How simulate a transformer with Maxwell Ansoft?

[lba]

Hi,

I would like to simulate a standard transformer (2 coils + 1 closed iron) with Maxwell 3D (V15) but I don't find tutorial or example. Maybe someone can explain how can I do ? I want to see the flux in primary and secondary. And I want to put a tension in primary and see the current in secondary. Is it possible ? I need 3D example becaus e my transformer have a non 2D piece.

Thanks
Lba

 

[fleem]

You'd use the transient solver. Here are some rough steps (from my notes) for creating one winding:

Create a solid copper ring.

Create a section:
create a coord system that intersect the part that should have a section, and make sure that coord sys is selected: modeler/surface/section

Delete uneeded sections:
modeler/boolean/separate bodies
delete uneeded sections

Assign an "excitation" (which can also be a load rather than a power source)
You can use the ckt simulator or just specify a current or voltage source. For example:
select section
maxwell3d/excitations/assign/coil terminal, and specify total number of turns (you'll specify how many turns in each winding later)
maxwell3d/excitations/add winding
Project tree: MyWinding/assign coil terminal

Now you're ready to transient solve:
First, copy the design into the same project and solve it with magnetostatic solver (you'll need at least one magnet somewhere or it won't solve. You might, for example, slightly magnetize the transformer core--just don't do it in the transient design!). You do this solely to create a reasonable FEA mesh.
(There is a bug who's work-around requires that only the transient and magnetostatic designs for this model be in the project--no other designs should be in the project.)
Then go back to transient design and create an analysis that imports the mesh from that magnetostatic design.
Before solving, IF you've changed any geometry, be sure to click 'Clear linked data' in transient solution, or it might use data with wrong geometry.

If you get an error about "mesh quality" not accepted by solver, then delete the output directory and resolve.

 

[lba]

Ok thanks :) , I will try your solution, but there no secondary in your system ?

 

[fleem]

Ok thanks :) , I will try your solution, but there no secondary in your system ?

Just do the same steps you did to make the first coil to make the second coil. And it will have a different 'excitation', of course. (Remember an excitation can be a source or a load). You don't have to tell Maxwell anything about the fact that its a transformer. Maxwell's job is to simulate the real world (EM-wise, anyway) and you don't have to tell the real world that two coils near each other are a transformer.

 

[lba]

Ok, I will try but how can I add a little

slightly magnetize

? I create new material ?

 

[gerbi]

Hmm.. some strange anserws here.

First, I hope You have some basic knowledge about how transformer works - I'll skip the basic principles and get to the point.

If You have Your geometry ready (two windings and core reduced by half) go to Transient solver, unless You want to be more precise and check solution in magnetostatic (You'll need to obtain number of turns and currents for both windings).

If You want to be strict, go to magnetostatic and set ampere-turns on both windings (opposite directions). Check the flux in core, generate proper mesh in volumes.

If You want to simply check how flux flows go straight to transient. Here is part of one of my previous post here on the board

[...]
4. Create "region" that will connect magnetically all solution domains (draw->region)
5. Now select the faces one by one and create coil terminals (select faces -> left click to select face -> right click and assign coil terminals. Remember about proper names and direction of the terminals, I mean in-out. like the current flows).
6. Now go to 'excitations' in Project manager. Create two windings and attach coil terminals to them. In windigs excitations select solid or stranded (solid conductor or parrallel conductors) and "external" type of winding - it means it will be connected with external electric circuit.
7. Launch Maxwell Circuit Editor. Create proper circuit (one winding excited from voltage or current source, second winding connected to some impedance load). Remember about naming correctly Your windings, assigning current source and zero point to each independent circuit.
8. Save circuit and export "net list". Go to Maxwell and import it (Maxwell-> Excitations->external circuit->edit external circuit->import circuit).
9. Set up "solution setup". Choose step interval and stop time (depends on frequency You have selected for Your current source). At least 20 steps for period will be fine. Then go to next tab - "save fields". Choose the same time step and stop time (watch units) and add this to list.
10. Validate Analysis and solve it.
11. Go to results in "project manager". Create transient report-> rectandular plot and select quantities You'll like to observe.

