triangular trusses

[dougantholz]

I have a long span (120') porte coche roof truss that is triangular in cross section. There are two bottom chords, one top chord of pipes, with pipes web members tying everything together.
My question is this: does anyone have a reference for designing such a truss accounting for overall stability, especially lateral-torsional buckling of the whole truss?
Any reference help is greatly appreciated.

 

[JAE]

I'm not so sure you have an LTB issue. The triangular truss is similar to a closed triangular tube, (the diagonals serving to transmit horizontal shear to the chords) so the main concern would be correctly modeling the truss and its applied loads, taking into account eccentricity of loading on the truss c.g.

Curious as to what others think here.

 

[csd72]

If your analysis package can do buckling analysis then you can get your effective length from this and then design it by hand.

 

[doejohn]

I'm inclined to agree with JAE about LTB--the section is torsionally efficient, so there isn't much tendancy for any localized "compression flange" instability to develop. The truss stability should be controlled by the global bending and torsion buckling limit states. If this is a concern, then your buckling analysis method should include the effect of shear deformation (e.g. modeling the truss using discrete elements for the chords and web members in an analysis package), as my Timoshenko reference indicates that this can be the source of considerable difference between elastic bending buckling of a solid section and that of a "laced column" of the same composite EI.

 

[Denial]

LTB buckling tendency will depend upon how acute-angled the triangle is.

 

[271828]

To have LTB, it must be stiffer about the bending axis than the other one. It doesn't sound like you have that.

It sounds like you're worried about the top chord flexurally buckling about a vertical axis.

Like one of the others typed, you could do an eigenvalue buckling analysis if you have a program like SAP. Just make sure you do a simpler model first to make sure you know what it's doing and how to use it.

You could also use the AISC Appendix 7 Direct Analysis Method. Explicitly model the top chord to be out-of-straight (bent laterally), use the reduced stiffnesses of the DAM, and run a rigorous second-order analysis. That should tell you if it's stable for the required load combos. Just looking at your ratio of second-order to first-order deflections, forces, etc., will give a good feel for the truss stability.

It would be interesting to see how the eigenvalue approach and the DAM approach compare.