scissor joist support

[LB BE]

I have searched the other threads on this topic. However, I don't see a real consensus out there and I am not able to find any research or studies on the topic. I have steel scissor joists bearing on steel beams and columns. I want to determine the horizontal load to design the supports. If the joists are designed with pin-roller connections and a max horizontal deflection is requested, is there a horizontal load to design the supports for?

 

[Agent666]

If the supports (column or beam) are free to move to accommodate the spreading of the truss, then they won't see any horizontal load. If both of the columns are not free to move over (consider things like return walls propping the columns for example), then there will not only be a horizontal force, but also the truss forces will potentially be quite different.

I believe both columns would require some constraints to them moving to stop the spread of the truss, if one side is restrained all the spread simply goes to the unpropped side (assuming roof diaphragm is not present). If they are supported on beams, then make sure you account for any minor axes moment generated from the horizontal thrust being applied to the top of the beam. The answer can also depend on how they are erected, lifting them in place might release a lot of the self weight spread for example, so an assessment of the forces that come after everything is lock in might also be appropriate.

 

[hokie66]

If the "scissor joist" is a scissor truss, the thrust should be resolved within the truss, much the same as for a truss with straight chords. Now if this is instead a kinked beam, the thrust and/or the horizontal deflection have to be catered for.

 

[LB BE]

Agent666,

I do not have any constraints on the columns. This is kind of a free standing structure. I was planning to fix the bases.
What horizontal load would I use to design the beam if all I have is a horizontal deflection?

 

[LB BE]

hokie66,

The roof is a metal deck on steel bar joists. The bar joists are labeled scissor joists by SJI even though they are really sloped parallel chord bar joists. The joist supplier has asked for a horizontal deflection limit to be given in the specifications. Are you thinking there is no horizontal load to design the beams and columns for?

 

[jayrod12]

They are asking for an allowable horizontal deflection to see if they can get the joist to work without restraining thrust. If you allow for some lateral deflection, then the horizontal loads are resolved within the joist.

Give them whatever you feel is a reasonable allowable horizontal deflection of the beam. For me I would probably start at H/500 or H/400 depending on the wall finishes.

 

[charliealphabravo]

If these are parallel chord joists (and not scissor trusses) then I don't think the thrust forces can be resolved within the members. Is there a beam along the ridge?

 

[LB BE]

jayrod12,

For allowable horizontal deflection of the beam, I was going to use 1/2" (which is about L/400). But the support is stiffer at the columns and that 1/2" might really be in the center of the beam span. Or is the theory that the whole beam and column system deflects horizontally the 1/2" along the whole length?

 

[LB BE]

charliealphabravo,

This is a special steel joist shape that is one piece per the Steel Joist Institute. No beams along the ridge.

 

[charliealphabravo]

It sounds like you need to specify the maximum amount of horizontal deflection that you can live with at any point along the bearing beam to avoid damage, code limits, etc. The actual deflected shape of the wall will depend on spans, base fixity, length of the wall etc.

The joist designer will attempt to resolve the thrust forces internally so that the horizontal deflection doesn't exceed that amount under design loads. The steeper the scissor and the shallower the joist depth the harder that will be since it takes more steel to stop the spread of the legs.

As long as the wall is softer than the joist (most likely case) then the actual joist-to-beam bearing condition doesn't matter much since the wall will just move outward until the truss bottom chord goes into tension (hopefully within your deflection criteria). There is no actual rolling at the support (Edit: and no actual thrust load resisted by your wall).

If the wall is stiffer than the joist at design load/deflection then you would either want to allow for rolling at one bearing (probably preferred scenario in which case your wall again sees no load) or your wall will need to resist the spreading load of the joist and there will be a stress reversal in the joist (probably not what you or the joist designer is expecting as these loads can be large).