Table of contents

Introduction ------------------------------------------------------------------------------------------------2

Objective ---------------------------------------------------------------------------------------------------2

Behaviour of Reinforced Concrete Slabs subjected to in-plane restraint --------------------------3

Compressive Membrane Action in Reinforced Concrete Beam-and-Slab Bridge Decks -------4

Literature Review -----------------------------------------------------------------------------------------6

Conclusion -------------------------------------------------------------------------------------------------8

References -------------------------------------------------------------------------------------------------8

List of Figures

Figure 1.1 Load-deflection curve for slab with membrane forces (Rankin et al. (1991)) -------3

Figure 1.2 Plan and X-section of forces and deformation in a beam-and-slab bridge
deck(exaggerated) by Alan Hons, PHD thesis book, Monash University -------------------------5

Page 1 of 8
Introduction

Reinforced concrete slabs are one of the important elements in most structural systems and
largely employed in engineering practices in a variety of situations and applications. The slabs
are relatively thin structural elements, whose main functions are to transmit the vertical loading
to their supports.

One of the needs of structural designers is a reliable method for predicting the ultimate capacity
of reinforced concrete slabs subjected to uniform distributed loadings. Elastic or plastic
methods of analysis are widely used. For the prediction of loaded slab behaviour, an important
factor that is normally ignored in the design of continuous slabs is the strength enhancing effect
of arching or compressive membrane action.

This report is concerned with the behaviour of reinforced concrete slabs which are subjected
to vertical loads and in-plane translational end restraints. The end restraints induce in-plane
compressive forces within the slab. This is referred to as compressive membrane action
(arching action) and it significantly enhances the stiffness and load-carrying capacity of the
slabs.

This report contributes to the knowledge that currently exists in this field and leads to greater
confidence in the prediction of the load-carrying capacity of reinforced concrete slabs where
end restraints are present.

Objective

The objective of this report is to investigate the effect of compressive membrane action in
reinforced concrete slabs.

Page 2 of 8
Behaviour of Reinforced Concrete Slabs subjected to in-plane restraint.

The solid line in Figure 1.1 shows a typical load-deflection curve for a reinforced concrete one-
way slab where membrane forces are present. This figure shows that from A to B, the stiffness
of the slab is greater than for an identical slab where compressive membrane action is not
present (shown by the dashed line in Figure 1.1). For an under-reinforced slab, the tension steel
will yield somewhere between points A and B. At point B the cracks in the slab associated with
the yield-line pattern have developed and the load carried by the slab is at a maximum. As the
deflection increases after point B, the load carried decreases due to the crushing of the concrete.
The maximum compressive membrane force occurs in the descending portion (between B and
C) of the load-deflection curve (Eyre (1990)). This is followed by the failure of the compressive
membrane mechanism. At C, there are no membrane forces in the slab, and the load is carried
purely in bending. After C, the membrane forces become tensile and cracking extends through
the depth of the slab. From C to D, the load is almost entirely carried by the reinforcement
acting as a tensile membrane. At D, the reinforcement fractures and the slab completely
collapse. The load at D will be higher than B when there is a moderately high reinforcement
content. Tensile membrane action is generally only considered in catastrophic failures of
structures and can only occur if the reinforcement is adequately anchored to the surrounds. This
report is not concerned with tensile membrane action and is only concerned with the behaviour
up to point B in Figure 1.1

Figure 1.1 Load-deflection curve for slab with membrane forces (Rankin et al. (1991))

Page 3 of 8
In general, the behaviour described above occurs for a slab subjected to uniform loading. When
subjected to concentrated loads (as in bridge loading) the slab often fails in punching before
the peak (point B) is reached. The dashed line in Figure 1.1 shows a typical load-deflection
curve for a slab where membrane action does not exist. If the slab is under-reinforced, the
tension steel will yield early on and cracks will develop in the slab. Eventually the concrete
will crush, and the slab will fail. By comparing the two load-deflection curves in Figure 1.1, a
change in behaviour occurs when the slab is restrained, and compressive membrane forces
exist. A significant increase in stiffness and load-carrying capacity can be achieved in the
presence of end restraints.

Compressive Membrane Action in Reinforced Concrete Beam-and-Slab

Many investigations have been conducted on compressive membrane action where the in-plane
displacements at the boundaries have been prevented with very stiff surrounding frames. This
is an idealized condition and does not occur in practical situations. However, it is known that
considerable compressive membrane forces can develop in practical situations such as bridge
decks. Compressive membrane forces can develop transversely across the interior slabs in a T-
beam bridge deck due to the presence of the surrounding beams and adjacent panels (see Figure
1.2). When the load is applied, the edges of the interior panel tend to move outward. This
movement is restricted by the presence of the adjacent panels and the stiffness of the beams in
horizontal bending. This induces a compressive force in the plane of the interior panel and
increases this panel's stiffness and load-carrying capacity. When looking at this process, the
horizontal bending stiffness of the beams, the in-plane stiffness of the adjacent panels and the
details of the connection between the end diaphragm and the beams are important
considerations. These factors will influence the development of the compressive membrane
forces since they affect the end restraint stiffness that exists for the slab.

Page 4 of 8
 Figure 1.2 Plan and X-section of forces and deformation in a beam-and-slab bridge
deck(exaggerated) by Alan Hons, PHD thesis book, Monash University.

Compressive membrane action can also develop in the longitudinal girders of a reinforced
concrete bridge. For this to occur, in-plane restraint for the beam-ends must be provided at the
girder supports. This could be due to the presence of bearing restraint forces or the effect of the
abutments and other support conditions. The development of compressive membrane action in
the longitudinal girders is harder to predict than the development of membrane action in the
deck slab. This is due to the large variability in bridge support conditions that are found in
practice.

Page 5 of 8
Literature Review

The development of the modern theories of compressive membrane action began when
Ockleston (1955) conducted load tests on a reinforced concrete building in Johannesburg,
South Africa.

Ockleston tested a 10ft. 6in. long one-way slab under a uniform load. The slab was continuous
over one edge, with the other edges discontinuous. The first cracks appeared over one of the
supports, followed at a later stage by cracks over the other support and at the underside of mid-
span. These cracks developed further, followed by the yielding of the reinforcement at the
supports, and then yielding of the reinforcement at mid-span. The slab then failed when the
concrete crushed at these locations. The ultimate load turned out to be nearly 3.5 times the total
working stress design load.

The two-way slabs that Ockleston tested were 16ft. by 13.5ft. and 4.5in. thick, including a 1in.
thick layer of mortar on the top surface. These slabs were continuous over all four edges. The
reinforcement content varied from 0.04% to 0.24% of the nominal gross area. The two-way
slabs were also tested under a uniform load. Cracking was first observed in these slabs at the
upper surface near the centre of the short edges at two to three times the working stress design
load. These cracks had almost extended right around the perimeter before cracks appeared on
the underside of the slab. At four times the working stress design load, the negative
reinforcement at the edges of the slabs began to yield. Eventually, all the negative
reinforcement around the edge and the positive reinforcement in both directions at the bottom
yielded and the slab failed. The ultimate failure load was found to be nearly six times the
working stress design load. Unlike his explanation for the behaviour of the one-way slabs,
Ockleston believed this discrepancy was too large to be attributed to the tensile strength of
concrete. Ockleston wrote in his paper that 'this large discrepancy cannot be accounted for by
strain-hardening of the reinforcement or by (tensile) membrane action, and no satisfactory
explanation has yet been found'.

Ockleston (1958) re-examined his results and came up with an explanation for the unexpectedly
high ultimate strengths of the two-way slabs from his previous tests. By studying the changes
of geometry in the slab, during the tests, he suggested that the additional capacity of the slab
might be due to compressive membrane action created by the development of membrane
Page 6 of 8
stresses in the slab. His theory was that, as cracking developed, the neutral axis moved towards
the compressive face of the slab. As a result, further vertical deflections caused outward
horizontal displacements at the slab edges. These displacements were prevented by the adjacent
panels, which formed a stiff surround for in-plane forces. Therefore, compressive membrane
forces developed in the slab and its ultimate capacity increased. In his 1958 paper, Ockleston
produced a mathematical analysis of a slab under compressive membrane action. The analysis
did not predict the ultimate load but could be used to determine from test results whether
compressive membrane action could explain the behaviour of the slab. Ockleston wrote in his
paper that compressive membrane action was most likely to occur in lightly reinforced slabs,
where the deflections were small, and the horizontal spreading of the supports was effectively
restrained. The result was an increase in the load-carrying capacity of the slab.

In 1961, Wood was one of the first researchers to carry out an analysis of compressive
membrane action in slabs. He developed a mathematical analysis of a circular slab with edges
restrained against sideways expansion. Wood stated that the presence of the edge restraint
resulted in a "kind of self-induced pre-stress" and "collapse can only commence with greatly
increased yield moments". Wood wrote that the effect of in-plane translational restraint can be
more significant than the effects commonly attributed to restraining moments.

In 1991, Rankin et al. carried out seven tests on uniformly loaded laterally restrained
rectangular slabs. These results were compared with unrestrained slabs. The basic formulation
of a simple, rationally based method for predicting the increased load capacity of rigidly
restrained slabs are given. The method is shown to give good correlation with a wide range of
test results from various sources. The in-plane restraint was provided by a steel supporting
frame, which was designed to minimize horizontal and rotational movements at the edge of the
slab. A uniformly distributed load was simulated using16 point loads, applied by a 1000 KN
capacity hydraulic jack. AH the slabs had a length of 950mm and a span-to-depth ratio of 19.
The percentage of reinforcement was varied and placed at different effect.ve depths. Test
results showed that the presence of in-plane restraints did not affect the first cracking load, but
significantly increased the post-cracking stiffness of the horizontally restrained specimens. The
ultimate capacities of the restrained slabs were found to be between 1.7 and 5.9 times higher
than similar unrestrained slabs.

Page 7 of 8
Conclusion

1. The load carrying capacity of the slabs is increased by compressive membrane action.
2. Estimates of the rotational restraint stiffness and the horizontal translational restraint
stiffness can be obtained using linear-elastic analyses of suitable beam models, and this
was found to give good predictions of the failure load.
3. A significant amount of compressive membrane action can be expected to develop
transversely in the slabs of actual beam-and-slab bridge decks. The stiffness and
strength of the slabs can then be expected to be significantly higher than that predicted
using methods that do not incorporate the enhancing effects of compressive membrane
action.

References

Ockleston, A. J. (1955). “Load tests on a three-storey reinforced concrete building in

Ockleston, A. J. (1958). “Arching action in reinforced concrete slabs.” The Structural Engineer
Vol. 36(6): pp. 197-201

Wood, R. H. (1961). “Plastic and elastic design of slabs and plates.” Thames and Hudson,
London.

Rankin, G.I.B., Niblock, R.A., Skates, A.S. and Long, A.E. (1991). “Compressive membrane
action strength enhancement in uniformly loaded, laterally retrained slabs.” The Structural
Engineer Vol 69(16): pp. 287-295.

Rankin, G.I.B. (1997). “Arching action strength enhancement in laterally-restrained slab

Hon, A. (2003). “Compressive membrane action in reinforced concrete beam-and-slab bridge

Page 8 of 8