I am working on flow-induced vibration of a double wall carbon nanotubes. I have solved the problem analytically and approximately. The results of both solutions are almost identical. Then, I used COMSOL multi-physics to do the same problem. However, the COMSOL results do not match with my previous solution. I may do something wrong with COMSOL modeling ( I am not COMSOL professional). Any suggestion to modify that may help for this.
Bests, Ali
On
When working with FEM, you discretise your domain into finite elements. Perhaps you could try to progressively decrease the element size in your domain, solve the problem and check the results for each simulation. When the results do not significantly change anymore (say 5%), you could argue that your results do not depend on the mesh size, and so you have eliminated (at least) one potential source of error. This is the essential idea of mesh convergence analysis.
It is always hard to know if the simulated answer is right or wrong, so you cannot really know if the previous solutions are wrong. With my Comsol and other FEM software experience, I can say that the hardest part is to validate the answer. You should do some measurements with real nanotubes to be sure. But anyway, you should provide more information about your case to answer more precisely.