Learning Forum

[CassioEM]

I am working in a transient analysis and my simulation is taking too much time... I started the resolution on Friday afternoon and when I arrived on Monday morning, I realized that the progress bar did not finish even half of the resolution...

I have thee steps:The first two are a quasi static analysis and the last one is a transient analysis.

The first two steps take around 1 hour and 30 minutes. The problem is the third one. I know that it is possible to change some parameters to optimize the resolution process. I already changed the number of points and the mesh. Is there any other parameter that we can change in order to reduce our time of resolution? I mean, parameters those are quite common to change in this kind of problem.

[peteroznewman]

There is a much faster way to calculate a Transient if you have a linear model, and that is the MSUP Transient.
You can learn more about that by reading the Transient Dynamics section of the ANSYS Help.

Another analysis that is even faster than a Transient is a Response Spectrum. That analysis, also only for linear models, can estimate the worst case state of the model over the duration of the transient. It is an estimate, not an exact calculation of the peak response, but it is very fast and useful on models that are too large to run as a Transient.

[CassioEM]

The model is non linear because I have large deflections. I changed the number of points in the solution and the time step for the parts those are not so important. I have not calculated the solution yet but I believe that is not enough.

I was thinking maybe I can change the resultants those are recorded by the software, because I have a considerable number of probes. But I am not sure if it will be enough...

 

[peteroznewman]

Do you need Strain to be computed at the end, or only Stress?  If you don't need Strain in the solution, you can turn it off in the output and that will save some time.

You can also optimize the mesh to be as small as possible.