BMC 1500w Magnet Expoy failure / motor failure

[magudaman]

Well this is a really big bummer. I was out riding the new prototype around tonight and my controller shut down with a hall sensor error. I made it back to the shop to try to figure out what happened.

Well after doing some test with a multimeter I cracked the motor case open to see magnets pushed out of their spots cutting open wires !!!

So here is the carnage:

So what the heck happened?!?! I have used this motor for over a year on another bike with a weaker controller and have not had a single problem. It got hot quite a few times on that bike with out issue (85c max). Now I put it on this new bike with a bigger kelly controller more airflow and from what I could tell was staying quite cool then this! Is it possible the additional magnetic forces are broke the expoy bond? Or was it get very very hot inside with out me even knowing it was happening?

Any idea appreciated...

What are hopes of repair? I tried pushing the magnets down with my hands with no luck. I would be afraid to hit them with anything, maybe an arbor press?

 

[Gordo]

WAG;
Either the epoxy let go from the heat or the expansion from heat broke the bond. Does it really matter what caused the problem? Aren't you looking for a solution?
You can press the magnets back into position with the jaws of a vise or a large "C" clamp. But first you have to get ALL the magnets out, clean the case and magnets and find a hi-temp product to stick them back in. Then you will have to devise a method to monitor the temperature and limit it.

 

[magudaman]

Any suggestions on getting them out with out cracking them? Guess I was venting since I really can't afford a new motor.

 

[flyinmonkie]

WAG;
Does it really matter what caused the problem? Aren't you looking for a solution?

If you know what happened, then you what to do for a solution and prevent it from happening again. No go to just glue them back in and have it happen again.

magudaman...
Sorry, not much help with magnets flying out of motors. Hope you get it sorted soon.

Clay

 

[magudaman]

What about pressing the magnets back into place and then creating a retaining ring , say out of aluminum to hold them down, keep them from moving up. Of course I am not sure of the forces at work here so maybe that would not be strong enough. The remaining expoy would keep the magnets from spinning round and round.

 

[neptronix]

haha, how about not overheating it in the first place?

That is usually how magnets get detatched, from what i have seen.
Looks like you may have to re-heat it and then press the magnets back down.
What could have happened is the the glue re-melted after it cooled down... hence they are stuck in an odd position now..

 

[magudaman]

Hey that would have been a good idea :wink: Well I really think the bond was destroyed before I had my temperature sensor installed in the motor. The magnets probably held fine under the lower stresses of the smaller controller but as soon as I went up to the 6kw controller it was having issues.

 

[whatever]

1500watts thru a motor designed initially for 250watt continuous, something had to give,

 

[texaspyro]

Just looking at, I don't think that they were ever epoxied in. They might have been tacked in with some superglue (which does not care for heat).

 

[Gordo]

Any suggestions on getting them out with out cracking them? Guess I was venting since I really can't afford a new motor.

You should be able to get the displaced magnets out by heating the outside of the housing directly where the displaced magnets are. Toss the motor in the freezer for an hour, take it out and quickly heat the small area. DO NOT HEAT THE HOUSING ALL THE WAY AROUND THE CIRCUMFERENCE. Mark each magnet with a felt pen so you don't flip it over by mistake when you put it back.

I don't think the adhesive is a type that is reusable by heating it. I am guessing it is a type that has cured and needs to be cleaned out of the housing and off of the magnets. Use 2 part hi-temp epoxy and a small "C" clamp to hold the magnet in place while this new adhesive cures. Use small pieces of soft wood between the jaws of the clamp, the magnet and the housing. Magnets are easily broken by distorting them sideways, so be gentle.

 

[magudaman]

1500watts thru a motor designed initially for 250watt continuous, something had to give,

Hmm not sure where your 250w number came from but these are rated from the factory at 1500w continuous.

The motor is rated for 1000 watts when used with the internal controller. Once the internal controller is not used and not feeding heat into the motor the continuous rating jumps up to 1500w.

This is model 1 of this series, 2 is 3000w and 3 is 4500w. Of course I am running this motor at a peak of 6KW so your numbers sort of work, 6x the rating.

You should be able to get the displaced magnets out by heating the outside of the housing directly where the displaced magnets are. Toss the motor in the freezer for an hour, take it out and quickly heat the small area. DO NOT HEAT THE HOUSING ALL THE WAY AROUND THE CIRCUMFERENCE. Mark each magnet with a felt pen so you don't flip it over by mistake when you put it back.

