Ferrite devices simulation

    • laila.marzall

      Hello Jacob,

          My name is  Laila Marzall and I am a Ph.D. student in Microwave Engineering at CU Boulder. My advisor is Professor Zoya Popovic.     Last year I've started to work on a project concerning to externally and self-biased, unsaturated ferrite circulators. Since then I've been using HFSS coupled with Maxwell3D for obtaining non-uniform Magnetic DC Bias Fields.     I'm are facing some difficulties to match measurements and simulation with my designs.     HFSS apparently only works with Polders approach, that presumes a saturated ferrite condition. This condition is not valid in the cases where ferrites are weakly biased (unsaturated condition). I'd like to confirm this constraint and would appreciate some instructions to overcome this limitation.   I also have many specific doubts about how to make the right simulation approach, for instance:

      • How to choose the right order for base functions;

      • How and when to use the permeability tensor instead of magnetic saturation;

      • What are the limits in terms of bandwidth since we are working with non-linear materials (which permeability depends on the operating frequency);

      • How to take into account demagnetization factors;

      • How to configure the material to include damping effects (losses);

      • How to perform a check of FMR ferromagnetic resonance under unsaturated conditions and different geometries.

        I'm having some difficulties in finding content about ferrites in your support platform, so if I could contact some of your team members to help me, it would be great.   To clarify my problem, please find attached a paper that compares the same design simulated in HFSS, Matlab, and CST.   Thank you in advance.

    • slouie
      Ansys Employee

      Hi Laila,

      You could use the following approach:

      • Solve in Maxwell and plot H field in ferrite to determine areas in the ferrite that aren't saturated (and therefore Polder approach is invalid).

      • In HFSS, divide the ferrite into multiple 3D objects - one or more that represent saturated regions and one or more that represent unsaturated regions.

      • For saturated objects, you may apply Polder tensor in material properties and Maxwell solution for bias

      • For unsaturated objects, define appropriate complex tensor model in material properties

      All material properties in HFSS can be defined as frequency-dependent models. Accuracy across the solved frequency band will depend on defining frequency-dependent material properties. 

      Start with the above procedure using default HFSS settings for things like order of basis functions and let us know how it goes!

    • laila.marzall

      Hello, thank you so much for the tips!

      Do you have an HFSS example using complex tensor?

      How I have a frequency dependent permeability (and is a very sensitive parameter for unsaturated regions) I kindly as you an example where the tensor is configured as frequency dependent.



    • laila.marzall

      Hello, do you have any position about the example?

      I'm working on a huge project of nanomagnetic materials and really need to try new simulation approaches. We are not reaching good results in HFSS.


    • laila.marzall


      I'm having a new issue with simulations of HFSS coupled with Maxwell3D.

      When I coupled the nonuniform bias field generated by Maxwell3D the sweep simulation stops with an error and I am not being able to detect what is the problem.

      Could you take a look at the file and tell me what I am doing wrong?

      Thank you!

    • laila.marzall

      Hello, do you have any position?

Viewing 5 reply threads
  • You must be logged in to reply to this topic.