paper: byfield

Analysis of composite beams with widely spaced shear connectors

Synopsis
BS 5400: Part 5 requires consideration to be given to the nonuniform flow of longitudinal shear and to the slippage between the slab and steel member of composite beams, where the longitudinal spacing of shear connectors exceeds 600mm. Analysis of such effects is complex and involves finite-element analysis with contact logic. This paper presents a simple spreadsheet based alternative solution. The method is suitable for serviceability limit state (SLS) checks and assumes that longitudinal shear is transferred between the steel member and slab only at shear connector locations. This method can be used for the assessment of existing bridges with widely spaced shear connectors. It can also be used for designing the strengthening measures for bridges built before composite construction was widely recognised. analysis of the inter-connector stresses. The design moment (Fig 2a) is applied to the combined sections with the shear connection removed (Fig 2c). It is assumed that the slab remains in contact with the section. Thus the radii of curvature of the beam (Rs) is equal to slab (Rc) + the distance between the two neutral axes. Since Rc Rs the analysis can be simplified by assuming that Rc= Rs= R. Analysis accounting for differences in radii is tedious and alters final stresses by less than 1%. From simple bending theory the moments in the concrete slab (Mc) and steel beam (Ms) are: Mc= and Ms= Es I s R ....(2) Ec I c R ....(1)

M. P. Byfield
BEng, PhD, CEng, MICE
Engineering Systems Department, Cranfield University, RMCS Shrivenham, Swindon, SN6 8LA, UK Received: 14/09/01 Accepted: 19/02/02 Keywords: longitudinal strain, shear, slip connections, composite beams, slabs, analysis. spreadsheets, serviceability, bridges

Introduction
The full interaction transformed sections method assumes that a continuous shear connection exists between the concrete slab and steel section in composite beams. Therefore, it is not suitable for carrying out the SLS checks required by BS 5400: Part 51 in the assessment of bridges where shear connector spacing exceeds 600mm. Such checks are required because of the non-uniform distribution of longitudinal shear in the connectors and because of the potential for slippage between the steel and concrete. This work presents a method for assessing the effect shear transfer discontinuities have on bending stresses and shear connector forces. The effect of discontinuities in the shear transfer in composite beams was considered in the pioneering work on composite beam design by Caughey and Scott in 19292. Later experimental work showed that imperfect shear connection leads to slip at the steelconcrete interface and a method was developed to analyse this incomplete interaction3. More recently a method has been developed for designing composite beams with a relatively low degree of shear connection4. Both methods implicitly account for the slip created by the flexibility of the shear connectors, although the shear connection is assumed to be continuous along the length of the beam. By contrast, the method presented here assumes longitudinal shear is only transferred between the sections at discrete locations along the beam. The method may be classified as full interaction because shear connectors are assumed to be rigid under elastic conditions.

Since M = Mc + Ms and R = Ey/, the stresses in the bottom of the concrete (c.b) and top of the steel (s.t) can be easily calculated, i.e.:

The method
The method is illustrated in Fig 1. In brief: Step 1: The shear connection between the slab and beam is released. The bending stress distributions are then quantified assuming no shear is transferred between the slab and the steel. The slip between the two surfaces is calculated once the bending stresses are quantified. It is assumed that the concrete can resist tensile stresses at this stage, although, in the final state the tensile stresses become compressive stresses, when the neutral axis lies within the steel. Step 2: Externally applied loads are removed. Instead, internal shear connector forces are applied to produce a reverse slip of equal and opposite magnitude to the step 1 slip. The stresses in the section caused by the shear connector forces are quantified. Step 3: The stress distributions calculated in steps 1 and 2 are summed to produce the final stress distribution.The force in individual shear connectors is equal to the change in shear connector force. Shear connection released (Step 1) Fig 2 illustrates the method in more detail. An element of beam located between two shear connectors is subdivided to facilitate

Fig 1. Basis of method

Fig 2. Determining slip at the shear connector locations on release 2 July 2002 The Structural Engineer|31

paper: byfield

v c.b = and

M i Ec yc Ec I c + Es I s M i Es ys Ec I c + Es I s

....(3)

v s.t =-

....(4)

remain the same, the equivalent eccentricity of T is independent of moment and dependent on sectional and material properties alone.The method can be adjusted to account for the minor differences in radii of curvature, although results are altered by less than 1%. The force T causes stresses c.b and s.t to be developed in the bottom of the concrete and top of the steel respectively, where: v c.b = - T and v s.t = T

