Ansys Maxwell Simulation- Torque Fluctuating

AI Thread Summary
The discussion revolves around troubleshooting torque fluctuations in an Ansys Maxwell simulation of a 1KW axial flux switching generator. The user reports that while the induced voltages are acceptable, the electromagnetic torque appears to fluctuate around a non-zero average, raising concerns about the simulation's accuracy. Suggestions include reviewing the model for geometry and excitation issues, testing in no-load conditions, and ensuring proper circuit connections, particularly regarding the delta configuration. The importance of grounding in the circuit setup is emphasized, as well as the recommendation to verify the model against original design data. The user is encouraged to follow up on these suggestions to achieve more reliable simulation results.
Hamed Farnaghizad
Messages
7
Reaction score
0
Hi guys!
I do have a problem in ansys maxwell simulation of a generator. The moving torque (which represents electromagnetic torque) fluctuates around zero which is not correct. Induced Voltages are ok. How can I fix this problem?
 
Engineering news on Phys.org
The first step to solving the problem is to describe it.

1) What are you simulating? Just saying "a generator" is not sufficient.
2) What is the purpose of your simulation? What do you want to get from it?
3) Post an image of your outputs. Describe why you think some are correct, and some are wrong?

Then ask for help.
 
Hi @jrmichler
Thanks for your reply, actually I thought it may be a common problem in ansys maxwell.
I am simulating a 1KW axial flux switching generator. I need to have its output voltages and electromagnetic torque.
Output voltages are acceptable but torque (moving toque in rectangular plot) is as the picture below:
Untitled.jpg
 
Hamed Farnaghizad said:

Well, this is not a "torque fluctuating around zero" in the first place. Its seems like being periodic and with non-zero avg value. Negative value means, well, that it's a generator.
First thing: is thit a no-load operation ? And second thing: how would expected result look like ?
(Sometimes model reduction + symmetry boundary may introduce some issues).
 
Dear @gerbi
I connected three phase windings in delta connection and connected to a three phase Y connection 100 ohms load. As I calculated torque should be around 47 Nm. One surprising thing is that , No load and Loaded Torques are the same!
 
100 Ohms is rated load ? I mean calculated from ohms law for generator rated voltage and current. I assume the load is also arranged in 3-phase manner.
 
Well, I changed output circuit as the picture attached below, but no good result.
Actually by changing the resistance of the load output voltage of the generator varies!
 

Attachments

  • Untitled.jpg
    Untitled.jpg
    8.5 KB · Views: 347
Phase voltage unbalance may suggest that your model has some major issues. I'm sorry, but you need to review your model (geometry, excitation, parameter, motion setup and so on) and double check it for differences when compared to original design data. Then test it in no-load conditions (increase resistance in your circuit to values like 10^6 ohm). Observe all winding flux/voltages (they must be symmetric) as well as flux distribution in generator components. That's quite standard approach for design review. If no-load operation is fine then you can move to loading your generator.
One thing I may say: it's rarely an Ansys Maxwell issue that causes a simulation to give so bad results.

Additionally, you may want to read this topic from beginning to the end: https://www.physicsforums.com/threads/ansys-maxwell-transient-torque-fluctuating-problem.926444/
 
@gerbi
Hi again
Sorry for delay in response
I checked everything and all the things look ok.
I think the problem is of the circuit, I used circuit editor of maxwell.
I wanted to have a Delta-Delta connection but It needs GROUND so I don't know what to do to have a DD connection.
 
  • #10
Well, you can always have Delta-Wye connection (Delta motor + Wye load with ground)...
Otherwise just ground one of load legs (ground is needed as a reference point).
I just hope you do FEM simulation to prove some previously designed and analytically analyzed concept. Designing a machine in Maxwell is not the most optimal approach.
 
  • #11
@gerbi
It is actually simulation of an analytically designed axial flux FSPMG, which is based on a journal paper. I will do your suggestions. I hope getting good results.
 
Back
Top