Question: I have an HSS 203 x 76 x 3.09 mm (Class 4 cross-section) with a known F_y of 426 MPa (from experimental coupon tests). To obtain the axial compressive capacity, I referred to CSA S16-09 Clause 13.3.5 dealing with “Members in compression subjected to elastic local buckling.” I evaluated the capacity based on the two methods given – (a) and (b) – but I get two vastly different results. Which one is more correct?

Answer: Initially, it should be reiterated that a measured yield stress is not permitted to be used in structural designs – the minimum guaranteed (or nominal) yield stress must be used.

As implied in this clause of the Canadian code, both methods for handling slender cross-sections are permissible:

Method (a) is based on the determination of an effective area, calculated using reduced element widths. This can be used for square and rectangular HSS, and box sections, but not for round HSS or pipes.

Method (b) is based on computing an effective yield stress for the whole cross-section and can be used for any shape of cross-section.

The two methods do not give consistent results for square/rectangular HSS. If the member is stocky with respect to overall flexural buckling (i.e. has a low KL/r) and fails by yielding, computing an effective (reduced) area – method (a) – whose elements satisfy the local buckling requirement, gives the greater buckling capacity. If the member is slender with respect to overall flexural buckling (i.e. has a high KL/r) and fails by Euler buckling, computing an effective (reduced) yield stress – method (b) – that satisfies the local buckling requirement, gives the greater buckling capacity.

In summary, it is safe to compute the member compressive strength of a rectangular HSS by using both methods and to take the larger of the two.