig. 17 Simplified Methods for
linear second order analysis Pin Ended Column 1/r = 0yi0.45d.
(i.e. considering geometrical
non-linearity but assuming • I= length
Differing first order end moments
linear material behaviour);
this method is called stiffness • c = factor for curvature MOland M02may be replaced by an
method. distribution

• Estimation of the curvature Though IRC:


1/r corresponding to a second ..
order deflection for which
112 requires to .
consider geometric
the second order moment is . ,
calculated; this method is imperfections in to
called curvature method, account III many
recognised in IRC: 112. places, there is no
specific guide lines AI!)I

The total moment including


are provided. It
second order moment for a simple
isolated member is: IS recommended Fig. 18 Illustration of Equivalent 1st Order
to consider the Moment in Case of Differing end Moments
following to account
for geometric imperfection for equivalent 1st order end moment
compression member on the Moe:where,
• M = total moment
basis of Euro code and fib model MOe=0.6M02+O.4M 2: 0.4 M .
OI 02
• MO = first order moment code.
including imperfections. M2 = MOl and M02 should have the same

second order moment The effect of imperfection may be sign ifthey give tension on the same
represented by an inclination, side, otherwise opposite signs.
• N = axial force Furthermore 1 M02 1 t 1 MOl I·
an
• y = deflection corresponding Q1= 200 The 2nd order moment M2 IS
to 1/r deduced by the expression
Where an IS the reduction
• 1/r = curvature corresponding factor
to y
for the clear length M2 = N ei X e2 = N ei X lLL
r' c

Journal ofthe Indian Roads Congress, October-December 2015


ULTIMATE "LIMIT STATE OF INDUCED DEFORMATIONS: SECTION 11 OF IRe: 112 SVNTHESISED 201

Where, Fig.19 in which is is the radius of K is a factor for taking account


<p
gyration ofthe total reinforcement of creep which is given by
Ned is the design value of axial area. K is a correction factor
r
force depending on axial load. In order
to reduce the curvature in cases
e2 is the deflection
where yielding is not reached Where
lIr is the curvature. in the tensile reinforcement,
a factor Kr is introduced. 1/1 = 0.35 + fck - ~ I
200 150
Ie is the effective length. K, = (TJ" -TJ(.JTJ" ))$ 1
-TJI><>J
Comparisons with the
c is a factor depending on general method indicate
the curvature distribution. For that in certain cases the
constant cross section, c= 10 = method can give unsafe
(rc2) is normally used. If the first results if allowance for
order moment is constant, a lower creep is not considered,
value should be considered(8 is and the factor K <p has been
a lower limit) corresponding to introduced for this purpose.
c
It has been calibrated against
constant total moment.
nbali--------2I
calculations with the general
As mentioned in the Fig. 19, method.
l/r is estimated on the basis of
• <P ef=is the effective creep
reaching yield strain in tensile and Fig. 20 TJba1as0.4 at Maximum ratio
compressive reinforcement. Moment of Resistance
• TI = relative axial force = NEi
• A = is the slenderness ratio .
Afcd
c 3.1.5 Biaxial Bending The
1 -K" ,.1\.'I'--1\.r
I" 1 _ VK'"",x--
&,rd
-- effects of slenderness for
I' 1'0 0.45d • NEd = design value of
columns bent biaxial are most
-Where: axial force.
V'eyd =~/ E, accurately determined using non-
,f d = is the effective depth linear analysis. The simplified
V' I, = is the radius of gyration of the • TI u =1+00
total reinforcement area; methods can also be used for
• TIbal = value of n at the case of biaxial bending.
Effective depth In maximum moment of The second-order moment is
cross sections with
reinforcement resistance; the value 0.4 first determined separately in
distributed In may be used (Fig. 20 ) each direction following any of
direction of bending
the above methods, including
imperfections. Itis only necessary
Fig. 19 Accounting for lIr while
to consider imperfections in
Calculating M2 • A = total area of one direction, but the direction
s
reinforcement.
Here correction factors K r and should be chosen to determine
Kq, are included. Where there is • the most unfavorable overall
A = area of concrete cross
unambiguous definition of d, the s;ction effect.
effective depth is worked out as per

