Iron loss calculation method Loss surface formular

yugang

Hello

Recently i use Flux to calculate the iron loss in the ring shaped core. I use loss surface method. When i finish transient simulation in Flux3D, i calculate the loss in the point.

So i get the B, H, power loss at this point.

Before, i get training from Altair, the loss is Loss=integral(B*dh)/T.

I use the B, H value at this point to calculate the power loss, while i can not get the same result as Flux result.

I attach the excel of the data in this post. It is kind if any information or support can be get from this forum.

Can any one repeat the calculation process to get same result of software. Because, we want to know how it works.

Thanks.

 

 

 

 

 

Lavanya Vadamodala_20519

Hello Yugang,

I am looking into the computations. Will get back to you.

Thank you,

Lavanya

Lavanya Vadamodala_20519

Lavanya Vadamodala_20519 said:

Hello Yugang,

I am looking into the computations. Will get back to you.

Thank you,

Lavanya

Hello Yugang,

There is a difference between Flux density computed on a point and Flux density computed by loss method in Flux (which is not supposed to be there). This is the reason for difference in power computed.

I created a ticket to our development team for answer.

I will let you know, as soon as get an answer from them.

Meanwhile, can you share you email and company name? so that I can create a ticket and keep it open until this is solved.

Thank you,

Lavanya

yugang

Lavanya Vadamodala_20519 said:

Hello Yugang,

There is a difference between Flux density computed on a point and Flux density computed by loss method in Flux (which is not supposed to be there). This is the reason for difference in power computed.

I created a ticket to our development team for answer.

I will let you know, as soon as get an answer from them.

Meanwhile, can you share you email and company name? so that I can create a ticket and keep it open until this is solved.

Thank you,

Lavanya

Hello

Thanks for your answer and help on this topic.

My mail address is gang.yu@se.com

My company is Schneider Electric(Shanghai R&D center).

 

 

yugang

Lavanya Vadamodala_20519 said:

Hello Yugang,

There is a difference between Flux density computed on a point and Flux density computed by loss method in Flux (which is not supposed to be there). This is the reason for difference in power computed.

I created a ticket to our development team for answer.

I will let you know, as soon as get an answer from them.

Meanwhile, can you share you email and company name? so that I can create a ticket and keep it open until this is solved.

Thank you,

Lavanya

Hello,

Do you have information on this topic ?

Thanks.

Lavanya Vadamodala_20519

yugang said:

Hello,

Do you have information on this topic ?

Thanks.

Hello Yugang,

Below is the explanation on difference in Flux density seen on point on a face and point on edge, which is leading to difference in losses computed.

1. Point in the middle of a face:

The slight difference between the results is explained by the fact that the points are defined on a user-defined surface. Thus, when the domain is meshed, that surface is subdivided in 2D elements (faces) that are shared between two 3D elements: one element to each side of the geometrical surface as seen in the attached figure. The computation done through the sensor interpolates the B values at the selected point using the 3D element on one side, whereas the computation through the LS model uses the element on the opposite side. Therefore, a slight difference appears between the two methods. This difference is negligible and does not affect the accuracy of the results.

2. Point on the edge:

Results obtained with the sensor and LS can be very different if the chosen point is at the interface of two physical regions: Air (non-magnetic) and magnetic. Moreover, if the point coincides with a node and such node is shared by multiple 3D elements, the difference in flux density will be there.

In order to obtain approximately the same results, the sensor needs to be computed on the physical region “Air” (as seen in the image). Indeed, when Flux computes B for LS, it defines “Air” as the parent region of the point. Results will not be exactly the same because a 3D element has to be selected for the computation, which means that we find here as well the problem described above for the center point.

This problem can be avoided by choosing a point that does not belong to the interface between to different physical regions. 

Does the point on which you have extracted the values, come under any of the above cases?

Thank you,

Lavanya