Where: Mi is the applied moment at any cross-section, see Fig 2a yc is half the depth of the slab ys is half the depth of the steel beam (if beam is symmetric) E and I are elastic modulus and second moment of area respectively (the subscripts c and s represent concrete and steel) Given these stresses the end displacement of the concrete (c.1) and steel (s.1) can be calculated from = L/E, i.e.: Ly c M Ec I c + Es I s Ly s M Ec I c + Es I s

e Ac
1 1 +

(y c - e) y c Ic

....(12)

e As

(ys + e) ys Is

....(13)

These stresses cause the end displacements sketched in Fig. 3, where: d c2 = and Lvc.b LT =Ec Ec

d c.1 =

....(5)

e Ac
1 1 +

(yc - e) yc Ic

....(14)

d s.1 =-

....(6) d s2 =

Where the average moment between connector pairs is: M= 1 n

Lv s.t LT = Es Es

e As

(ys + e) ys Is

....(15)

!M
i = 0

....(7)

L is the shear connector spacing n is the number of sections between shear connectors, see Fig 2b. Thus, the total end slip, sm (equal to s1 +s2, Fig 2c) is: s m = d c.1 + d s.1 ....(8)

The total end slip (2.St, see Fig 3) = c2 s2. It is assumed that the stress distributions resulting from the shear connector forces correspond to those produced by the application of a moment and axial force to a long beam. Local stress concentrations at the shear connectors are ignored and strain distributions are assumed to be linear in both the steel and concrete. In simply supported beams, where the maximum shear connector forces occur remote from the maximum bending forces, this assumption will be approximately correct. It is also assumed that no slip occurs at the connectors under elastic conditions and that inter-connector slip produces zero friction. Combine steps 1 and 2 to determine final composite stresses (Step 3) It is assumed that zero slip occurs at the shear connectors, i.e: step 1 end slip + step 2 end slip = 0. Given this the previous formulae can be combined to produce the total shear connector thrust T, at any section, where: + e Ec I c + Es I s Ec Ac M (yc + ys ) 1 (yc - e) yc Ec I c (ys + e) ys 1 + o E s As Es I s
-1

T=

....(16)

Shear connector forces applied to produce reverse slip (Step 2) The externally applied loading is removed from the combined sections and internally applied shear connector forces T are applied to produce reverse slip, see Fig 3. The longitudinal force T is the sum of shear connector forces between the section under consideration and the end of the simply supported beam or point of contraflexure. Physically, the centreline of the force T must be near to the base of the shear connectors. However, the force is also accompanied by an interface normal force to maintain equilibrium5. Thus the moment induced in the slab,Tyc, is reduced by the moment resulting from the normal force.This reduced moment may be accounted for by an equivalent eccentricity, e, of the force T. Thus, the moments in the concrete and steel become: M c = T (y c - e) M s = T (y s + e) ....(9) ....(10)

Fig 3. A section of composite beam located between shear connectors, shear connector force applied and external loading removed

T is equal to the sum of the shear connector forces between the end of the span and the section under consideration; connector shear is given by the change in T.This is illustrated in Fig 4, where the average moment between the first two shear connectors is double that of the adjacent section. Thus, the change in total thrust (or shear connector force) is 1T.

Fig 4. Calculation of the shear connector force The stress in the steel and concrete can be determined by adding the stresses calculated in steps 1 and 2 to produce the following formulae: Final stress in the top of concrete slab: vc.t = M i Ec yc Ec I c + E s I s -T 1 e Ac (yc - e) yc Ic

No gap is allowed to form between the slab and section. The position of the equivalent eccentricity can be determined given that that the normal forces maintain the interface contact. Since Rc Rs =R and M/I = E/R the position of the equivalent eccentricity, e, can be defined where: e= yc Es I s - ys Ec I c Es I s + Ec I c ....(11)

....(17)

Thus, by ensuring that the curvatures of the steel and concrete 32|The Structural Engineer 2 July 2002

Final stress in the bottom of concrete slab:

paper: byfield

vc.b = +

M i Ec yc Ec I c + E s I s M i E s ys Ec I c + E s I s M i E s ys Ec I c + E s I s

-T

1 e Ac +

(yc - e) yc Ic

o o o

Fig 5. (top) Direct stresses in the concrete slab Fig 6. (above) Direct stresses in steel section Fig 7. (left) Section properties of worked example ....(18) Fig 8. Direct bending stresses in a noncomposite steel beam (Fig 7) strengthened through the addition of shear connectors 3m from each end.

