Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

Hysteresis loss Modelling of Inductor coil

Please login with a confirmed email address before reporting spam

Hello
I have simulated inductor using AC/DC module and I have calculated the inductance and magnetic field around the inductor but now I want to calculate the hysteresis loss in the inductor. For that I am using steinmetz equation to calculate the hysteresis loss in one element and then summation of loss over infinite elements, but for now I am unable to figure out how to add the steinmetz equation in the magnetic field simulations to compute hysteresis loss. Any suggestion and advices would be really appreciated

Regards
Khalid

9 Replies Last Post Jul 17, 2014, 9:19 a.m. EDT

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago May 17, 2013, 3:50 a.m. EDT
Hello Khalid,

hysteresis losses cannot be well met by the Steinmetz equations, but by an adequate model such as Jiles-Atherton or Hodgdon which uses differential equations. As I have workeed for many years on this field I could implement the latter model into a workable and precise way in order not to sub- or overestimate losses, which would rener a simulation worthless. I attach two papers , which may help. If you need parameters of a particular material class, please let me know.

Regards,

Mauricio
Hello Khalid, hysteresis losses cannot be well met by the Steinmetz equations, but by an adequate model such as Jiles-Atherton or Hodgdon which uses differential equations. As I have workeed for many years on this field I could implement the latter model into a workable and precise way in order not to sub- or overestimate losses, which would rener a simulation worthless. I attach two papers , which may help. If you need parameters of a particular material class, please let me know. Regards, Mauricio


Please login with a confirmed email address before reporting spam

Posted: 1 decade ago May 23, 2013, 8:37 a.m. EDT
Hello Mauricio

Thanks for your reply. I am working on Jiles Atherton Model to compute the hysteresis loss but I am afraid it will not be easy to find the parameters needed to model the hysteresis loss. The parameter values are not even provided by manufacturers and even they dont know.I am using the MF106 ferrite material at 100 kHz.
I have attached the pdf of the material
Hello Mauricio Thanks for your reply. I am working on Jiles Atherton Model to compute the hysteresis loss but I am afraid it will not be easy to find the parameters needed to model the hysteresis loss. The parameter values are not even provided by manufacturers and even they dont know.I am using the MF106 ferrite material at 100 kHz. I have attached the pdf of the material


Please login with a confirmed email address before reporting spam

Posted: 1 decade ago May 23, 2013, 3:28 p.m. EDT
Hello Khalid,

for both the Jules-Atherton and the Hodgdon models the knowledge of the majos hysteresis curve is indispensable to determine the parameters., This curves are missing in the data sheet and so the only possibility I see is to use an equivalent material from a competitor (i.e same core loss level @ 100 kHz, 200 mT). Once you know the temperature you are probably to be working use the curve closest to it (most surely 100°C). Determining the parameters for the Hodgdon model via my procedure is straight forward, but I don't know if you can implement the equations in COMSOL. Determining the parameters for JA is not so simple and can yield wrong results if not done correctly.

Regards

Mauricio
Hello Khalid, for both the Jules-Atherton and the Hodgdon models the knowledge of the majos hysteresis curve is indispensable to determine the parameters., This curves are missing in the data sheet and so the only possibility I see is to use an equivalent material from a competitor (i.e same core loss level @ 100 kHz, 200 mT). Once you know the temperature you are probably to be working use the curve closest to it (most surely 100°C). Determining the parameters for the Hodgdon model via my procedure is straight forward, but I don't know if you can implement the equations in COMSOL. Determining the parameters for JA is not so simple and can yield wrong results if not done correctly. Regards Mauricio

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jun 3, 2013, 4:04 a.m. EDT
Hi,

i also have same intention. But until now i dont have good result and more questions to answer, because i dont know where to ask (i've tried support also).

it works with Jiles-Atherton.

pdf-file explains how inductor_w_hysteresis_nosolution.mph works. problem: pdf-file was made on comsol 3.5 the mph also. if mph opened on comsol 4.3 some special operators wont change automaticly. comsol doc tells minimalist about this operator.

so i addopted the old mph into 4.3 (with help from another user) and become hysteresis_example_2D-2-nosolution.mph.
Question: BH or HB-Curve from real material are available, but i stil dont understand how to make it as an input. Or JA-parameter should adjusted so that the hysteresis curve same as the real material.

i hope we can discuss more about it.

regards
akmal

Hi, i also have same intention. But until now i dont have good result and more questions to answer, because i dont know where to ask (i've tried support also). it works with Jiles-Atherton. pdf-file explains how inductor_w_hysteresis_nosolution.mph works. problem: pdf-file was made on comsol 3.5 the mph also. if mph opened on comsol 4.3 some special operators wont change automaticly. comsol doc tells minimalist about this operator. so i addopted the old mph into 4.3 (with help from another user) and become hysteresis_example_2D-2-nosolution.mph. Question: BH or HB-Curve from real material are available, but i stil dont understand how to make it as an input. Or JA-parameter should adjusted so that the hysteresis curve same as the real material. i hope we can discuss more about it. regards akmal


Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jun 10, 2013, 7:33 a.m. EDT
Thanks Akmal for your suggestions. I am working on your given model to get an inside to the JA method. One can vary different JA parameters to get desired shape, how accurate thats a question
For eg varying k results in dilation of the loop and changes the coercive field (Hc)
In addition I am simulating soft ferrite, I dont know how should I input the material properties in the comsol material browser. I have datasheet of the material but I am not been able to find other related parameters in the datasheet as asked by comsol. How should I simulate particular material, I have attached the datasheet of MF106 Ferrite