Journal of the Indian Roads Congress, October-December 2015


HEGGADE ON

There are areas where biaxial For circular columns, it is possible • III persistent situations:
bending moments can be neglected, to take the vector resultant of lOt 50
-<---- and h/b s2.5
like after considering bending moments in two orthogonal be - (h/by/3
in each direction separately, if directions, thus transforming the
the slenderness ratios in the two problem into a uniaxial bending •
transient situations:
principle directions do not differ problem with M2 considered only
by more than a factor of 2 and in the direction of the resultant lOt:s; 70 /3 and h/b ~.5
the 'relative eccentricities' do not moment. be (h/bY
differ by more than a value of 0.2
as represented in Fig.21. In general, however, it IS
./ lOtis the distance between
recommended here that M2
torsional restraints
• tJ)Caf{lcwntri¢tfl}$
conservatively be
r••cluding2"'order) .. calculated for both
__ 0 ~ l¢~ ~"enlJ1<itl<$ ./ h is the total depth of beam
B' /,1<lorde(Q<cenu. directions. Bending
--- .•..
-~~=--.. ord",oo:._

should then be checked


in each direction
in central part of lot

./ be is the effective width of


Fig. 21 Different Slenderness independently, and compression flange
in 2 Direction then biaxial bending should be
considered (with M2 applied in The IRC: 112 proposes the clear
Where this is not satisfied, the both directions together) unless distance between the restraints of
moments in the two directions second-order effects can be b 2

(including second-order effects) neglected in one or both directions. 60 b or, 250-e-,


e d
must be combined, but Imperfections should only be whichever is the lesser, for the
imperfections only need to be considered in one direction. lateral stability of simple or
considered in one direction such continuous beams, where'd'
as to produce the most unfavorable 4. LATERAL STABILITY is the effective depth 'be' is the
conditions overall. Section design OF SLENDER BEAMS breadth of the compression flange
under the biaxial moments and of the beam midway between
axial force may be done either by IRC: 112 requires to investigate restraints. For cantilevers with
a rigorous cross-section analysis lateral instability of precast bridge lateral restraint provided only
using the strain compatibility girders before it becomes a part at the support, the clear distance
method or by simple interaction of the finished bridge system, from the free end of the cantilever
during casting, transportation
l£...)2
( MRx + (~)a
MRy
<1 and erection. It specifies 1/300 as
to face of the support is limited to
25b or 100bc 2 whichever is the
casting geometric imperfection
less~r. d'
for unbraced conditions where I =
• Mx/y design moment in the total length of beam. The above empirical expressions
respective direction, including
seem to be applicable for non-pre
nominal 2nd order moment. While carrying out investigation stressed beams. The pre stressed
for lateral instability, the 2nd order beams have more complexities
• M Rx/y corresponding moment
effects could be neglected on due to induced pre stress resulting
resistance of cross section.
fulfillment of following empirical in locked up compressive and
• a exponent varying up to 2 . conditions. flexural stresses in beams.

Journal of the Indian Roads Congress, October-December 2015


ULTIMATE LIMIT STATE OF INDUCED DEFORMATIONS: SECTION 11 OF IRe: 112 SYNTHESIS ED 203

Robert F Mast deliberateslimits and in fact decides the safety against lateral stability
extensively on 'Lateral stability ofmaximum tilt (8max) to which the by virtue of reduced deflections
long prestressed concrete beams' beam can be subjected to. After caused by rotations about the
in two parts i.e. part 1 and part 2,the tilting is initiated by the initial weak axis. However, it has to be
published in PCI Journals. angle 81 near the support locations, ensured that the stresses are within
the beam achieves its equilibrium the limits in overhang portions.
Lateral stability aspects of PSC with a uniform lift angle 8 (shown
beams are discussed below by at midspan) with CG of the mass Classic studies of lateral buckling
illustration of an example as of the deflected beam right under of beams are based on the
shown in Fig. 22. The improper the roll axis. assumptions that the supports are
lifting hook placement and casting restrained for rotations. However
imperfection can cause the beam In the figure as Zo approaches for prestressed beams in the
to be tilted at an initial angle SI Yr' the beam starts rotating and casting yard, the same has to be
near the lifting hook location becomes totally unstable even supported on elastic pads as the
about the roll axis. Normally the without the initial imperfection prestressing has to be induced.
and without improper Due to the casting imperfections
'~~
location of lifting
IHITW.su~rfCEN'RlCHYOSU-""OR:T·025'",3e<fn
buckling is caused by the middle
J"'.._
~ ••. ~~~~:"'(lJMD

hook.
'--"'W'IttU.tMS
CASnNGIHPfRFECTlCI!·t19X1(A$SUMEOJ

e.,'INITlALlATE!tIt.Ea:NlRIClTY'I3S'IU16'227$6r.wn
'16«)6mon

part of the span twisting relative


.~ ~ • .......x"U"'THEOAIrICALOEfI.£CtlON O(l25/JII
':::::::-... AC1Ut.l.DEFLECTICl4.!' ~SW'I$

