Fly brace supporting Purlin 3

[Italo01]

Hello,

I participate in a Whatsapp group with some engineers of my region(Brazil) and one of them said that he uses fly brace to reduce the span of the purlin. He introduces the fly brace on the software and the purlin acts as a continuous purlin.

I said that consider this wrong because the bottom chord of the truss is not adequate to receive the horizontal load and that he may introduce torsion on the truss, but he doesn't believe it because the analysis show no problem when he introduces the fly brace and the purlin, who otherwise fails, pass.

I think that the loads are balanced and the horizontal loads are absorbed by the opposite fly brace and suggested, without success, that he analyses the model with live load on only one span.

What do you guys think of this?

I've seen that a lot of the engineers on the group agree with him and do the same.

 

[human909]

The double fly brace will load the rafter, but not really consequentially.

Agreed. I see no issue with it, I was merely pointing it out as it matters in the context of the original question.

I think by incidental they’re saying you can safely ignore the fly brace supporting the rafter, not that the fly brace provides literally zero support.

If that is what is being said then I guess I'm in furious agreement. I do agree you can safely ignore modelling a fly brace as supporting the rafter, I've designed structure with fly braces and I've effectively ignored the load path in them while allowed for their presence with flange/chord restrain of the rafter/truss. I believe this pretty much the approach other here have outlined.

That said, some people here have then made the leap to suggest since their model doesn't include rafter loads on the fly brace, the fly brace is therefore not experiencing loads imposed by the rafter. Which should be a clearly erroneous assumption.

Why does this matter? It doesn't really. Except when you start using the same argument to justify rejecting the initial question posed. Arguing that a fly brace doesn't reduce the span length, when in fact it can do exactly that.

 

[Italo01]

Human909 is correct. I'm sorry for creating the thread and disappearing but i was too busy on these last two weeks. I created a model to check and, for the purlin, the balanced situation is more critical than the unbalanced. The Fly rafter reduced considerably the bending moments and displacement of the purlins.

Thanks Guys.

 

[human909]

Thanks for running the model and showing the significant reduction in bending moments and deflection.

I'll likely still be safely ignoring my fly braces in my models and considering only as flange restraints. That said I do believe this is a good exercise. You cannot always safely ignore redundant load paths. In some cases taking what you might believe is a conservative assumption may cause load to go where you don't expect it causing issues in strength or deflection criteria.

One of my early demonstrations of this was when a single story roofing system was used by other engineers with slender HSS columns. The rafters ran over the top of the columns with what was effectively a rigid or semi rigid connection, but it was modelled as a pin connection. Instead of only being a load path for axial load the columns took some of the bending moment from the rafters. Having columns visibly bending in your building is not what the client wants!


Likewise in the case of the fly braces. If you size them ONLY strong enough to withstand the 2% of rafter flange load then you could end up in a situation where the fly brace begins to buckle under high roof loads. (It is unlikely that this would get noticed though as you only would need some very minor elastic buckling of the fly brace before more load transfer to the purlins occur.)

As soon as we supply multiple load paths then this can get complicated rapidly. Statically determinate frames are much more simple.

 

[Tomfh]

The possibility of the fly brace buckling due to purlin load is perhaps another good reason for twin fly braces, rather than the cheaper single side only option.

Re. The bent columns, our office made that blunder too in the distance past. A long truss on top of slender tubes. The tubes visibly deflected as the truss sagged into position.

 

[hetgen]

I don't think you can have it both ways.

Either the truss has sufficient lateral torsional rigidity to provide support to the purlin in bending or the purlin has sufficient flexural stiffness to support the truss for LTB.
The fly brace is just an axial link.

You can't have your cake and eat it!

 

[human909]

I don't think you can have it both ways.

You can have it in any way you can rationally analyse it. A double fly brace does accomplish it 'both ways'.

Either the truss has sufficient lateral torsional rigidity to provide support to the purlin in bending.....

When you have a double fly brace the rafter/truss doesn't need lateral torsional rigidity.

You can't have your cake and eat it!

Why not? It is pretty routine to use our flooring system as lateral restraint on major beams, load bearing surface and as a diaphragm for lateral loads.

I'm a big fan of having cake and I might as well eat it too!

 

[hetgen]

When you have a double fly brace the rafter/truss doesn't need lateral torsional rigidity.

That is true if the purlin has sufficient stiffness and acceptable vertical deflection spanning between points A & B.

If the purlin doesn't have sufficient stiffness spanning between pint A and B then the truss bottom chord should not be considered as laterally braced on point Z.


The statement below by OP's colleague implies the purlin span between points X & Y. I don't think that is entirely true, but I agree that, for uplift load, the fly brace acting in tension could help transfer some of the load from the purlin spacing between points A-B to the adjacent purlin between B-C.

....that he uses fly brace to reduce the span of the purlin.

...He introduces the fly brace on the software and the purlin acts as a continuous purlin.

 

[human909]

That is true if the purlin has sufficient stiffness and acceptable vertical deflection spanning between points A & B.

If the purlin doesn't have sufficient stiffness spanning between pint A and B then the truss bottom chord should not be considered as laterally braced on point Z.

These two stiffness requirements are not directly correlated so such an assertion cannot be made.

I won't respond with alternative definitive statements because so much depends on the individual circumstance. Once we start dealing with continuous purlins, dual fly braces as supports and multiple bays we have a quite a complicated statically indeterminate system.

But what I will say is that you analyse the structure in a rational way with a reduced span of the purlins by accounting for the purlin load that the fly braces support. Whether this is a worthwhile approach is another question which I believe I've already commented on.