I don't think the adhesive is a type that is reusable by heating it. I am guessing it is a type that has cured and needs to be cleaned out of the housing and off of the magnets. Use 2 part hi-temp epoxy and a small "C" clamp to hold the magnet in place while this new adhesive cures. Use small pieces of soft wood between the jaws of the clamp, the magnet and the housing. Magnets are easily broken by distorting them sideways, so be gentle.

I think you are correct about the adhesive, it seems to be an epoxy type stuff. I may just try to press the magnets that are already sticking up back down while putting a bit of 2 part high temp stuff around the areas that are exposed and press them into that. Then I was thinking about putting that same high temp stuff across the tops of the magnets where they were sliding out along with aluminum retaining ring and a bunch of fasteners CNC drilled into place.

I may also have a source for a new one of these motors with a different winding count. I could always pull out the rotor from that one and toss it into this one but that isn't a cheap alternative. What would the difference be between a 18 turns of wire verses ones with 12 turns and a heavier gauge?

 

[Gordo]

Most 2 part adhesives expand as they disintegrate when they are overheated. This is how you make them release. I would get the magnets out, clean the housing and magnets and start again.

On the effect of the number of turns, do a search. There is tons of info here.

At some point, jamming more power to a motor does not give you more torque. Once you saturate the field, you are just producing more heat. Your efficiency goes all to hell and as you have found out, so does the construction of the motor. Maybe you want to rethink what you are doing? It is one thing to say you can't afford a new motor, and another to keep destroying the one you have by exceeding it's ability to survive overpowering it.

If you pound on your finger with a hammer, it will not hurt less the longer you do it.

 

[Alan B]

...

I may also have a source for a new one of these motors with a different winding count. I could always pull out the rotor from that one and toss it into this one but that isn't a cheap alternative. What would the difference be between a 18 turns of wire verses ones with 12 turns and a heavier gauge?

It will produce the same torque (as the first motor) at 12/18ths of the current. It will produce 12/18ths of the speed at the same voltage, or require 18/12ths of the voltage to produce the same speed.

 

[magudaman]

Thanks Alan! That is very simple, and I wasn't really finding it doing searches, hence why I asked.

I ended up not pulling all the magnets bc I tried to get the one that was sticking up the most but had little luck, I even managed to chip off a tiny piece of the outer coating. They are press fit into the ring in addition to being glued. So I got some high temp glue coated the exposed surfaces after roughing them a bit, filled in gaps elsewhere and press the magnets back down. I've let it cure for a couple days and repaired my damaged hall lines. I think I am going to put it back on the bike today and drop the current levels down 3000w or so and watch my temperatures closely.

Thanks for all the help!

 

[Alan B]

No problem neighbor. Someday I want to see this machine!

 

[magudaman]

Gordo seems to have had the right idea from the beginning I opened the motor back up after a couple more rides and sure enough the magnets had moved again. SO today I ended up risking breakage of the magnets and pull them all out. A couple ended up with a some very very small coating chips but basically came off unharmed! So now I need to clean the thing up real good do some roughing and find a good adhesive. I of course didn't mark the magnets direction as I a bit flustered but they just each are either push or pull and don't seem to have two poles (except for the back side) . I know for a fact every other one on the rotor alternated push pull.

Is there a suggested adhesive to get this babies back in? I think I'm going to use Loctite 648 which seems like a good fit but loctight actually makes what they call magnet adhesives like 331 and 3060. Those magnet adhesives seem really bad when I started looking at the data sheets as there high temperature strengths dropped to 15% at 100c! I don't plan on running them at this temp but it could hit that again.

I believe the rotor and magnets are coated in what looks like Zinc Dichromate which only give me a 75% strength with 648 but maintains 100% strength at a continuous 150C, so it doesn't break down over time at heat of 150c.


Honestly I think this motor has the potential to run 6kw as my torque kept increasing but of course getting the heat out of such a small space became more of a problem. Man I really want to get a dyno!

 

[Gordo]

Did you pull them out cold or heat them a little first? Even a hair drier will often destroy even an epoxy bond.

 

[magudaman]

Warmed them in the oven at about 175F. All came out but one. I then tried doing the freezer and blow torch to quickly warm it up, but not too hot. Still didn't really work, but I wasn't really getting it too hot. As I continued to pry away it finally broke off fairly violently and luckily it didn't break. I plan on ordering the glue tomorrow.

