Modeling Temperature Gradient Across Thickness of a Plate Element

[dcarr82775]

I am trying to model a temperature change through the thickness of a plate element to simulate edge curling. Far as I can tell you can not specify a temperature gradient through the thickness. So I am trying it through some frame elements above the plates with rigid links down to the plates to force shortening in the top. Anyone have any experience trying this in RISA?

 

[JoshPlumSE]

You cannot currently create that type of loading in RISA-3D. Though we have had that request before. I will add your request to the list of folks who have asked for it. Hopefully we can do that one of these days.

Personally, I'm not truly familiar with the theory upon which that loading would be based. But, once a reference is obtained, I can't imagine the programming would be all that difficult. If you have a technical reference on the subject that I can point our developers to, that might help. Not saying that I can promise that it will be added in a particular time frame. Just saying that if I can anticipate (and deal with) any "bumps in the road" along the way that can only help the development process.

 

[dcarr82775]

Thanks Josh. No technical paper, I am working through a process in my head, and hoping somebody else had already beat their head against the wall.

 

[JoshPlumSE]

I'm sure someone has. I ran into the same issue 20 years ago when I was working for a heavy industrial engineering firm. Back then I'm not sure any of the structural software we had did this. And, my company had SAP, STAAD, RISA.... GT Strudl might have done it, but I don't remember exactly. I have to believe that someone has added this capability since then.

The derivation of the equivalent nodal forces for this will likely be pretty tricky. But, ultimately, it's just math. And, once it's been done it never has to be done a 2nd time. What concerns me is that this is likely something that would be really tricky to verify. So, we'd have to develop a pretty good testing / validation plan to make sure that our numbers were truly correct.

 

[PMR06]

Could you model this using layers of thinner plates stitched together with rigid links? Then apply a different temperature load to each layer?