Can Software Simulate Variable Magnetic Fields in Thin Film Sputtering?

[Dan Kennedy]

Hello everyone. I need to simulate flux densities generated by thin films of permanent magnets of varying composition. This is to guide selection of parameters for sputtering of Nd-Fe-B permanent magnets. Does anyone know of a program which will allow me to simulate fields from permanent magnets of an arbitrary geometry? And is there a way to vary the magnetic properties? Basically what I'd like to do is model a process in which sputtering is used to create regions of different magnetic fields by varying the process parameters (ie carrier gas pressure). Thanks for any help anyone can provide.

 

[jsgruszynski]

The realm of simulators appropriate is pretty limited. http://www.ansoft.com/products/em/maxwell/" does 2D, 2.5D and 3D magnetostatic simulation on arbitrary geometry. There are some similar competitors but most other simulators presume something about the symmetry or component shape. Ansys is pretty much the dominant vendor in this space.

Even with Ansys you are just providing macro parameters like mu, Hc, etc. for magnetic materials. And still the elements in the simulation assume "bulk" constant values rather than continuously changing values.

Beyond that, not at all dealing with chemical/process variance of these basic magnetic parameters or their formation of the material. There probably are only empirical or basic physics research sources for that. Or they are commercial proprietary knowledge you'll never see in the open.

Realistically and most likely you'll need to have the equipment to make the films, run empirical experiments, do the measurements yourself, create your own models and build your own simulator to generate inputs to feed something like Ansys Maxwell.

Other industries have such process simulators (semiconductor) but those are mature industries (so lots of time has been available to develop fancy tools) and they generally do not play with magnetic materials much. The disk drive industry does but most of that stuff is kept proprietary because of the competitive value - the industry is too small and has a different technology adoption dynamic than semiconductor.