You are on page 1of 5

Chris Hendy 27/1/06 Modelling bracing systems in grillage analysis of steel-concrete composite bridges

Modelling bracing systems in grillage analysis of steel-concrete composite bridges


1. Introduction For multi-beam decks, beams are often braced in pairs such that the bracin is not continuous trans!ersely across the deck" #he presence of bracin locally si nificantly stiffens the brid e trans!ersely" $oments and shears in the deck slab are attracted out of the concrete slab and into the bracin as sho%n in Fi ure 1 !ia the trans!erse stiffeners" #his effect is not modelled in a con!entional rilla e analysis unless the increased stiffness in the location of bracin s is included usin a shear fle&ible member %ith inertia and shear area chosen to match the deflections obtained from a plane frame analysis of the bracin system" #hree-dimensional space-frame or finiteelement representations of the brid e can be used to model these local effects more directly" #he transfer of moment causes tension in the shear connectors on one side of the flan e and induces compression bet%een concrete and flan e on the other" 'elds at tops of stiffeners must also be desi ned for this moment, %hich usually leads to throat si(es much reater than a )nominal* 6 mm that mi ht be appropriate for shear stiffeners %ithout bracin " +n e&treme circumstances, it has led to the pro!ision of butt %elds bet%een top flan e and stiffener or e!en additional usset stiffeners such as that on the left-hand irder in Fi ure 1" Clearly it is necessary to model these effects in some %ay so the forces induced in stiffeners, studs and bracin s can be pro!ided for in the desi n"

Figure 1 Typical transfer of moment from slab to bracing system 2. Modelling of bracing ,ne option for desi nin the bracin s is to i nore their presence in the rilla e and e&tract slab rotations and deflections at the bracin locations" #hese deformations are then -forced. on to a plane frame model as sho%n in Fi ure 2" #his approach is used in /C+ publication 0210 2ref3" +t often produces e&cessi!e forces and moments %ithin the stiffeners and bracin system because it i nores the reduction of deflections and rotations that %ould occur if the bracin system %as included" + norin the presence 1 of 4

Chris Hendy 27/1/06 Modelling bracing systems in grillage analysis of steel-concrete composite bridges

of the bracin system in the rilla e also i nores its benefit in impro!in distribution in the lobal analysis" ,ne %ay of includin the bracin in lobal analysis 2%ithout usin a 56 spaceframe or shell finite element model %hich i!es a superior representation3 is to represent it as a shear fle&ible member in a rilla e model" #he trans!erse member representin the bracin system and slab abo!e is i!en a fle&ural stiffness, EI, and a shear stiffness, GA" 7oth are determined from plane frame idealisations of the bracin system and an associated %ith of slab" #he torsional stiffness of the trans!erse member is based on that of the slab alone" ;a ;b :a :b

Figure 2 Deflections and rotations applied to bracing system to determine bracing forces and moments Fi ure 5 sho%s a typical plane frame representation of the bracin system and slab" #he bracin members in the plane frame ha!e the area and bendin stiffness of the bracin s themsel!es" #he stiffener members ha!e the area and bendin stiffness of a -#ee. section comprisin the stiffener and an attached portion of %eb plate of %idth 52t %here t is the %eb plate thickness" #he choice of %idth of slab to use is more complicated" #he actual %idth chosen %ill usually ha!e relati!ely little effect on the !alues of EI and GA for the idealised shear fle&ible member, and thus the distribution in the lobal analysis, but it %ill make a si nificant difference to the distribution of moment bet%een deck slab and bracin system" #he actin %idth of slab depends on the eometry of the local moment introduction by the stiffener and studs and the ability of the slab abo!e the beam to distribute load in torsion 2%hich can be enhanced by a local haunch3" #he effecti!e %idths for trans!erse members in 7/4800 0art 5 clause 1"14"2, based on shear la considerations, are therefore not directly applicable here but can be used as a uide" #he smaller the %idth of slab chosen, the reater the proportion of moment %hich %ill pass into the bracin system rather than the slab" 9 smaller %idth of slab is therefore conser!ati!e for the desi n of the bracin system, stiffeners and shear studs" ,ne option is to assume that moment spreads out into the slab from the stiffener location at 84 de rees" #his i!es an actin %idth of B mid%ay bet%een the main beam %ebs and an a!era e %idth of B/2" #he rules in 7/4800 0art 5 clause 1"14"2 lead to a %idth of B/8, but not reater than a!ailable %idth" /ince a smaller %idth is conser!ati!e, it is recommended here to use the latter, althou h this could be refined if the forces produced in the bracin system are found to be unacceptably hi h"

2 of 4

Chris Hendy 27/1/06 Modelling bracing systems in grillage analysis of steel-concrete composite bridges

