Magnet Demagnetizing from to much power not heat?

[Arlo1]

Ok So one of our members has posted on diy electric car forum that he doesn't want to over power his BLDC motor because it will demagnetize the magnets from to much power run into the stator... Not the heat the stator makes when it gets hot but the actual magnetism it self! I have seen a few people claim to be worried about this now and are afraid of pushing permanent magnet motors to hard because of this. Is this real or a misconception?

Biff???

 

[Njay]

It is real that a strong magnetic field can demagnetize a magnet - that's actually how they are created, by giving them a very strong magnetic field.

Now, if in normal conditions on a motor such strong magnetic field can be created, I don't know... although I know that heating the magnet will allow a less strong magnetic field to have a permanent effect on it.

 

[grindz145]

Yeah it's very real Arlo. http://www.electronicsteacher.com/succeed-in-physical-science/magnetism/magnetic-materials.php

 

[Arlo1]

I do understand they are created this way. But is it really possible to demagnetize them inside a motor without heating them. For instance look at Luke running motors more then 5-10x their rated power... I have put 30,000+ watts into my x5 and they never demagnetized even though the most power it makes is at ~10,000 watts. The stator it self could not make anymore magnetism it just turned to heat.

Does anyone have realy world proof of the magnets demagnetizing from to much opposing magnetic force inside a motor???

 

[ZOMGVTEK]

Is it seriously practical to get those kinds of magnetic fields out of a motor without the copper melting?

Anyone have a ballpark for the relative amount of power you need to demagnetize the magnets in a typical design? 100x nameplate? 1000x?

People around these parts run 10x the power the motor was intended to handle without much fuss.

 

[John in CR]

During BLDC commutation I thought it was always an attraction force from the electromagnets created in the stator iron. Doesn't it require a repelling magnetic force to demagnetize a permanent magnet?

 

[Arlo1]

During BLDC commutation I thought it was always an attraction force from the electromagnets created in the stator iron. Doesn't it require a repelling magnetic force to demagnetize a permanent magnet?

Technically it does both. It attracts one magnet while pushing the previous one away.

 

[amberwolf]

AFAIK it would take quite a field to induce a change in the field of the magnets themselves, but I cant' find the info I once read on that. The closest thing I can find is this:
25. Will neodymium magnets lose strength if they are held in repelling or attracting positions for a long time?
In most applications, the answer is simply "no". If the magnets will be exposed to higher temperatures while in repelling applications, the answer is "possibly". The exact answer is a bit too complicated for a FAQ answer, and requires specifics about the application.

Then this page:
http://www.duramag.com/neodymium-technical-info.html
which defines some of the levels of "Hci" or "intrinsic coercive force", which is the level of field needed to overcome the existing magnetic field and remove it. Unfortunately I have no idea what level of field the motors output at any particular power level, or how concentrated it is at the magnet location relative to the stator teeth and coils.

 

[Mr.K]

I think it's more about the direct axis current than the power, if you drive the d-axis current past the MTPF line there is a chance of demagnetization. But not all motors has the short circuit point inside the maximum current circle and will therefore not be susceptible for demagnetization, they will probably burn the windings before it happens.

https://dl.dropbox.com/u/13304011/demag.PNG

 

[bigmoose]

Mr.K best first post I have ever seen! Any chance of you uploading the graphic to the Endless Sphere Server? Something tells me you are a motor designer...

For Info: MTPF = Maximum Torque per Flux Linkage Line

edit: Thanks for the graphic!

PS: A question that you might be able to answer. How come removing the rotor from a stepper motor permanently "knocks down" the magnetic strength of the rotor magnets when you put it back in if you pull it out without a flux "containment tube" around it.

 

[Njay]

Ok, perfect time to sneak-in something I saw a few years ago:

http://www.phys.unsw.edu.au/hsc/hsc/electric_motors5.html

This page shows a pancake motor whose magnets are "made" after the motor is assembled, because for them to keep their strongest field, they must be in a closed magnetic loop. So they placed a tick wire around each magnet so that they can magnetize them after motor assembly. I always wondered why a Lynch type motor can be opened without "magnet issues".

 

[Arlo1]

So... To Answer my question its possible but how possible lol. I now have to understand the graph that looks like a compressor map for a turbo.
Thanks MrK.
But I do think its safe to say the windings will be hot before hurting the magnets right?

 

[Kingfish]

I’m not professionally steeped in motor theory; however from my readings I can imagine another way to evaluate the problem. Keeping it simple, a scenario that could demagnetize is through high flux density and through losses created by hysteresis.

