[SOLVED] Square cross-section helical coil in ANSYS HFSS ( or Maxwell)

Hello everyone.

I wanted some help regarding the design of a helix with a square cross section.

I intend to simulate an electromagnet dipole with a rectangular cross section ferrite core in HFSS(or Maxwell)

Is there any library model like the user defined primitive segmented helix? If not any suggestions as to how may I approach it would be really helpful.

Thanks

- - - Updated - - -

I was wondering whether maybe using circular helix would do with the core being rectangular and not designing the coils being wounded exactly to the sides cuboid core structure. Would it effect the simulations very much ? Apologies for my ignorance as its my first time of designing such structure. Any help will be greatly appreciated

- - - Updated - - -



I have attached a picture of the type of winding I am talking about

I think I might have something that can help you, I'll attach an HFSS file.

There's an option in SegmentsPerTurn that you can change to 4. I think that will make the coil you need. Other things like Pitch, Wire Thickness, Radius, etc. are also all easily changeable. I'm not the creator of this file but it has worked well for me in a lot of my work.

@kthackst : Thank you v much. This is what I was looking for. But I am facing another problem. I designed my dipole with this design but the mesh calculations(by adaptive meshing) are taking a huge time. My 2 helix specifications are : Pitch : 10 mm, turns : 60, radius of helix : 25 mm. Is this abnormal meshing time required just because of these segmented helix ? I had to give manual meshing with very few elements(maximum no of elements to 10) to do the simulation.

Another thing , when I couldn't get a solution before this, I used the method of designing the coil by sweeping a rectangle around a vector as mentioned in this tutorial. https://www.youtube.com/watch?v=2O8y58XiDlU. These are taking normal calculation time. Do you know if this can provide me with the similar results as that of using actual helix coils ?

Well for the method in the video you would have to sweep one rectangle for every turn? That seems like a lot of excitations. I would bet the current distribution in the true helix is just a bit more complicated than that of the repeated loops, making HFSS take more time to get an accurate solution. My personal preference is to make my simulation as accurate as possible and let it run overnight if I have to, unless you plan to do a parametric sweep in which case obviously that's not feasible.

Another solution would be going into Analysis --> Setup1 (or whatever its named) --> General --> Maximum Delta S and increasing the number here. I would compare with a longer sim to make sure you are still getting the data you are most interested in.

Alright. . No its actually sweeping a single rectangle for all the turns and then specifying the no of turns and total ampere turn as the current excitation. Yes its true that using a helix will give more accurate results. But its not becoming feasible for my design. Actually I am using Maxwell for my simulation right now and referred to their guide manual where they have used the same way of designing. Maybe I will try the simulation using helix coils in HFSS later and see what happens. Anyway thank you very much again for your guidance.

Best regards

Hi. So this is the model I am trying to simulate.( Maxwell file) . If someone could look into it and tell exactly why it is taking so long to do the mesh calculations

I won't expect much impact of using square versus round coil shape. The strongest effect is probably on coil resistance and skin effect. Surely the coil and core edges will increase the meshing effort.

I'm not using HFSS so I can't check you simulation files.

Yes that's true. Actually my problem is that these helical coils aren't being able to do the meshing only no matter what's the shape. Planar helical coils are simulating fine. If anyone has simulated these kind of elongated helix structure, kindly suggest what is needed to be done.