EI is determined from the model in Fi ure 5" 9 moment M1 is applied and the deflection :1 obtained" For a simple beam element supported and loaded in the same %ay< M 1B 2 1 = 2 EI #hus M B2 EI = 1 213 21 GA is determined from the model in Fi ure 8" 9 moment M2 and force F2 are applied and the deflection :2 obtained" For a simple beam element supported and loaded in the same %ay, usin standard deflection formulae< B2 M B 2 F2 B 2 2M 2 2 = 2 + = M2 + %here F2 = 6 EI GA 6 EI GA B #hus 12 EIM 2 GA = 223 6 EI 2 M 2 B 2

M1

M1

:1

Figure 3 Determination of EI for shear fle ible grillage member M2 F2=2M2 / B M2 F2=2M2 / B :2

Figure ! Determination of GA for shear fle ible grillage member

5 of 4

Chris Hendy 27/1/06 Modelling bracing systems in grillage analysis of steel-concrete composite bridges

3. Determination of moments and forces in bracing system #he shear fle&ible member deri!ed abo!e is included in the rilla e model and the analysis performed" #he results from the rilla e member are then e&tracted and split into symmetric and anti-symmetric cases correspondin to those in Fi ures 5 and 8 as sho%n in Fi ure 4" Fi ure 4 assumes that load is applied to only the nodes in the rilla e" #his is preferable as load applied bet%een nodes %ill in reality act on a member %ith the stiffness of the slab only, rather than the fictitious member deri!ed abo!e" #he distribution of moments bet%een deck slab and bracin system from this case %ill not then be correctly predicted" #he moments and shears from the t%o cases produced are re-applied to the plane frame models Fi ures 5 and 8 as appropriate and the effects in the bracin members, stiffeners and slab determined" 2/ince the plane frame model is the same for the t%o cases, there is actually no need to subdi!ide the load cases as described abo!e3" #he local effects from loads applied bet%een beams can be e&amined %ith the plane frame model in Fi ure 6" #he %elds bet%een stiffener and flan e need careful attention" #he moment transferred bet%een slab and stiffener may also result in uplift on the studs" 9nother difficulty is determinin the number of studs in!ol!ed in transmittin the moment bet%een slab and stiffener" #his depends on the studs locations %ith respect to the stiffener and the out-of-plane bendin stiffness 2and stren th3 of the flan e" 6etailed uidance on the effecti!e len th of flan e to consider, and thus the number of studs on this len th, can be found in => 1115-1-? clause 6"2"6"8" !. "hec#ing $elds at tops of stiffeners and shear studs 9s noted abo!e, the analysis is likely to predict uplift on the studs" /trictly, the studs should be checked for this effect in addition to the lon itudinal shear resultin from composite action" +n reality, only a fe% studs in the !icinity of the stiffener %ill be affected" #his %ill represent a small proportion of the o!erall lon itudinal shear resistance pro!ided, so e!en if these studs are fully utilised in tension, any effect on reducin the effecti!eness of composite action in the main beam %ill be small@ redistribution of lon itudinal shear can occur alon the beam" #he fati ue check of the %eld is usually the critical one" =&perience on sAuare or li htly ske%ed multi- irder brid es indicates that 12 mm fillet %elds usually suffice bet%een tops of sin le-sided stiffeners and top flan e, unless the stiffener outstand is unusually small" For ske%ed brid es, the predicted %eld si(e can be reater" +f the %eld si(e becomes unmana eable, it may be possible to pro!ide a usset as in Fi ure 1 2but not on the outer face of e&ternal irders for aesthetic reasons3" 7utt %elds bet%een stiffener and top flan e should enerally be a!oided because of the risk of flan e distortion" %. Things to bear in mind & 'imitations of the method $oment is transferred from slab to bracin system by bearin on the stiffener and uplift on the shear studs" #ension actin on the shear studs does not i!e a completely ri id response due to both slip and bendin of the flan e plate" #his means that the stiffness is likely to be an upper bound B conser!ati!e for

8 of 4

Chris Hendy 27/1/06 Modelling bracing systems in grillage analysis of steel-concrete composite bridges

bracin desi n but sli htly un-conser!ati!e for lobal distribution and hence main beam desi n" #he reater the %idth of slab used in this model, the less the moments that are attracted to the bracin system"

F $a $b

0"42$aC$b3

0"42$aC$b3

F 0"42$a-$b3 Figure % (eparation of grillage results into cases to re-apply to models in Figures 3 and ! 0"42$a-$b3 F

Figure ) (eparate plane frame to e amine local effects of $heel load bet$een stiffeners *repared+ "hec#ed+ Date+ Chris Hendy Chris $urphy 51/5/06

4 of 4

You might also like