Altair FLUX - Sensor for Iron Losses

Imanol Eguren

Hello,

I am simulating an electrical motor that uses M800-65A sheets. I am using the parametric distribution option in the scenario so the computation time is reduced. 

I want to obtain the iron losses from predefined sheet models of FLUX. However, I do not know how to define a sensor for this result.

The lack of a sensor increases the processing time considerably, because the computation of the iron losses would also be distributed if I had a sensor defined, whereas now, I have to wait for quite a long time after the simulation has finished in order to obtain those results.

Any idea whether this is possible in Flux? I am currently using version 2020.1.

Best regards,

Imanol

 

Farid zidat_20516

Hello Imanol;

 

There is no sensor to compute iron losses. Even if it exists, the iron losses are not computed during the solving, but only at the end of the solving.

 

There is a big speed up of the iron losses computations, to see that, please use laminated non conducting regions, this will help to speed up the iron losses computation.

 

Please note, that if you are doing this to compute the efficiency map, and you are sudying a permanent magnet motor, you can use FluxMotor or you can use FEMT available from the flux supervisor.

 

Best regards 

Farid zidat_20516

Hello Imanol;

 

There is no sensor to compute iron losses. Even if it exists, the iron losses are not computed during the solving, but only at the end of the solving.

 

There is a big speed up of the iron losses computations, to see that, please use laminated non conducting regions, this will help to speed up the iron losses computation.

 

Please note, that if you are doing this to compute the efficiency map, and you are sudying a permanent magnet motor, you can use FluxMotor or you can use FEMT available from the flux supervisor.

 

Best regards 

Imanol Eguren

Hello Farid,

I did not know that the loss computation was faster with the laminated regions. I will modify the regions and try again.

Thank you!

Farid zidat_20516

Hello Imanol;

 

You are welcome!

 

This improve is recent actually, it was done by steps to accelerate the computation of iron losses. So it is normal you did not the info. 

 

I sugget you to have a look on the release not each time you install the newest version, so you can see what are the new features, and also issue corrected.

 

Best regards 

Gaizka UGALDE

But there is no need to wait for the postprocessing to compute the modified Bertotti model. It is true that the equation needs to calculate the derivative of B but it can be solver either with the initial magnetostatic simulation or even neglecting the first calculation steps in time or position. Actually, when losses are postprecessed, the first 2 point are not calculated. ¿Can somebody from Altair’s staff comment on that?

The post processing time increases a lot as the problem size increases because as Imanol said we are not taking advantage of the distribution of the computation load.