• Registered: 12-01-2017
  • Posts: 12

Topic: Compression experiment simulation

Hello,

I want to use OOFEM to simulate  the compression process of the metal  part. I know how to set static load, however, the load changes over time during the compression process, I do not know how to set dynamic load. Do have any examples i can follow in "..\tests\sm"?

2 Reply by bp

  • bp
  • Administrator
  • Offline
  • From: Prague, Czech Rep.
  • Registered: 02-03-2007
  • Posts: 1,026

Re: Compression experiment simulation

Hi,
any load can change during the time, this is defined by loadtimefunction parameter. In short, the value of a load is evaluated as product of load value and value of loadtimefunction evaluated at specific time. Typically you use loadtimefunction different from ConstantFunction, see tests/sm/Mieses01.in for a simple example.

Borek

bp's Website

  • Registered: 12-01-2017
  • Posts: 12

Re: Compression experiment simulation

Hi Borek

Thanks for your reply.  It's a good example for me to learn.  According to the demo, I test a model. Here are some parameters:
LinearStatic nsteps 6 maxiter 20 nmodules 1 lstype 4 smtype 10
...
NodalLoad 2 loadtimefunction 1 dofs 3 1 2 3 components 3 0.0 -5000.0 0.0 set 2
PiecewiseLinFunction 1 t 2 1.0 6.0 f(t) 2 0.0 5.0
OOFEM can run the model successfully. However, I have several doubts about it.
1)  Is it right the analysis type i chose?  Is the compression process not a dynamic analysis type?
2）Is there a relationship between “nstep” and parameter “t”?
3) Can OOFEM output a time process, not a time point?

Thanks

4 Reply by bp

  • bp
  • Administrator
  • Offline
  • From: Prague, Czech Rep.
  • Registered: 02-03-2007
  • Posts: 1,026

Re: Compression experiment simulation

shili wrote:

OOFEM can run the model successfully. However, I have several doubts about it.
1)  Is it right the analysis type i chose?  Is the compression process not a dynamic analysis type?
2）Is there a relationship between “nstep” and parameter “t”?
3) Can OOFEM output a time process, not a time point?

1) It depend on the speed of the process, if it is relatively slow, you can neglect inertia and viscous effects and use static analysis. If you could not neglect inertia and viscous forces, you have to use dynamic analysis
2) In static analysis nstep parameter defines the number of loading steps you want to apply to specimen, t is then a discrete pseudo-time which equals to individual loading step numbers.
3) Not really sure that I understand your question. In dynamic analysis you will get the history of the process computed at discrete points. You can choose the time discretization, so you can control the resolution in time. You can also interpolate between solutions in individual discrete times to  get values in between, if needed.

bp's Website

  • Registered: 12-01-2017
  • Posts: 12

Re: Compression experiment simulation

Hi, Borek

Thanks for your reply. Your answer is clear. It helped me a lot.

I have other question need your help. Here is it:
How do OOFEM show plastic yield? I want to know the yield point of the model during the compression process.

6 Reply by bp

  • bp
  • Administrator
  • Offline
  • From: Prague, Czech Rep.
  • Registered: 02-03-2007
  • Posts: 1,026

Re: Compression experiment simulation

HI,
you should use appropriate nonlinear plasticity-based constitutive model. To choose the best one, please consult oofem materialmanual. On output, you can get plastic strains and status description (elastic, yielding, etc) in every integration point. Also the nonlinear solver reports for every solution step the reached load level. You can also export plastic strain field to vtk for visualization.

bp's Website