We have two circuits: Electrical and Magnetic. Electrical handles current to produce flux. Large current with particular design can produce very high flux without damage to the system. We can easily envision that with confinement of plasma in a magnetic bottle. 8)

In the magnetic circuit using PM such as in a BLDC motor, if the back iron is large and with high permeance but the magnets themselves are undersized, I can envision two modes for failure: High Flux density due to inferior magnet construction – or simply insufficiently rated for the power, and through high-frequency hysteresis. Both cause losses (heating) within the magnets that are not directly related to losses within the copper core, however are associated with iron motors (stators, back iron)… and more to the point – a bad design or operator exceeding designated parameters.

When we consider how to extend the performance of a DD hub, I often though it would make more sense to add fins (or vanes) perpendicular to the back iron ring and parallel to where the islets of the spokes mate; I’ve seen this on some electric motorcycle hubs. The laminar flow of air pulls heat away from back iron which is heated by flux and from bonded magnet materials through conduction, and less effectively from copper coil convection and radiation.

However, to be concise - it seems to me that if we are concerned about having such issues right out of the box then don’t buy the product.

My ½ μh, KF

 

[Arlo1]

However, to be concise - it seems to me that if we are concerned about having such issues right out of the box then don’t buy the product.

My ½ μh, KF

Its not a question to do with purchas of a motor it seems to be a common concern with people on BLDC motors and I have a feeling its a misconception.
Here is one eg. from Ryan800 on DIY electric car..

No, the DC/DC converter more than doubles the voltage to the controller, to as much as 650V (I think the output voltage is actually variable depending on need, but I could be wrong). 500V max in early Prius models.

Be careful though. The full 650V will give you more torque at higher rpm, but it won't increase the low speed maximum torque. You have to run more amps through the motor to do that, which at some point will demagnetize the magnets. Tritium will supply up to 360Arms, so that's not the problem. I have it limited to 80Arms right now, and it's slow. And my car only weights 2300lbs/1050kg. I could increase that, especially in winter, but I hate to find the motor's limit the hard way. Maybe 100A would be fine, but if not I've just ruined my motor. remember that it's only rated at 100ft-lb, which is not much starting torque with 6.8:1 ratio.

Basically, it's a neighborhood vehicle until I get a bigger motor in there. I know you spent a lot to get that motor over there, but maybe Ford sells hybrids there? Those are designed for 300V and good power, so they're a good match for the Tritium. You just have to remove the extras.

One last warning is make sure the controller you buy is compatible with IPM motors. I'm discovering that they're a little trickier than induction or surface mounted PM motors.

That said if the price works out and you're ok with sluggish performance, go for it!

http://www.diyelectriccar.com/forums/showthread.php/any-use-highlander-rear-diff-motori-43920p5.html

And this post from Ripperton

I didnt mean from being overheated, I meant from being overforced ?
there is so much electro magnetic force coming from the stator that it wipes the magnetic field out of the permanent magnet.
I still have this rotor, it would not have enough power to ride across flat ground.
But this was when I had the 120v 1000A Kelly in there and I think it was turned up full.
This is why I cant understand why John Fiorenza is making more powerfull rev-A stators but the magnets are still that same 7mm thick ones.

http://www.diyelectriccar.com/forums/showthread.php/ripperton-electric-track-bike-41173p71.html

 

[Kingfish]

Ahhh thanks Arlo, that’s basic

E = mc², P= IV. Unless we’re at 0°K there will be losses from shoving too much @#$% through sock if it ain’t built for it; things will get hot. In that application it appears the original magnetic design can’t handle the high flux. I have to ask: Do these guys have disposable income? :lol:

With FEMM, I had toyed using thicker magnets for RF and for AF, and I didn’t see as large of an improvement with RF as I did with AF. That said, there comes a point where momentum becomes an issue – and there is good reason why we want thinner magnets. Those guys on that other forum are asking for too much from the design.

In freewheel, KF

 

[liveforphysics]

I can tell you from experience, I've never damaged a magnet from anything but radical over-temperature.

I've also heard it's possible to do, but I've done massive current tests, driving so far past saturation on the tooth that adding another few hundred phase amps makes a barely measurable difference in the torque to cog-over the motor, and the windings go from room temp to smoking the 220degC insulation in just 2seconds, and still never had the magnets get damaged (which would be easy to see because the Kv would increase).

I think you might need a stator specially designed to have extremely high saturation points to even create a situation that makes it possible to damage a magnet from field strength. Perhaps requiring higher saturation point alloys than anything we've got in RC and Ebike motors as well. I've got a 1,000amp constant current supply setup at the moment I can stick on any motor winding and see what happens (find saturation torque knee etc), so far nothing has caused detectable magnet damage that I've seen or tried.


