Magnetic field simulation

[HeavyIon]

How to simulate the speed of appearance of a magnetic field, for example, in a conductor that is surrounded by a solenoid? I try to use CST Studio, but I did not found suitable solver or I did not understand how to set the right impulse.

 

[argotier]

Interesting question.

For most engineering applications you can consider that the field establishes instantaneously, as the velocities involved are pretty high. I'm wonder what kind of problem are you working on, care to share?

But more precisely there is a finite speed associated with the propagation of electromagnetic fields (waves actually), speed of light if the medium is vacuum or more generally:

1/√(με)

Being μ: medium's magnetic permeability and ε: medium's permittivity.

But be aware that if you are dealing with an iron core, for example, things will get much slower, as the iron magnetic domains will take some time to align with the inductor field.

In the physics realm I get that this matter is not so simple.

 

[waross]

Google "Induction Coil Time Constant".

This time constant τ, is measured by τ = L/R, in seconds, where R is the value of the resistor in ohms and L is the value of the inductor in Henries. ... The transient time of any inductive circuit is determined by the relationship between the inductance and the resistance.

In one time constant the value will reach 63% of the difference.
In the first time constant both the current and the magnetic field strength will reach 63% of the terminal value.
In the second time constant the value will reach 63% + (63% x 37%) = 86.3%
In the third time constant the value will reach 86.3% + (63% x 13.7%) = 91.4%
For most applications it is accepted that the voltage, current, magnetic field strength or anything else subject to time constants will have reached the terminal value.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[HeavyIon]

For most engineering applications you can consider that the field establishes instantaneously, as the velocities involved are pretty high. I'm wonder what kind of problem are you working on, care to share?

This question arose due to the solution of the beam focusing problem by pulsed magnetic quadrupole lenses. If you look at the focusing lens, then I am interested in the speed of the appearance of the magnetic field in working area of the lens per 1 ms pulse. The core of the quadrupole lens will be made of arrange in layers of iron.

 

[IRstuff]

It seems to me that the time delay is substantially more due to the electrical componentry vs. actual propagation of the field through free space.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[waross]

Hi IRstuff: As the electrical current builds the magnetic field, the magnetic field is inducing an opposing EMF in the electrical circuit.
Many instantaneous events, if analyzed on a suitably expanded time scale will be seen to actually rise on an inverse exponential curve.
It is relative.
If you are wanting responses in the range of 100 milliseconds, a time constant in the order of nanoseconds is relatively instantaneous.
The time constant of the magnetic circuit is independent of the time constant of the electronic circuits behind it.
Hopefully, the time constant of the electronic circuits is instantaneous relative to the time constant of the magnetic circuit.
If not, the math becomes interesting.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[IRstuff]

As the electrical current builds the magnetic field, the magnetic field is inducing an opposing EMF in the electrical circuit.

There are other delays in the circuitry, propagation delays in the control circuitry, as well as RC time constants in the coil drivers.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[waross]

There are other delays in the circuitry, propagation delays in the control circuitry, as well as RC time constants in the coil drivers.

Yes. Don't forget the L/R time constants

The time constant of the magnetic circuit is independent of the time constant of the electronic circuits behind it.

I guess that I should have said "total delays"

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[Compositepro]

The response of the magnetic field to changes in current is at the speed of light (which varies with medium of transmission). The response of the current though the coil with time is what you need to calculate. That depends on the inductance and driving voltage. The voltage can be increased to drive a faster change in current, and then reduced to prevent overshoot.

 

[waross]

The voltage can be increased to drive a faster change in current, and then reduced to prevent overshoot.

That works.
In the industrial motor field it's called field forcing.
Field forcing may also be used to reduce a magnetic field quickly.
In some applications the magnetic field does not decay fast enough and must be forced down with reverse current.
The largest application of that that I have seen was field forcing on four interconnected DC generators and four motors on a large drag-line to improve control response.
Total HP 5200.
There was also a second smaller set of four generators and motors with a total of 4000 HP.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[argotier]

I agree that in this case the most relevant time delays will be from the circuitry and not the field itself.

Good news is that this effect will be taken into account by most simulation tools.

You probably will need some fine adjustment later on when testing the equipment, depending on your required precision.

 

[HeavyIon]

Good news is that this effect will be taken into account by most simulation tools.

Which simulation tools, for example?

 

[argotier]

Which simulation tools, for example?

Well, from the humble PSPICE and beyond, all will handle the RLC time constant of the circuit and take it into account from the voltage wave generator (whatever might be) to the current developed when the coils are energized. Here a timely example.

My only doubt is how (if you can) to link all this with the magnetic field simulation in one tool. I'm not familiar with CST Studio, but it looks pretty good.

And as IRstuff and waross pointed out you still have the control circuit delays, wich I don't know if can be simulated (maybe estimated?) before hand. That's why I said that some final adjustments will be probably needed nonetheless.

 

[IRstuff]

There are mixed signal simulators based partly on SPICE that can handle everything from the switch to the driver attached to the inductor.

From there, you potentially have a choice of coming with a black-box inductor/magnetic field function in the circuit simulator, or running iterations between a FEM model and the circuit simulator.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[HeavyIon]

Yes, of course, there are delays in the control circuit, but in the beginning, I need to estimate the time the magnetic field appears in the center of the lens.
PSPICE, as it seemed to me, is a tool for modeling in electronic circuits, whereas here I need a tool for modeling physical processes in volume. I could not create the required conditions in CST, now I'm trying to use ANSYS, but I have no experience there. Still there is COMSOL, but there is also almost no experience there.

 

[IRstuff]

It's already been mentioned several times that magnetic fields propagate at the speed of light in the medium; do you need to parse that further?

You said that your signal (electrons?) come in 1-ms pulses, so why is a few nanoseconds of field propagation of such interest?

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm