C
HORIZONTAL SHEAR CONNECTIONS IN COMPOSITE BEAMS
OMBINATION
of precasting and in-situ concreting often proves to be the most economical method of construction. The structure so pro- duced consists of prefabricated con- crete beams (either reinforced or prestressed) connected with a cast in-situ concrete slab.
An essential requirement of com- posite construction is that the precast and the cast in-situ components act together as one unit under all loading conditions. Incomplete monolithic action is caused by relative movement at the interface, in the form of horizontal slip, or vertical separation, both. To maintain complete
or
monolithic action at all loads, the connection between the two com- ponents must be strong enough to resist the horizontal shear and the vertical tension, if any, acting at the interface,
The horizontal shear stress at the interface is estimated with the conven- tional formula:
S
where s
=
V =
Q
I
b
=
VO
bl
the horizontal shear stress at the interface:
the total external shear force at the section; the first moment of the transformed composite section on one side of the interface about the neutral axis of the com- posite section;
the moment of inertia of the transformed com- posite section:
the width of the section
at the interface.
The validity of this formula is doubtful when applied to a cracked section, especially under ultimate load condition. Nevertheless, in the ab- sence of a more realistic assessment, the "nominal" horizontal shear stress given by this formula is generally ac- cepted as an approximate measure of the actual stress.
According to the classical theory. no stresses develop in a composite member in the vertical direction. However, secondary stresses occur which sometimes produce tension. These secondary stresses are smali, but even so they might cause vertical separation at the interface if there is no reinforcement crossing the joint. There is no method yet to assess the magnitude of the probable vertical tension, or even to locate where it might exist.
Far East Architect & Builder February, 1967
by Dr. H. W. Chung
Department of Civil Engineering, University of Hong Kong
Types of Horizontal Shear Connections
In a composite member, transmis- sion of horizontal shear across the interface can be achieved by three
means:
(i) concrete-to-concrete bond: (ii) shear keys; (iii) steel ties.
Concrete-to-concrete bond
In a composite beam, concrete-to- concrete bond always exists at the interface unless it is locally destroyed through poor workmanship. The surface of contact may be smooth or rough. A rough surface is preferable because, firstly, it gives a larger area of contact, thus increasing the effi- ciency of bond; and, secondly, it promotes an interlocking action be- tween the precast concrete and the cast in-situ concrete.
While a smooth surface can be easily defined as one which has been cast against a form, trowelled or floated, a "rough surface" is a vague term. It does not mean a surface at which there has been no consolida- tion, but rather one made up of coarse aggregate particles about half embedded in mortar. A rough sur- face is best achieved by brushing off a surface layer of chemically retarded
concrete.
However, to produce an exposed- aggregate surface is a tedious job which must be done at the right time after casting, and this might not be convenient in practice. An alterna- tive method is to scrape the surface of the consolidated fresh concrete with the edge of a sheet of steel. This is undoubtedly an easy treat- ment, but care must be taken to avoid over-scoring of the surface. Some engineers advocate a cross-tamped surface, generally formed with a wooden tamper which gives con- solidation right up to the top of the beam and adequate ridging in the direction of stress. The cross-tamped surface is claimed to be rough enough to effect good bond, despite the fact that all the coarse aggregate particles are smoothed into the mortar.
A simple way to determine the shear capacity of a particular type of contact surface is by push-off tests. Though the stress condition at the
interface of a composite beam cannot be precisely simulated in a push-off specimen, the results of push-off tests do give a rough indication of the be- haviour under load and the ultimate strength of the connection. Such tests have been performed by Hanson at the Portland Cement Association (U.S.A.),1) and the findings concern- ing bonded concrete surfaces are as follows:
bonded (i) A
surface (either smooth or rough) develops its maximum shear resistance with very little slip (less than .002").
(ii) The shear strength of a smooth surface is about 300 lb in2.
(iii)
The shear strength of a rough surface is about 500 lb in2. (iv) The shear strength of a rough surface seems to be insensitive to variations in the depth of roughness.
Since 1950, a number of composite beams have been loaded to destruction in Britain, Germany and the U.S.A. Most of the test specimens showed no sign of distress at the joint even under the ultimate load. indicating that the bonded surface is a reliable connection. Available information on the beams that failed at the joint is summarized in Table 1. The ultimate horizontal shear stresses in this table show poor agreement. the scatter being greater for the beams with a smooth interface than for those with a rough interface. Moreover, these stresses are in general lower than the corresponding values from Hanson's push-off tests.
An important point is obvious in the Table: that the connection is markedly strengthened by a small amount of steel placed across it.
Shear Keys
Rectangular shear keys provide no resistance to vertical separation at the interface other than the concrete-to- concrete bond. On the other hand, dovetailed shear keys tend to tie the cast in-situ concrete to the precast concrete, the effect being similar to that of stirrups extending across the joint. However, formation of dove- tailed shear keys on the concrete surface is a time-consuming job and thus not practical.
49