Learning Forum

[Anunda Chuenlerssakul]

Hi all,

I would like to asking for a help on the mesh metric and force-torque validation. I am working on the motor simulation in the 3D transient but I wonder how could I validate the accuracy of the result since it doesn't have Mesh metric ex.Skewness in Ansys Electronics. 

Can I prove it with integral of maxwell stress tensor as the transient is using virtual work principle?? and if it possible, how can I validate by this? Could anyone suggest on the validation for the force and torque result?

 

Thank you very much

Anunda

[GLUO]

Hi,

I don't think comparing integral of VFD/MST with the default method(virtual work) would prove anything. They are using the same mesh. Normally we increase the mesh size and compare the new torque/force result with the result using the original mesh, and if the results are close, then it means the mesh is acceptable.

GL

[Anunda Chuenlerssakul]

 

I try to decrease the mesh size and select the suitable one but my study is about varying the width of the stator. So when I try to use the same mesh size with different stator thickness, the torque result is so different. It's not linearly increase nor decrease it really sprase on the result and also the force result is sprase too. 

The mesh setup is here, I have attach the image below.

decreasing the stator width is giving a higher torque I don’t know why?

 

[GLUO]

Hi,

If you see toque varying when change the mesh size, that means the mesh might not be optimal. Please make sure there are multiple layers of mesh in the air-gap and stator. If you are using segmented geomery, please also make sure each time step, the rotor rotates integer number of segments (N*360/#segment degree). This improves the accuracy of torque ripple.

GL

[Anunda Chuenlerssakul]

I have seen that many tutorial, also official tutorial of ansys doesn't care that much about the meshing in case if magnetostatic. So, if it poosible for me to neglect all mesh at the air gap and just provide the mesh at surface of material that I would like to observe the force and torque?

[GLUO]

Hi, 

Magnetostatic uses adaptive mesh, but it is still based on mesh operations. I would still recommend to make sure there are multiple layers of mesh in the air-gap and stator by adding dummy band objects. Also, please still use a dummy "motion band" and assign torque to the motion band and everything inside the motion band.

GL