What do you like best about SimScale?
No data loss so far. No crashes. No fatal errors. Do not have to worry about simulation stopping if I send my computer to hibernate, or if I have to restart it.
Great graphical user interface. Be it CAD editor, mesh creation, or results post-processing.
And if I have to do something more complicated, I can just export/download the results to my PC and look at them in ParaView.
CFD calculations being done in OpenFOAM (an open-source, well-known CFD software) is appreciated.
Great responsive support team. It's almost like talking directly to developers of SimScale.
I can use as many resources as I want. The only problem is when I run out of resources - because that can be costly for whoever is paying for my subscription.
I generally import CAD files in STP file format, but SimScale has a very good compatibility with a wide range of file formats.
Screenshots up to around 20M pixels (probably depends on the RAM you have available on your PC?) can be taken, and they are very lightweight - less than 1MB each. Review collected by and hosted on G2.com.
What do you dislike about SimScale?
Occasional glitches with screenshot taking. Took a screenshot? Download it (and make sure it's not just an empty rectangle), because if SimScale forgets the "view" for the simulation run, it will forget all the screenshots.
Occasional glitches with CAD editing - if an operation has a problem, I cannot interrupt the operation, and I have to wait for operation to time out - which can take a while. But don't worry about it, CAD editor is being actively developed, developers are aware of the issue, so this problem will be solved. In two months™?
Periodic boundary condition can be done for translational periodicity, but not for rotational one. Very annoying! It could have reduced computational usage by a factor of 2, or 4, or 12, depending on geometry, if periodic boundary condition just worked. [I hope SimScale people are reading the review. Because I am sure that OpenFOAM has this feature, but SimScale for whatever reason does not have it, yet]
Predictions of running time and computing resource usage are unreliable. The software responsible for that probably needs to take more factors into account - even with similar geometries, meshing can take 40min for one case and 80min for another (and prediction is probably 15min? to be fair, my geometries are inordinately complex). Review collected by and hosted on G2.com.