Von Misses and Local Buckling

[Arun4567777]

Dear All,

I am trying learning FEM. I have a plate with small b/t ratio thus qualifying as a slender section. It is made by cold rolling process. Since this section will be subjected to local buckling, I want to know that if I model this plate in Staad and check Von Misses Yield Critera. Will Von Misses take that local buckling effect into account or I need to check that separately.

 

[ldeem]

I am not sure what you are designing but plate buckling is a complex subject. I typically use Blodget to get an idea of buckling stress and then stay well away from that.

 

[WARose]

...I want to know that if I model this plate in Staad and check Von Misses Yield Critera. Will Von Misses take that local buckling effect into account or I need to check that separately.

Von Misses criteria is independent of any buckling.

 

[Arun4567777]

So do I need to check the plate seperatley for buckling.

 

[WARose]

So do I need to check the plate seperatley for buckling.

Yes.

 

[SWComposites]

Agree, YES. Buckling is a completely separate check from basic strength checks.

 

[Lomarandil]

I'll add that while I don't know the capabilities (or lack thereof) for STADD specifically, not all FEM is capable of modeling thin shell buckling appropriately. You might need something specifically written to handle that, like CUFSM.

General-purpose software that can be adapted to handle local buckling would still require very careful use and validation.

 

[271828]

A small b/t ratio would indicate local buckling is NOT a problem, so I assume you mean "high b/t ratio."

von Mises is a yield criterion. It takes the general stress state with six components (sigma_x, sigma_y, tau_xy, ...) and computes whether or not that stress state should result in yielding. It has zero to do with local buckling.

To check local buckling with an FEA program, you would need to take one of two approaches.

Option 1 is to do an eigenvalue buckling analysis. These are typically linear, so would have nothing to do with yielding.

Option 2 is to give the element a small out-of-plane displacement (imperfection) and then run a large-displacement second-order analysis and track the stress-displacement curve to find buckling. This could be material linear, which would have nothing to do with yielding, or material nonlinear, which could consider yielding.

I haven't used STAAD since 1998, and disliked it a lot back then (LOL), so I don't know if it would do Option 1 or 2. SAP2000 will do them, so you don't need a very high end and expensive program like Abaqus or Ansys.

 

[WARose]

To check local buckling with an FEA program, you would need to take one of two approaches.

Option 1 is to do an eigenvalue buckling analysis. These are typically linear, so would have nothing to do with yielding.

Option 2 is to give the element a small out-of-plane displacement (imperfection) and then run a large-displacement second-order analysis and track the stress-displacement curve to find buckling. This could be material linear, which would have nothing to do with yielding, or material nonlinear, which could consider yielding.

I haven't used STAAD since 1998, and disliked it a lot back then (LOL), so I don't know if it would do Option 1 or 2. SAP2000 will do them, so you don't need a very high end and expensive program like Abaqus or Ansys.

I use STAAD every day (in fact, just about every day for the last 25+ years).....and I wouldn't use STAAD (or most any other FEA software I am familiar with) for a buckling analysis. I'd use hand calcs. For most things we do, it is faster anyway.

 

[ldeem]

Make sure you read up on how STAAD is showing results. You can see a Von Mis stress in results but that is a calculated stress RESULT. I tried to find a link to the STAAD technical reference but couldn't. The value in the results has nothing to do with any calculated allowable stress.

In my case when I have had to look at plate buckling it has always been large items (well over 40x50x50'). I use STAAD to determine plate stress due to loads then I figure out an allowable stress using Blodgett (mostly) sometimes Roark.

 

[271828]

I use STAAD every day (in fact, just about every day for the last 25+ years).....and I wouldn't use STAAD (or most any other FEA software I am familiar with) for a buckling analysis. I'd use hand calcs. For most things we do, it is faster anyway.

Buckling analysis in an FEA program is my last resort also. It's important to do enough manual calcs so that I know approximately what the programs should predict.