Got any qustions - ask

 

[lba]

The external input is ok with voltage from Maxwell circuit editor but the I want to charge the transformer with a resistor and Maxwell give me an error see below the drawing please. How can I put a resistor to the secondary winding ?
thanks :)

 

[lba]

i have find a solution for the last problem, need to have Winding 1 and 2 in the same circuit editor.

but i have another problem (see the drawing) maybe someone know what ?

 

[gerbi]

The external input is ok with voltage from Maxwell circuit editor but the I want to charge the transformer with a resistor and Maxwell give me an error see below the drawing please. How can I put a resistor to the secondary winding ?
thanks :)

It looks like the name of Your coil in Circuit Editor is different than this one in Maxwell 3D (they must be the same).

Using resistor as a load of transformer ? Nothing more simple, connect it in series with winding. No "magic" here ;)

Regards

 

[gerbi]

i have find a solution for the last problem, need to have Winding 1 and 2 in the same circuit editor.

but i have another problem (see the drawing) maybe someone know what ?

Add a minimum resistance (like 1e-6 ohm) in series to circuit with voltage excitation. This small resistance has almost no effect on results but helps Maxwell to start simulation

Regards

 

[lba]

ok, it's better I don't have the error about loop current but I have always the first error in the message "don't have enough memory or killed by user".

thanks for your help

 

[lba]

ok, all is fine it was the mesh I think (different) but now how see the current and the tension ?

 

[gerbi]

ok, all is fine it was the mesh I think (different) but now how see the current and the tension ?

Create a Rectangular Plot and select what do You want to see.. (in results, project tree).
Watch what are you selecting, some of the currents come from maxwell 3D design (maxwell always calculate current and voltage on windings, other come from Your circuit -> those are named branch currents and node voltages).

Regards

 

[lba]

ok, thanks, great ! :)

now, I would like to put 1000 turns to the winding but Maxwell said it's not possible because it's only 1 turn for solid winding, it's possible or not ?

 

[gerbi]

ok, thanks, great ! :)

now, I would like to put 1000 turns to the winding but Maxwell said it's not possible because it's only 1 turn for solid winding, it's possible or not ?

"Solid winding" means that it's one mass of copper - so it has to be just one turn.

Check "stranded" (multiple parallel wires in selected mass) as winding structure and You can change number of turns (on both coil terminals of coil).
For next projects use rather "stranded" than "solid" unless You have a reason.

Glad it helps
Regards

 

[MaxwellFan]

fleem, how exactly to overcome the Import mesh" bug in Maxwell, transient analysis. I'd like to import mesh from eddy current into transient and kept only these two analyses in the project directory (as you suggsted), but clicking on the "setup link" button doesn't open the browser (as it used to do in the previous versions)

 

[lba]

Hi,

All seems good except the voltage at secondary => µV or mV when I charged with 10 Ohms, and if I put R = 10000 Ohms the voltage is good ! what's wrong ?

 

[gerbi]

Hi,

All seems good except the voltage at secondary => µV or mV when I charged with 10 Ohms, and if I put R = 10000 Ohms the voltage is good ! what's wrong ?

Have You designed this transformer or just draw it in Maxwell ? Are You EE ?

 

[lba]

Hi Gerbi,

I'm teacher at University, I have an "Agrégation" in Electrical area (France). I now 2D simulation but I'm new in the 3d.
I have attached the project in this message. I don't have this transformer in reality, I test before build some transformers.

P.ZIP is the project
P MESH.ZIP is the mesh
P_circuit.ZIP is the circuit

I hope you can help me.