,.,...·IolA,)(I •••U"'PVNSI8eU!TLT·1~·(I)I)UI1I:IJ
to the support creating a sideway
~~~.;, ~'~::;:e~L:;=TO~:;WH'I37t'" Thus the safety against deflection. In case of the beams
::~::::~TY~H:;~~INPt:rncT~'$.~1
the lateral buckling is that are lifted by lifting hooks as
.~'_~*.~,-,~
~'r::::S ~: - .----~~-.
.~::;~:::::::~:::::::,.=:,::,;.'::::
(lfF,.tSOle.TKENMAXINITW.tATEAAl.ECETRIOTYIS(e.-.J·TJ~1
a measure of Y vis-
a-vis Z and is ~alled shown in Fig. 22, the roll axis is
(_11I"01'

:::_~
~~'~~
,..

·N. ~=~~
.~\P
gross factor of safety
w
0 at top while the beams that are
supported during casting as well
(FOS = Y r/Z)0 for a as during transportation have the
Fig. 22 Equilibrium of Beam at perfect beam without roll axis at the bottom as shown in
Tilted Position imperfection. If one has to account the Fig. 23. Thus the value Yr for
for imperfections causing the the beam being lifted is below the
casting imperfections considered initial angle S1 and limiting the
roll axis whereas for those beams
gets manifested itself by way of maximum lift to S ,the factor of
max laterally buckling and bottom
curvature in plan of beam after
supported have Yr above the roll
transfer of prestress force. Lifting safety reduces to ~ (1- ~ ) .
hook placement tolerance needs
Bmax z, axis, all the other other principles
and concepts of lateral buckling
to be considered during casting. However, it is more logical to
being the same.
The above tilting of beam induces deduce the factor of safety
the lateral deflection about weak against lateral stability! ...\1-._
axis of the beam. Because of by dividing maximum I
the transfer of prestress, there is possible tilt 8 with that ~~"""
max
already tension at the top fiber of equilibrium rotation 8
OF......! 0 J~j
of the beam for which the tensile at midspan. 11\:
C.OOF , _

stress caused about the weak axis


by the component of the self- Moving the lifting
""'"

l I I

weight due to tilt gets added which position inwards ~


needs to be within the permissible Improves the factor of·
=:
••.
""""'AI)
Fig. 23 Equilibrium of Beam on Supports

Journal of the Indian Roads Congress, October-December 2015


204 HEGGADE ON ULTIMATE LIMIT STATE OF INDUCED DEFORMATIONS: SECTION 11 OF IRC: 112 SYNTHESISED

5. CONCLUSIONS beams are discussed which may 4. BS EN 1992-2 : 2005, euro


be useful for the readers. code 2 - Design of concrete
In the paper above, the background structures - part 2 : concrete
of clauses for Ultimate Limit Most of the deliberations bridges - design and detailing
State for induced deformations including the figures are taken rules.
have been presented which will from background documents
help the users to understand the prepared by the project team 5. V. N. Heggade, A vision of
intent of the clauses there by for EN 1992-1-1 (2004) as such modem structural code of
avoiding the misinterpretations. the same is acknowledged very practice: bridge between
As there IS always bone of profusely. code making and practice,
contention about the effective the bridge & structural
heights of cantilever piers, the REFERENCES engineer, B & SE_Volume
same has been deliberated in 44 - Number 2- June 2014.
detail. Though the geometric 1. IRC: 112: Code of practice
for Concrete Bridges. 6. Robert F Mast, Lateral
imperfections are said to be
considered in many places ofIRC: stability of long prestressed
2. Euro code 2 commentary,
112, there is no guidelines in the concrete beams Part 1, PCI
Published by the European
code as such the recommendation Journal/January-February
Concrete Platform ASBL,
has been made for compression 1989.
June 2008.
members which can be adopted
7. Robert F Mast, Lateral
in the code. The provisions made 3. P.A. Jackson, the buckling
stability of long prestressed
in IRC: 112 seem to cater for slender bridge piers and the
concrete beams Part 2, PCI
only non-prestressed concrete effective height provisions
J ournal/J anuary-February
beams as such the aspects of of BS: 5400: part 4, CECA
1993.
lateral buckling of prestressed Technical report 561, June
1985.

Journal of the Indian Roads Congress, October-December 2015

You might also like