Historical Considerations

Up to the 1980s, the Canadian shear design provisions directly paralleled the US provisions of the American Concrete Institute (ACI) Building Code Requirements (ACI Committee 318, 2005). These provisions, which are still in use in the United States today, were empirically based on test results from a large number of shear experiments, generally of small size and with a large amount of longitudinal reinforcement.

Staring in 1984, the Canadian provisions (CSA, 1984) began to diverge from the ACI rules. Specifically, the Canadian provisions became divided into a “simplified method” and a “general method”. The simplified method was essentially the same as the ACI provisions but the general method was a new set of design rules based on the Compression Field Theory (CFT) (Collins, 1978; Collins & Mitchell, 1980). This general method required the explicit evaluation and satisfaction of constitutive and compatibility equations. While the method was significantly more complex than the ACI provisions, it did apply to general loading conditions and would usually produce more economical designs. Unlike the ACI provisions, but like the then equally new CSA strut-and-tie provisions, the engineer needed to explicitly select the angle of diagonal compression in the web of a member. Because of this, different engineers would produce different designs for the same applied loading. While some engineers appreciated the freedom that this allowed, others were uncomfortable in that it became more difficult to check each others work; more than one solution to the same problem could be equally correct. An additional limit to the method was that it could not determine the strength of members without transverse reinforcement such as slabs.

In the 1994 CSA shear design provisions (CSA, 1994), a number of changes were made. For the simplified method provisions, a size effect was added to the shear strength of members without transverse reinforcement. This resulted from the experimental observation that members of a larger overall depth tended to fail at lower shear stresses, which was relevant for thick slabs for example. The 1994 general method of shear design was based on the Modified Compression Field Theory (MCFT) (Vecchio and Collins, 1986) and now provided a single design solution for a given set of applied loads. The use of the MCFT instead of the CFT also extended the general method to the design of members without transverse reinforcement, including the size effect. To use the new general method, however, the engineer needed to consult tables of values of в and в, the two main variables for the shear strength of members. Partly due to the difficulty of using table-lookups with spreadsheets, the general method was still significantly more difficult to use than the simplified method, particularly for the analysis of existing designs. Despite this, the method has allowed many impressive structures to be built that exceed the limits of the ACI code. As just one example, Figure 3 shows a 176 metre tall tower in Toronto designed by Yolles Partnership which has high demands on the lateral load resisting system due to the narrow nature of the tower (a height to width ratio of 11:1). Were this tower designed according to the ACI shear provisions, it likely could not have been built as economically since significantly larger coupling beams would have been required by that standard. The necessary increase in floor-to-floor height would have lowered the sellable number of condominium units for the same height of building. Similar shear provisions to the CSA general method were also adopted by the Ontario and Canadian Highway Bridge design codes (OHBDC, 1993; CHBDC, 2000), and the US AASHTO LRFD Bridge Design Specifications (AASHTO, 2004).

Despite the advantages of the CSA 1994 general method, there remained difficulties with its practical use. As noted above, the difficulty of using tables of values with spreadsheets made the method difficult to automate for day-to-day design work. In addition, it was difficult for engineers to develop confidence in the values given by the tables of в and в as it was not clear where these values came from or why they varied the way that they did.

For the generation of the 2004 shear design provisions, a conscious decision was made to make the provisions as general as the 1994 general method, but as easy to apply as the 1994 simplified method. The intention was to make design with the general method non-iterative and to replace the tables of behaviour with simple equations. As an intentional side effect of doing this, it became possible for the simplified method to be a special case of the general method rather than being based on the ACI provisions. The 2004 provisions remain based on the MCFT but are now simple enough to explain and use that they represent a significant advance in research on shear.

Fig. 3. CSA 1994 general method allowed this tower Fig. 4. Basic shear resisting mechanism assumed in to be more economical. 2004 code.