Thermal Movement in Portal Frames

[nd0586]

Hi there,

I've been using eng tips as a resource for a while, but this is the first time i haven't found my question already answered in another thread!

I'm designing a large (9x20m bays) portal frame, and the software that I'm using to model the frame doesn't seem to have the option to input temperature variations. This (and a lack of discussions about the topic on the internet/in my office) suggests to me that this is generally ignored in portal frame design. However, I'm worried about locked in thermal stresses and thermal movements in the plane of the frame. I know that the axial forces in the rafters are fairly critical to portal frame design, and I think that a combination of two things will happen, depending on stiffnesses -

- The stiffness in the frame will lock in additional axial compressive forces in the rafters (during thermal expansion), causing failure of the frame at lower loads than predicted by the portal frame modelling software
- The thermal expansion/contraction will not be sufficiently restrained by the stiffness of the frame and will cause the total deflections to increase beyond acceptable limits.

Does anyone have (preferably!) an argument to put forward to assuage my concerns, or a suggestion of how to proceed with the design? My thoughts at the moment are to do a hand calc on the expected thermal movement, and apply the forces by hand in the model to see what happens...

Thanks!

 

[hokie66]

Portal frames are not stiff enough to restrain the members against thermal volume change. You can therefore model this by determining the change of length and imposing a deflection accordingly. You are correct in your assumption that this is usually ignored in portal frame design. For a 20 m span, it doesn't make much difference.

 

[nd0586]

Thanks for your reply - that's good to know! When I said 9x20m though I wasn't very clear - I meant 9 bays in a row i.e. 180m long in the direction of the portals (by around 220 m in the other direction, but bracing in that direction is another topic...) - is it still acceptable to just impose temperature deflections at the ends of the portals?

 

[hokie66]

That is a different story indeed. I thought you meant a 20 m span for one portal, and the bay spacing was 9 m. Fairly typical.

For your large expanse of roof, you will need expansion joints. There are a number of threads on the site, including some with references, about expansion joint spacing. I would look into dividing your roof into separate structures, each 60 x 55.

 

[kieran1]

20m is very large bay size considering the size of secondary steel that would be required. Ie: purlins and sheeting rails.

It all depends on the expected temperture variations and co-efficent of thermal expansion.

For a structure this long i would typically split into three sections with two movement joints.

Kieran

 

[nd0586]

Thanks for your answers!

If my cladding is designed to take the thermal movement from the steelwork, and the portal frames are not stiff enough to cause any locked in stresses, then if i design the columns for the resulting out of verticality would that not be sufficient?

I agree that a movement joint would be preferable, but I am trying to avoid them if possible.

Expected temperature variations are pretty low (UK) - assuming a conservative 40 degree variation then the max movement i would be likely to get would be around 45mm - within horizontal tolerances for my column heights.

 

[hokie66]

The following is one well regarded reference for expansion joint location.

http://www.modernsteel.com/Uploads/Issues/April_2005/30740_fisher.pdf

I think avoiding expansion joints in a portal framed and braced building 180 x 220 metres is destined to result in despair.

 

[nd0586]

Thanks! I'll have a read through.