I was trying the frequency analysis feature in simscale recently and I had the opportunity to compare the results with data derived from another FEM-tool (comsol). I was expecting the same results as the problem is not to difficult and the naterural frequencies should be constant. However, simscale doesn’t find the lowest resonance frequencies. Moreover, the lowest frequency in simscale is already a mixed mode vibration.
The only difference I was able to find was the orientation of the global coordinate system. But beside that both simulation setups (simscale, comsol) were the same.
Did somebody encounter similar differences? What might be the reason?
The study can be found in my public projects. The lowest frequency by comsol is 112Hz while simscale says 180Hz.
Hi there, would it be possible that you include the link to the project, as well as some images/data that you have gathered so far?
Here is the link to the project in simscale:
Here is the lowest frequency according to comsol: 112Hz along the x-direction without distortion of the upper lengthwise beam (completely red)
This is the lowest resonance according to simscale: 180Hz along x-direction but with a distortion of the upper beam
I repeated the whole experiment, this time with matching orientation of the coordinate system. But the result is still the same.
Hi, could you give it a go with a second order mesh, by enabling this option?
When looking for more accurate solutions, second-order meshes are the way to go.
Switching to 2nd order elements reproduced the results from Comsol. Thanks for the tip! However, it is a bit scary that the results between 2nd order and no 2nd order are so different. Non-professionals might get lead astray.