 

[whatever]

there is a glue they use on aircraft, its made by 3m cant remember the exact name but can look it up if of interest, its one hell of a strong glue. Your mags are fine, just pop em out, even if they crack you can glue back in place. just need a damn good glue
make sure you get polarity correct as you take mags out, mark each one with permanent marker showing how they came out,

 

[salty9]

You might want to test the magnets to see if the heat has weakened the magnetic field. If the temp got close to the magnet's Curie point it could be weakened or destroyed.

 

[bigmoose]

Here is the best magnet bonding epoxy out there: Hysol EA9394 177 deg C capability

NASA Use reference (Page 264, marked 252-3 in report) http://www.scribd.com/doc/40109738/NASA-Glenn-Research-Center-2006-Annual-Report

Data sheet: http://www.mtpinc-exporter.com/chemicals/tds/Hysol%20EA%209394.pdf

Sandia test report: http://prod.sandia.gov/techlib/access-control.cgi/1995/950229.pdf

Where to buy it for $16 bucks through KR Anderson:
http://www.kranderson.com/shop/Henkel-Loctite/549728/HYSOL-EA-9394-50ML-SEMPAK

 

[Arlo1]

Here is the best magnet bonding epoxy out there: Hysol EA9394 177 deg C capability

NASA Use reference (Page 264, marked 252-3 in report) http://www.scribd.com/doc/40109738/NASA-Glenn-Research-Center-2006-Annual-Report

Data sheet: http://www.mtpinc-exporter.com/chemicals/tds/Hysol%20EA%209394.pdf

Sandia test report: http://prod.sandia.gov/techlib/access-control.cgi/1995/950229.pdf

Where to buy it for $16 bucks through KR Anderson:
http://www.kranderson.com/shop/Henkel-Loctite/549728/HYSOL-EA-9394-50ML-SEMPAK

Hey Bigmoose....
How about something for higher temp and submersion in oil???
177 deg C might cut it I never remember what the max oil temp is...
But anyways yesterday I was fixing a customers bike a yamaha yz450 just put a new piston and valves in and it was hard starting I spent a lot of time diagnosing it till I pulled the flywheel to see if maybe the key moved and it was not the key the magnets were spining in the rotor they are all ok but I carfully got the tin cup out with the magents still glued to it and cleaned it all with breakclean for now and I told the customer with caution Im going to see about re-gluing them in....
Thanks Big moose. Hope to meet you one day Luke says you are a very kind dude!

 

[bigmoose]

Arlo, thanks for the comments. Like my Dad always said, "Life's short, try to make it good for everyone you interact with!" Hope to meet all you guys one day too. I sure like how you are all pushing the envelope and experimenting. I feel like I am a fly on the wall of Edison's first lab.

Take a look at the data sheet http://www.mtpinc-exporter.com/chemicals/tds/Hysol%20EA%209394.pdf

This stuff is amazing. It has been tested in JP4 and hydraulic oil and held up fine per the data sheet. The data sheet also shows 600psi shear strength up to 400 degF!!! I think this is the epoxy for you. A hint that makes the bond even stronger (applicable to all epoxies): Be very careful mixing to have as little air entrained as possible. Entrained air makes it weaker, but not at a failure level. I am talking about getting the last 15% out of this material with respect to bond strength. Also note the cure kinetics in the first NASA article and the data sheet. This material likes a higher than room temperature cure after one day at room temperature.

 

[fechter]

Also make sure the surfaces are clean and have no traces of oil. I think when they manufacture the steel parts they are coated with oil and they don't always do a good job of getting all the oil off.

After mixing epoxy, a good way to get the air out is to pull a vacuum on the mixuture. I like to load the mixed epoxy into a large syringe, then push as much air out as possible, block off the tip and pull the plunger back. This expands all the air bubbles and allows them to float to the surface. The syringe makes it handy for applying the epoxy afterwards.

You can get some flavors of the Hysol stuff from McMaster-Carr.

 

[bigmoose]

As a follow on since Fetcher brought up how to de-aerate the mixture. If anyone is using epoxy in a production environment, just buy the THINKY mixer that meets your specifications and that you can afford. These centrifugal, planetary mixers (some with vacuum capability) are just the finest for preparing epoxies and epoxy compounds. It doesn't get any better!
http://www.thinkyusa.com/products/