Pre-engineered Columns for Compression

[EngDM]

I am in the process of analyzing a pre-eng frame for a new loading case. I've calculated the column compressive resistance as per CSA S16 13.3.2 for flexural buckling, however in my case the web slenderness ratio forces me to use 13.3.5 and reduce my yield strength to Fye = 83MPa. I have the component/loading sheets for the existing pre-eng, and calculated the load on the columns based on the provided reactions, and the reaction well exceeds the resistance calculated using the effective yield strength. If I calculate it assuming no local buckling of the web, the column is nearly at 100%, which given how tight pre-eng usually gets designed doesn't surprise me.

I am wondering how they get around this slenderness check? I looked into the "Limit States Design in Structural Steel - Kulak, Grondin" hoping that maybe since it's a plate column there would be different requirements, but unfortunately it appears that the same slenderness ratios.

Anyone ever run into this before?

 

[lexpatrie]

Software problem that is "too expensive to fix" and hasn't lead to any documented failures? MBMA (I forget the name of the software) is a dark dark black box.

 

[EngDM]

The column in question is loaded by two crane's on outriggers as well as bearing roof loads. The total reaction is like 1600kN factored, and the resistance with the reduced Fye is 820kN. Is it actually possible that this column has been standing 20 years and hasn't been loaded above 50% of the max possible loading? Not to mention that it's a beam column due to the crane lateral impact loads so the utilization probably well exceeds 200%.

 

[SandwichEngine]

Your results sound off to me. Are you considering the girts and flange braces as weak axis restraints for bending and compression?

 

[Smoulder]

Not sure about Canada but there's a method where you take effective section as excluding the middle part of the web that's making it a slender section. So you can still use full yield on reduced section. Sounds like that would be a much bigger number for you.

 

[EngDM]

Not sure about Canada but there's a method where you take effective section as excluding the middle part of the web that's making it a slender section. So you can still use full yield on reduced section. Sounds like that would be a much bigger number for you.

Canadian code offers 2 methods. Either reducing Fy, or reducing effective area like you said. It wasn't until the most recent edition of the code where it actually specifies not to use the Fy method for anything other than circular shapes. Using the area method my capacity is much larger.