structures built before composite construction was widely recognised. Such structures, which could be railway or highway riveted girder bridges, could be strengthened through the installation of shear connectors at discrete locations. The potential of such strengthening is illustrated in Fig 8 which shows the direct bending stresses for the beam sketched in Fig 7. The analysis assumes longitudinal shear is only transferred between the steel and concrete by groups of shear connectors 3m from each end of the 12m long,simply supported beam.The figure shows the partial shear connection leads to a substantial reduction in the maximum direct bending stresses in the steel section. Calculations show that the position of the effective eccentricity of the shear stud force, e, consistently lies just below the centroid of the slab. Since the centreline of the force must be near to the base of the shear connectors, the normal forces in the shear connectors and at the steel concrete interface must be important in inducing this equivalent eccentricity. As shear connector spacing is increased, the tendency for the steel and concrete to separate will increase. Thus, partial shear connection may need to be accompanied by additional connectors along the length of the beam to prevent separation. Fatigue failure of shear studs is known to occur in bridges.This method provides a useful tool for analysing the effect that the loss of shear connection through fatigue would have on the loading on remaining studs. Such an analysis could reveal when stud loss through fatigue would become critical, causing a beam to unzip as the studs that remain become progressively overloaded.Finally, the method can also be used for analysing steel and wrought-iron sections built up from separate components riveted together at discrete points along their length.

Conclusions
A novel release/restore method is presented for analysing the direct bending stresses and shear connector forces in composite beams. The method implicitly recognises the breaks in the shear connection at the steelconcrete interface. It can be quickly programmed using, for example, spread-sheet software and can be easily adapted to analyse any type of composite member builtup from components connected together longitudinally along their length. The method finds application as a means to carry out the BS 5400: Part 5 SLS checks required for the assessment of bridges where shear connector spacing exceeds the normal code limits. Another application includes the design of strengthening measures for bridge structures built before composite construction was widely recognised.

Final stress in the top of the steel section: v s.t = 1 +T eA +

s

(ys + e) ys Is (ys + e) ys Is

....(19)

Final stress in the bottom of the steel section: v s.b = + 1 +T eA s

....(20)

Results
Figs 5 and 6 illustrate the model predictions for a beam (Fig 7) subjected to a dead and imposed load of 54.5kN/m run, with shear connectors at 600mm intervals. The impact of inter-connector slippage on stress is clearly visible, with stress oscillating about a mean value that corresponds exactly to that produced from the method of transformed sections. Both the top of the concrete and the bottom of the steel remain largely unaffected by the breaks in shear transfer.As would be expected, slip has its largest influence where there is the greatest rate of change of bending moment. Thus, the edges of the beam are greatly affected, with slip creating significant tensile as well as compressive stresses. Conversely, slippage has relatively little effect where bending is approximately uniform, as is the case at the centre of the beam. In reality, sharp changes in stress would be smoothed because of the flexibility of shear connectors and the effects of local stress concentrations at the connectors. The shear connector forces predicted from the proposed method correspond to those predicted from a conventional analysis of shear flow. A possible application for the method is the assessment of existing bridges where shear connector spacing is larger than the limits specified in modern design codes, such as BS 5950: Part 5. Where connector spacing (or groups of connectors) exceeds 600mm the code states5 that checks are necessary to take account of the non-uniform flow of longitudinal shear and of the slip between the slab and steel.Previously,no simple method existed for such calculations, so this approach may have application, including those old

REFERENCES
1. 2. 3. BSI.: BS 5400: Part 5: 1979 Steel, concrete and composite bridges: Part 5. Code of practice for design of composite bridges. London, 1979. Caughey, R. A. & Scott, W. B.: A practical method for the design of I beams haunched in concrete. The Structural Engineer, 7(no. 8), 1929, pp 275-93. Newmark, N. M., Siess, C. P. & Viest, I. M.: Tests and analysis of composite beams with incomplete interaction. Proceedings Society for Experimental Stress Analysis, 9, No. 1, 1951, pp 75-92. Oehlers, D. J. and Sved, G.: Composite beams with limited-slip-capacity shear connectors. ASCE Journal of Struct. Eng. Vol. 121, No. 6, 1995, pp 932-938. Oehlers, D. J. and Bradford, M. A.: Composite steel and concrete structural members: fundamental behaviour. Pergamon Press, Oxford, 1995. 2 July 2002 The Structural Engineer|33

4. 5.