housing analysis

[glad221]

Please find the attached model on which I am working. As you can observe, to make this two part (housing and the lid) assembly, I have used the rigid bar elements and spider connection to represent a screw connection between the housing and the lid. This surface contact between the housing and the lid locks all degrees of freedom.

However when I run the simulation in order to find the natural frequency of this assembly, I only get the results from one part (either housing, or the lid) and not the assembly. The software however does not recognizes this as an assembly.

if there is any better way of a surface contact between the two parts that allow the software (LMS Virtual-Lab) to recognise this as one assembly.


Thank you

 

[GregLocock]

I doubt many FEA users can discern much about mesh or modelling problems from a screenshot of a solid model.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[glad221]

A few pictures of models, please look and give suggestions if possible.
thanks

 

[glad221]

A few photos of models, please look and give suggestions if possible.
thanks

 

[BlasMolero]

Dear glad221,
The term "assembly" in FEA is not applicable the same way as in CAD systems, we manage assemblies all the time when we mix different element types in a FE model. But returning to your problem: I am not expert in LMS Virtual-Lab, in theory if you define rigid bar elements between touching faces of contacting solids then the model will behave like one part, and a freq analysis will be possible. Try to export the model in an ASCII TXT format to see that in effect the rigid bar elements are created properly. As a work-around you can imprint touching faces in each other solids and merge coincident nodes, this is in effect a similar approach, sorry not been able to help more.
Best regards,
Blas.

 

[crisb]

You mention about contact which inherently creates a nonlinear analysis, whereas a natural frequncy analysis is essentially a linear analysis.
I think it's only possible to carry out a natural frequency analysis at one particular stage of a nonlinear analysis, using a restart analysis or similar. The nonlinear contact is carried out as a static analysis, followed by the natural frequency analysis.
This is a general comment and I have no knowledge of the capabilities of your software for this type of combined analysis.

 

[flash3780]

It sounds like what you're trying to do is a pre-stressed modal analysis.

I'm not familiar with the software that you're using, but you'll have to solve the problem in two steps.

Firstly, you'll need to tie the beams to the tets at a node add a tension to your screws. Different software approaches this differently, but most have good help files. You should tie the flange together with constraint equations, bonded contact, or mesh right across the flange: Modal analyses cannot handle non-linear contact. From there, you'll want to go ahead and run your pre-stressed modal analysis.

I'd try to avoid both the natural frequencies of the individual components and of the assembly.

To simplify things, you could assume that the stress stiffening from the screw load is small and mesh across the flange & ignore the screws completely. That'll probably give the same answer with far less headache.