-
1
- #1
hadiandfem
Mechanical
- Oct 24, 2012
- 5
First of all, Abaqus might have a very good user interface,but the solver in coupled simulation is not generally as good as ANSYS.
Anyway, to those have a concern about "Creep Analysis"
Steady state (Secondary creep), the strain rate only depends on stress. Hence, You need A,n as temperature-dependent material properties (i.e Metal Handbooks, Published paper, etc). The dimension of A is MPa/s, where the Stress is MPa. And n is dimensionless. In the Abaqus, choose the Visco analysis in the step module. It has been recommended to use the strain-hardening (stress state can vary during an analysis)version of power law (Norton) in Abaqus. Keep in mind that in the time-hardening (the stress state is essentially constant) for non-zero values of m (total time order) the analysis has some inaccuracies. Furthermore, once m approaches to -1 the inaccuracies also increase. You need to decide for the value of time period, and increment size (initial, min and max). Let ask yourself what material did you choose? and how long it does take to get creep data in real tension creep test? (You can extracted this information in so many published paper)the answer is time period. For a very low stress you need to choose implicit/explicit integration method.
Finally, check your boundary conditions, per-defined thermal load and mechanical load as well. Also, check the size of your mesh and make a sure that has enough quality for your problem. Run the Job! You MUST plot the strain rate Vs. time.
Anyway, to those have a concern about "Creep Analysis"
Steady state (Secondary creep), the strain rate only depends on stress. Hence, You need A,n as temperature-dependent material properties (i.e Metal Handbooks, Published paper, etc). The dimension of A is MPa/s, where the Stress is MPa. And n is dimensionless. In the Abaqus, choose the Visco analysis in the step module. It has been recommended to use the strain-hardening (stress state can vary during an analysis)version of power law (Norton) in Abaqus. Keep in mind that in the time-hardening (the stress state is essentially constant) for non-zero values of m (total time order) the analysis has some inaccuracies. Furthermore, once m approaches to -1 the inaccuracies also increase. You need to decide for the value of time period, and increment size (initial, min and max). Let ask yourself what material did you choose? and how long it does take to get creep data in real tension creep test? (You can extracted this information in so many published paper)the answer is time period. For a very low stress you need to choose implicit/explicit integration method.
Finally, check your boundary conditions, per-defined thermal load and mechanical load as well. Also, check the size of your mesh and make a sure that has enough quality for your problem. Run the Job! You MUST plot the strain rate Vs. time.