Thermo-mechanical model ( explicit)

[nermine]

Hello people,

I'am working on a thermo-mechanical(explicit) model using the step Dynamic, temp-disp, explicit,I have two discrete rigid parts with a deformable part ( dies and a workpiece) because i'am modelling a hot forging model.
- Interactions : surface to surface contact / surface film condition
- system units: mm/s/mJ/T/MPA
- Mesh : C3D8RT for the workpiece,hex and structured elements
- BC : Velocity for the top die and the bottom die is encastred
- Predefined field : each part has its initial temperature of forging
- interaction property : Normal / tangential / thermal conductance
- properties: density, elastic plastic,conductivity and heat capacity ( temperature-dependent data)
My question is why do i always have such a problem ( see picture joined here) ? even when i refined the mesh, the problem seemed to occur again and again. Another problem is that with the mesh refinemend my problem takes tooo much time for the calculcation, really too many increments, for 1 sec time period , my laptop spent two days to be at 0.25 s !! What can be the cause please ? is it due to the complexity of the model ? or is it related to my laptop performances ? i have an intel core i7 processor( 2 cores) but a good graphic card ( NVIDIA quadro k2100M) and a 16 GB RAM ! if there is any adjustment to do in order to accelerate the calculation, i will be glad !!
PS : for more enlightment i can send the cae model by e-mail to make thigs more clear.

 

[FEA way]

Some small penetration may occur because of visualization issue so start from measuring this penetration. Make sure that master and slave surfaces are properly defined in contact pair settings. Add point mass/inertia to reference point of each rigid die. Try with general contact if it still doesn't work. And if you want to speed up the analysis, you can use mass scaling.

 

[IceBreakerSours]

What fraction of (slave) element thickness is the penetration you are 'seeing'? If it is an acceptable fraction, then why bother? Also, doesn't the compressing shells must have some thickness; nothing has zero thickness so may be you are being thrown-off by the post-processor.

A couple of thoughts: 1. The rigid elements are too 'big' in comparison with the deformable elements; you want to have them both be of similar dimensions. 2. If the final expected deformation isn't large, then try implicit. Otherwise, as long as the BC is applied 'slowly', then mass scaling is your best friend in explicit.

*********************************************************
Are you new to this forum? If so, please read these FAQs:

http://www.eng-tips.com/faqs.cfm?fid=376

http://www.eng-tips.com/faqs.cfm?fid=1083