Mr.K - Awesome contribution, I hope to see more from you in the future.

 

[bigmoose]

The reason I asked about the "knock down" on the stepper motors, is that I did it. About 15 years ago, I didn't know any better and took one apart. It only had about 60% of nameplate torque after I stuck the rotor back in and it met nameplate torque before. Always wanted to understand what happened there.

I contacted a compact BLDC manufacturer when we were disassembling those for issues, and they said no problem with BLDC only issue with stepper motors... but they couldn't tell me why.

 

[amberwolf]

How come removing the rotor from a stepper motor permanently "knocks down" the magnetic strength of the rotor magnets when you put it back in if you pull it out without a flux "containment tube" around it.

Are they AlNico magnets? Cuz according to that page I'd linked before, those can get damaged from opposing forces easily, like forcing two north poles together will permanently reduce the strength of both magnets. They ship with "keepers" to make sure that doesn't happen unintentionally.

So maybe something happens to the shape/strength of the field without the stator present?

 

[roller]

amberwolf is onto it there ; Coercivity is the mag force that weakens a PM ("intrinsic" coercivity seems to be a theoretical absolute maximum, practical results for manufactured goods is lower for Neo). It is a known property with a known value, thus it is *possible* to demag by field alone.

Neo coercivity values range from 750 - 1000 kA/m (9.5 - 12 kOe), increasing with grade (strength) of magnet. Typical grade for Neo magnet in most Chinese motors is N35 (33~35 MGOe), at lower end of coercivity (which is still about twice the coercivity of SmCo).

Ignoring heat (usually a bad move), 'typical' N35 mags should tolerate a field of 750 kA/m w/o demag (for how much time?) Will it see this much field? Up to winding characteristics and V, I . . . Typical saturation for Si Steel is on the order of 1.4 Tesla = 1.4x10^4 Gauss. Is that enough to demag Neo on its own? Maybe.

In my experience, however, we can't ignore heat and time (and frequency and rate of change, etc) . . . so the answer is even less straightforward. I did an experiment which sadly showed -- though I had never seen quantified in literature -- that coercivity changes with temperature. A hot magnet is "softer" than cold . . . and can demagnetize in the presence of a field that is otherwise not strong enough to do that. So, in the real world, it's a function of field x temp x time x planetary alignments, etc.

Attached pic shows (in the fuzzy light green vertical band) changes in magnet polarity. Only 1 edge of each magnet went to 0 or reversed . . not the center of magnet . . not the entire surface . . . I assume temperature was uniform, but field (or flux density) is not -- the edges do more work, as a motor is turning. the greatest flux exists between neighboring magnets - not at the center of a magnet.

 

[liveforphysics]

Just curious, has anyone had an experience with modern Nd based magnets having this problem?

Perhaps it's a legacy thing from days of magnets with field strengths so much lower relative to the saturation limits of iron that it's something that only used to be a concern back in the days before proper motor magnets.

 

[flathill]

I think the brammo race bike with the parker motion motor had this problem until they upgraded the magnets

Not usually a problem unless you are doing field weakening on a hot motor
Although extreme field weakening could possibly do it on a cold motor
Or you loose timing on a hot motor
Or have a controller glitch on a hot motor

It is also possible to demag with physical shOck
dropped motor on concrete

but all magnet motors wont work in a few millennia if you are are thinking of making a spaceship planet for when the sun starts acting up

 

[whatever]

just a quick google search gives some answers
http://www.x-magnet.net/faq.htm

 

[grindz145]

Wow. That is an epic first post. And then... Poof.

Anyway, I was definately under a misconception that it was easier to damage magnets due to high temperatures. It's good to have some realworld datapoints for sure.

 

[Arlo1]

Wow. That is an epic first post. And then... Poof.

If you are talking about Mr.k lol. Its all good. Its really cool to think who might be watching... I had a guy come to work telling me his buddy was raving about this forum and I asked if it was ES has was so stocked to know I am a part of this. He wanted info on a diy motorcycle project and he came to the right place. I am building an army here on the island one dude at a time.
But as for Mr.k well sometimes we have to accept there is some very busy guys who might pop up and some of them have signed a lot of NDA agreements and can only help so much. If it wasn't for some of the rule breakers I would be almost nowhere with my DIY controller project.

 

[Mr.K]

I didn't expect to be this popular I'm not a motor designer, but I'm doing my master thesis about control of salient pole permanent magnet motors right now.
The graph is from this book, but i don't think they cover demagnetization of the magnets in it.

Advanced Electrical Drives: Analysis, Modeling, Control
http://www.amazon.com/Advanced-Electrical-Drives-Analysis-Modeling/dp/9400701799