I have designed a 80m box truss with StaadPro. For the top chords that are in compression, do I need to specify their Lz to be equal to 80m? If I do that, the top chords will fail terribly due to slenderness. Is it a valid assumption to default the Lz for the top chords to the member length between nodes in this circumstance?
I have attached the relevant file; will appreciate some advices.
Hi John,
Have a look at Table 22 of BS 5950. I believe, strictly speaking, yours would be having a partial restraint in direction at both ends, so I'd take 0.85L would be the effective length. However, taking it as 1.0L would be a quick and conservative approach. Staad Pro will only help you check the buckling (actually the reduced compression strength, taking into account buckling effect) for a local member. You'll also need to ensure that your compound member (i.e. the whole truss) does not buckle, and normally a quick hand calculation would suffice.
Some general comments:
1. Truss analysis is a simplified method of analysis, something which I think, with computer analysis, we can do without.
2. Based on your model, you could have specified a member as a truss member, instead of specifying member start and end releases.
3. With steel design, be careful on how you detail your structure to achieve the support condition that you assumed, i.e. the roller condition in X-direction.
Hope this helps.
Hean