PDE magnetostatic electromagnet how to measure the magnetic field.

Hi dear support from Matlab.

I recently finished designing my step/stl file on an electromagnet.

I got the shape, how do I apply around the electromagnet a coil? Or is this not needed? I would like to know the magnet field between two poles in 3D. Second why do I need boundary conditions?. The whole nabla thing is new to me.

For what I understand is, I upload the stl file in to Matlab, somehow apply the Ampere windings, and permeability for the stl file component. and do a simulation to determine the magnetic field. Maybe also change the geometry of the magnetic poles to increase the amount of Tesla.

I build something from scratch, heat treated (permeability 15000-20000) it and applied 6A at 30V over 500 windings from 0.4mm insulated wire.

[Bastiaan Pierik]

The max amount of permeability for saturation is 2.15 Tesla for pure Iron.

Answer 1

[Ravi Kumar]

Hi Bastiaan,

It might be easier if you do a 2-D approximate analysis. For 3-D, modeling the air around the coil to solve for magnetic field can get very complicated if you are unfamiliar with FEA workflow, particularly dealing with mesh. On the other hand, you can draw simple 2-D geometry using PDEModeler, where you can complete the whole setup for magnetostatics analysis or create geometry using decsg function and use PDEModel object workflow.

Regards,

Ravi

[Bastiaan Pierik]

Hi Ravi thanks, I have some spare time in corona, perfect to do a simulation thanks for pointing That out.

I’ll have look at the 2 solver.

Best

B

Answer 2

[Bastiaan Pierik]

So I tried various combinations based upon the magnetic field of a stator example:

So I have a surrounding Air around the Coil

(R1+IRON+R2+R3-R4)*SQ1

so it cuts a region of interest, ok. I go to pde mode and select the regions that have a 15000 permeability.

at a current density I guess is I * N current x windings correct or how does this correlate?

so I select region 2 and 1 and add 15000 permeability and 6000 Ampere Windings.

how do I know the magnetic field between the poles?

its red and 40. do I need an equation for this?

thanks:) its fun

here is the code for simulation:

%

pderect([-6.0 6.0 -4.0 4.0],'R1')

pderect([-5.5 5.5 -1.5 1.5],' IRON')

pderect([-4.5 5 -0.5 0.5],'R3')

pderect([5.0 5.5 -0.1 0.1],'R4')

%cut out a section from the pde which you require

pderect([4.5 5.5 -1.0 1.0], 'SQ1')

%(R1+R2+R3+R4)*SQ1 which you want to zoom in on.

%

Bastiaan

[Ravi Kumar]

Hi Bastiaan,

Glad to see you make progress in setting up your model. You did not past the full code to fully understand and respond to your question. It would be good if you just save the whole set up in PDEModeler ->File ->Save As..., then I will have better idea of your setup.

Regards,

Ravi

Answer 3

[Bastiaan Pierik]

i calculated the current density which is through a wire 6A at 0.5mm diameter = 3.06*10^13A/m2.

I give in this value into the regions:5 and 6.

the following regions have zero current density

and permeability is 9,1,4,3,2 regions =20000

the regions:10,11,7,8 air/permeability or free space = 4*pi*^10^-7

if I run the pde solver after mesh refinement I get= step size to small...

[Ravi Kumar]

Hi Bastiaan,

I took a look at your model. I think you assigned the material properties (in PDE specification) incorrectly. Air and copper should have permeability close to a vacuum, which is 4*pi*1E-7 (for faces 5,6,8,10,11, and 7), no current for air (for faces 8,10,11, and 7) and a current value of 3E13 for face 5 and -3E13 for face 6 (not the change in sign). The iron core should have a permeability of 20000 and no current. Also, it looks like you selected nonlinear solver based on the example, however, in your case call properties are constant, so you don't need that. With these changes, I obtain the following results (I have attached the code file also):

This is still not correct, but I think you are progressing in the right direction. One thing I would like you to check is if you have all units in a consistent unit system. PDE Toolbox does not have units, but all your model data, geometry, material properties, etc. should be provided in a consistent unit system. As of this model all material properties are in SI units, this means geometry is also in m, I don't think this is true. Can you update the geometry to be m? And also make sure that the current density is also computed correctly.

Regards,

Ravi

[Bastiaan Pierik]

So I changed everything to the metric system and divided by 1000. meters to mm

based on my calculation I get a current density of 6A/(0.25*pi*0.0005^2) hoewever the values I get for the Magnetic Field are still to high, if I change the current density to around

the magnetic field ABS(H) in which value do I measure ? Gauss i am guessing Gaus

I made the shape of the Poles pointy to see if there is a concentration on flux