We are working to design the roof of a high-temperature tank and had the following concern:

If you are designing a supported roof (1, 2, 3 or more span), how does one account for the reduction in allowable stress of the rafters? There is a stress reduction applied to the rafter spacing (it has the yield stress taken into account in the formula). However, if at high temperatures, allowable stresses are reduced, but beams for rafters are selected on a Moment Resistance criteria according to the CAN/CSA-S16.1 regulations, how can one account for this? Instinctively, you would need a stronger beam.

The solution we came up with is based on the general beam stress formula: (Fy)(I)/C=Mmax

We replace (Fy) with (Fy*Mfactor), where (Mfactor) is the allowable stress reduction factor found in Appendix M. We can see that it should be valid to divide a given Moment Load Case (Mmax), calculated via uniformly and linearly distributed loads, by the (Mfactor). This would effectively account for the reduced stress allowance, but in a form in which we can utilize beam selection tables and the Moment Resistance Criteria, Mr & Mr’.

Appendix M only mentions allowable stress reductions in general, as well as self-supporting roofs. For self-supporting roofs, a ratio of elastic modulus at the design temperature to 199GPa is used to modify the calculated required roof thickness.

I am asking whether or not there is any agreement that our solution to account for Moment Load Cases at high temperatures is valid.

Please let me know if you require any clarification on this question.

Thank you.