Hi everyone!
I'm still trying to compare the electrostatic force calculated by the expression (F=0.5*epsilon*V^2*A/g^2) with the result coming by Maxwell stress tensor in COMSOL.
I used the suggestion of Ivar KJELBERG (in the Model Exchange part): mapped mesh, axial symmetry. But still the results don't match (3 orders of magnitude).
Once I set the variable Fes in the force table, I get the result from Post processing-->data display-->global-->Fes_z
If I try to integrate the stress tensor through the same bonderies (surrounding the Fes box) I get a N/m value. What is strange is that the relation between these two forces is the perimeter of the plates (in my cases 2*radius*pi.greco since I'm using circular plates with the axial symmetry).
Is there anyone that already tried to computer the electrostatic force by COMSOL with good results?
Has anyone an idea about the behavior of the boundary integration of the Maxwell stress tensor?
Thank you very much to everyone will give some suggestions
Riccardo
I'm still trying to compare the electrostatic force calculated by the expression (F=0.5*epsilon*V^2*A/g^2) with the result coming by Maxwell stress tensor in COMSOL.
I used the suggestion of Ivar KJELBERG (in the Model Exchange part): mapped mesh, axial symmetry. But still the results don't match (3 orders of magnitude).
Once I set the variable Fes in the force table, I get the result from Post processing-->data display-->global-->Fes_z
If I try to integrate the stress tensor through the same bonderies (surrounding the Fes box) I get a N/m value. What is strange is that the relation between these two forces is the perimeter of the plates (in my cases 2*radius*pi.greco since I'm using circular plates with the axial symmetry).
Is there anyone that already tried to computer the electrostatic force by COMSOL with good results?
Has anyone an idea about the behavior of the boundary integration of the Maxwell stress tensor?
Thank you very much to everyone will give some suggestions
Riccardo