Regards
Khalid Hameed
Thanks Akmal for your suggestions. I am working on your given model to get an inside to the JA method. One can vary different JA parameters to get desired shape, how accurate thats a question For eg varying k results in dilation of the loop and changes the coercive field (Hc) In addition I am simulating soft ferrite, I dont know how should I input the material properties in the comsol material browser. I have datasheet of the material but I am not been able to find other related parameters in the datasheet as asked by comsol. How should I simulate particular material, I have attached the datasheet of MF106 Ferrite Regards Khalid Hameed


Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jun 10, 2013, 9:05 a.m. EDT

Thanks Akmal for your suggestions. I am working on your given model to get an inside to the JA method. One can vary different JA parameters to get desired shape, how accurate thats a question
For eg varying k results in dilation of the loop and changes the coercive field (Hc)
In addition I am simulating soft ferrite, I dont know how should I input the material properties in the comsol material browser. I have datasheet of the material but I am not been able to find other related parameters in the datasheet as asked by comsol. How should I simulate particular material, I have attached the datasheet of MF106 Ferrite

Regards
Khalid Hameed


Hi Khalid Hameed,
on your data sheet, you have something that can be very helpfull. Pdf page 5. real- and imaginary of permeabilität. i've never tried this befor. did you tried?

define your soft iron like this:
Ampère's Law > Magnetic field: constitutive relation: magnetic losses: use permeability (real and imainary)

study time dependent
tell me if it works or not.

regards
akmal


[QUOTE] Thanks Akmal for your suggestions. I am working on your given model to get an inside to the JA method. One can vary different JA parameters to get desired shape, how accurate thats a question For eg varying k results in dilation of the loop and changes the coercive field (Hc) In addition I am simulating soft ferrite, I dont know how should I input the material properties in the comsol material browser. I have datasheet of the material but I am not been able to find other related parameters in the datasheet as asked by comsol. How should I simulate particular material, I have attached the datasheet of MF106 Ferrite Regards Khalid Hameed [/QUOTE] Hi Khalid Hameed, on your data sheet, you have something that can be very helpfull. Pdf page 5. real- and imaginary of permeabilität. i've never tried this befor. did you tried? define your soft iron like this: Ampère's Law > Magnetic field: constitutive relation: magnetic losses: use permeability (real and imainary) study time dependent tell me if it works or not. regards akmal

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jun 10, 2013, 12:21 p.m. EDT
That worked for ferrites, but if one opt this method then every material has to be defined via real and imaginary part of permeability even air, aluminium and copper. Is there any other method that each material could be defined by different constitutive relation i.e either magnetic losses, HB curve or relative permeability

Coming back to JA method, as you have define current via I*sin(2*pi*f*t) but you have inserted t=0 (but it still simulates, I dont know)and secondly if I change the frequency to lets say 100 [kHz], I get a very straight and short diagonal line instead of close hysteresis loop. Does this suggest that the material im using is lossless or the JA method that employs flsmhs function doesnt fit in the current simulation

Have you got success in matching the datasheet hysteresis curve with any of your simulated materials?

Looking forward for your answer


Khalid
That worked for ferrites, but if one opt this method then every material has to be defined via real and imaginary part of permeability even air, aluminium and copper. Is there any other method that each material could be defined by different constitutive relation i.e either magnetic losses, HB curve or relative permeability Coming back to JA method, as you have define current via I*sin(2*pi*f*t) but you have inserted t=0 (but it still simulates, I dont know)and secondly if I change the frequency to lets say 100 [kHz], I get a very straight and short diagonal line instead of close hysteresis loop. Does this suggest that the material im using is lossless or the JA method that employs flsmhs function doesnt fit in the current simulation Have you got success in matching the datasheet hysteresis curve with any of your simulated materials? Looking forward for your answer Khalid

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jun 11, 2013, 9:39 a.m. EDT
Hi,
ampere law: you can select soft iron (domain) only for real and imaginary part of permeability. And for air or else you make another anpere law node and define them with relative permeability.

t=0 just define initial condition. if time dependent study starts t=0 will be ignored. "a very straight and short diagonal line" have you zoom in? if p_loss unequal to zero, then the closed loop has an area (see another line graph on result). i cant answer this (Does this suggest that the material im using is lossless or the JA method that employs flsmhs function doesnt fit in the current simulation)

not yet.

akmal
Hi, ampere law: you can select soft iron (domain) only for real and imaginary part of permeability. And for air or else you make another anpere law node and define them with relative permeability. t=0 just define initial condition. if time dependent study starts t=0 will be ignored. "a very straight and short diagonal line" have you zoom in? if p_loss unequal to zero, then the closed loop has an area (see another line graph on result). i cant answer this (Does this suggest that the material im using is lossless or the JA method that employs flsmhs function doesnt fit in the current simulation) not yet. akmal

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jul 17, 2014, 9:19 a.m. EDT
Hello Guys,

I am designing inductor for Buck Converter and for this i have to find losses in the inductor to find the efficiency of the Buck converter. Can anyone please guide me what information do we require to find hysteresis loss and eddy current loss in the inductor. In my case, inductance is 3mH and frequency is 10kHz. Other parameters of buck converter are Input voltage 300V, output voltage 100V, output power is 1kW and output capacitance is 100uF.

Please dear friends help me in this regards.

Thanks in advance
Best Regards
Muhammad Saleem
Hello Guys, I am designing inductor for Buck Converter and for this i have to find losses in the inductor to find the efficiency of the Buck converter. Can anyone please guide me what information do we require to find hysteresis loss and eddy current loss in the inductor. In my case, inductance is 3mH and frequency is 10kHz. Other parameters of buck converter are Input voltage 300V, output voltage 100V, output power is 1kW and output capacitance is 100uF. Please dear friends help me in this regards. Thanks in advance Best Regards Muhammad Saleem

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.