Modeling compression only supports

[radiocontrolhead]

I am trying to achieve a behavior in FEA that considers a roller support but constrains movement in the negative gravity direction only allowing it to lift off the support if the deflected shape warrants it.

is there a way to do this? I've been playing around with tension/compression links that are extremely flexible but it's a bit wonky and not very accurate.

Thank you

 

[rb1957]

what code ?

gap/contact elements ?

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[BAretired]

One way to do it is to run with a roller support and observe the reaction. If it is upward, leave it in. If it is downward, remove the roller support and run it again.

 

[radiocontrolhead]

This is primarily for building structures. I am not sure i understand what gap/contact elements are unless it's equivalent to say a link?

 

[radiocontrolhead]

BAretired,

I think this approach will work just dandy. I will chime back in if I find I need anything more from the analysis. Thank you!

 

[SWComposites]

what FEA code are you using?

 

[NRam]

If using ETABS or SAP2000, you can define a link as a "GAP" Element.
If you define the Gap as 0", then it will function as a compression only link.
Or you could define a non-linear link, with a really high stiffness in compression, but a pretty low stiffness in tension (don't make the numbers too large as it can affect the accuracy of your model)
I've done the removing supports thing too.

Just an idea, but also depending on what program, maybe you can use some kind of soil spring support definition?

As others said, you can "iterate" the model yourself by just removing the support if it's in tension.

 

[sticksandtriangles]

NRam,

Etabs and SAP have an area, compression only support button. My understanding is that you assign this to your footing of interest and it will determine the soil spring stiffness based on the FEM mesh that is created.

S&T -

http://www.re-tug.com

 

[JoshPlumSE]

I am trying to achieve a behavior in FEA that considers a roller support but constrains movement in the negative gravity direction only allowing it to lift off the support if the deflected shape warrants it.

In most programs, this should be relatively easy. They should have a support that is a compression only or tension only support. Alternatively, the program might have a member that is a tension only or compression only member. Or, a gap or hook element.

I've been playing around with tension/compression links that are extremely flexible but it's a bit wonky and not very accurate.

Okay, if you have a compression only element, make it relatively rigid and connect it to your pinned boundary condition.

 

[NRam]

Sticksandtriangles,

Yes that could also work! The soil spring feature in ETABS uses the user defined stiffness property in pci (lb / in3 = lb/in / in2) which is essentially an area stiffness (the amount of force for one square inch to deflect one inch).
This stiffness in provided by the Geotech as the subgrade modulus.
ETABS then dumps this into the corner nodes by consider the tributary area of each node. So pci * in2 = lb/in (regular stiffness spring for each node).

So you could model, say, a "baseplate" with a high stiffness in the u3 direction and that would work as a roller too.

 

[cam_b]

I'm not sure what software you're using. I know that in Space Gass you can make any support unidirectional, which would achieve what you are after. I expect other software would have similar features.

 

[sdz]

BAretired,? Bachelor of Arts?
Most FEA structural programs have some method of doing what you want using gap elements or compression only elements. What program are you using?

Ooo eee ooo ah ah ting tang walla walla bing bang
Ooo eee ooo ah ah ting tang walla walla bing bang