Thanks
Regards

 

[gerbi]

Hi Gerbi,

I'm teacher at University, I have an "Agrégation" in Electrical area (France). I now 2D simulation but I'm new in the 3d.
I have attached the project in this message. I don't have this transformer in reality, I test before build some transformers.

P.ZIP is the project
P MESH.ZIP is the mesh
P_circuit.ZIP is the circuit

I hope you can help me.

Thanks
Regards

1. In "almost no load state" (R=1000) there is no voltage drop at secondary. Voltage difference between primary and secondary (both have 100 turns) comes from magnetic circuit design (not all flux links both windings).
2. In load state (R=10) You have almost short circuit. Voltage drops since the current goes above it's nominal value.

I see no simulation/Maxwell problem here. The problem is with the design.

PS.: If You want to observe how things change delete Your mesh operations (Maxwell will go weel enought for some coarse pre-analysis) and go with initial mesh. Change the time step to 0, 0.5, 1 ... 40 ms and You can do "quick change and test" analysis.
One more thing - in circuit editor change excitation to 12*sqrt(2)=16,97V (I bet You wanted to have 12 V RMS but You need to put amplitude in the excitation).

I also suggest gaining some more practical knowledge about transformers designing before building one.. You know, Maxwell is a nice tool but the truth is that, it's not a design tool.

Hope it helps, If You've any more questions - let me know.

 

[lba]

ok, thanks Gerbi but why:

2. In load state (R=10) You have almost short circuit. Voltage drops since the current goes above it's nominal value.

?

What do I need to do for take big current in the secondary, it is the number of turns ? I don't understand where my error is ...

regards

 

[gerbi]

ok, thanks Gerbi but why:

?

What do I need to do for take big current in the secondary, it is the number of turns ? I don't understand where my error is ...

regards

Modify the core

EDIT: ..and I mean material, shape (are those extra limbs needed ?) and it's cross secion (now You have rectangular core and windings - that also affects the results). Wound the windings closer to the core - some flux goes around winding because it's shape (look at the B distribution in X-Z plane).
In small electronic transformers this all may work - with low efficiency. In power transformers the design prinicples are completely different.

 

[lba]

I don't know how have B distribution

And how can I have the self equivalence in mH for each coil ?

 

[gerbi]

I don't know how have B distribution

And how can I have the self equivalence in mH for each coil ?

Select Plan X-Z then click Field Overlays in Projct Tree -> B -> select time moment in bottom left corner of main window for which you want to see field distribution

There is no easy trick to get inductance here.

 

[lba]

but I don't put time, the system don't want, the time is always -1s. I don't understand :"Select Plan X-Z" could you explain please ?

 

[gerbi]

but I don't put time, the system don't want, the time is always -1s. I don't understand :"Select Plan X-Z" could you explain please ?

Project manager -> Analysis -> Setup -> Save fields -> select time step and stop time the same as in the general setup -> click add

In the material tree You cal select Planes -> Global X-Z

 

[MaxwellFan]

Dear Gerbi, do you know by chance the work around of the import mesh bug I mentioned in post 16

 

[gerbi]

Dear Gerbi, do you know by chance the work around of the import mesh bug I mentioned in post 16

Hello

sorry. but no

 

[lba]

Ok, all is fine for transformer Gerbi ;) Thanks !
Now, I would like to simulate a simple motor (primary, iron and magnet). Is it possible to simulate the rotation of the magnet ? And how can I do ?
I would like the couple of forces on the magnet but I don't see it in the report.
Regards

 

[gerbi]

Ok, all is fine for transformer Gerbi ;) Thanks !
Now, I would like to simulate a simple motor (primary, iron and magnet). Is it possible to simulate the rotation of the magnet ? And how can I do ?
I would like the couple of forces on the magnet but I don't see it in the report.
Regards

My pleasure ;)

Know what.. write my Your email address in Private Message and I'll send You some tutorials about motors, ok ? I don't want to write a book here on forum ;)