Ventillation fans in a vacuum - fans are failing - windings fault to earth 4

[ryanmech]

Hi

Short question: What type of winding insulation should be used for a fan motor that is within a vacuum pressure?
also posted in the HVAC forum

http://www.eng-tips.com/viewthread.cfm?qid=394983&action=updated

Details:
We have an application where we have ventilation fans in a chamber under a vacuum pressure.
What is the point of a fan in an environment where there is no air you may ask? well there is air it is just 5% of what it would be at atmospheric pressure.
Yes I know that fan motors are able to be designed with their windings quite dense, due to the high cooling efficiency of the fan air running over the motor...And we are running les than 5% of that air over the fan. (FYI - the fan body/hub temp has not ben measured to be over 30deg C during any of the tests below)

The machine is a food vacuum drier.
The environment changes during the process but generally: is initially ~50% humidity and ending at very low humidity. The temp ranges between 5 and 50 deg C. The pressure ranges between atmospheric and -95kPag (or 101kPa to 5kPa absolute)

the fans we are using have squirell cage motors. we are using a VSD and a sinusoidal filter to provide variable speed to the fan depending on the process steps.

There are four fans - three have failed due to initially earth faults (recorded by the VSD) and then eventually they fail by windings short circuit to earth. the earth connection cannot be seen as it is within the innermost of the windings, and the construction is such that the resin has to be baked off to remove the windings and as such the evidence is not available to be sighted.

we are using:
Fans: FC056-VDQ.41.V7
VSD: Allen Bradley PF525 VSD
Filter: Schaffner FN5040 Sine Wave filters, typically they are run at 50Hz

Initially there were no faults in the conditions listed above.

At a period of ~30hrs, the earth ground faults were occurring only when the vacuum was greater than -80Kpag. (the VSD records the earth ground fault due to a measured current difference >25% between any two of the phases/windings).
When the chamber is returned to atmospheric pressure there were no faults. i.e. the insulation at atmospheric pressure is much greater at atmospheric than at a reduced air pressure. insulation test to earth from each of the windings is >1000MΩ at 500V after returning to atmospheric pressure.
Then the earth faults would get gradually more frequent and at pressures closer to atmospheric pressure, eventually the fault is continuous or instantaneous. the insulation resistance is then measured as very low i.e. short circuit.


we have undertaken several tests to try to isolate the equipment.(same environment as above Unless noted otherwise)
1. running without the VSD and filter --> direct online with the fans = failed windings short circuit to earth --> proves the VSD and Filter are not the direct cause.
2. running at atmospheric pressure continuous for 48 hrs --> =no faults -->proves that no degradation of the winding insulation occurs at atmospheric pressure, therefore VSD and filter are not the cause.

we think that the main factor is that the fan motor insulation subject to vacuum degrades over a period and eventually causes a permanent short circuit.

Our client has fan with the same part number that have been operating in the same conditions for 10 years without issue. To be clear - the specifications for the Fan is that it is not specified for anything other than atmospheric pressure environments. The client knows this, we know this and the supplier have clearly stated this.

We think that there had been a change in the insulation of the fans - our clients 10yr old fan has winding insulation better than the current version of the same fan.

that brings us to the short questions:
What type of winding insulation should be used for a fan motor that is within a vacuum pressure?
should the fan windings be totally encased/potted in resin to ensure the best chance of not degrading in the vacuum environment.

We have read a lot of theory about the degradation of winding insulation in vacuum. Practically we don't have the information about what type of winding insulation we should use.


thanks for getting this far down....

regards
ryan







Ryanmech
Scott Technology Ltd NZ
Machinery automation

 

[stevenal]

Since EEs like to use vacuum for current interruption and use added insulation at altitude, I think we intuitively know there must be a minimum somewhere even if we never heard of Paschen.

 

[MikeMay17]

Ryan,

The minimum spark-breakdown voltage (regardless of pressure) for Copper in Air is 250 Volts; if the voltage is less than this, you should be good to go. Just keep in mind that the voltage at your motor terminals (the voltage that matters in this case) may not be the same as the line voltage in to your motor driver... If I were you, I would put a meter on the motor terminals during operation of the fans and see what kind of spikes you are generating with that VFD and those (assumed) long cables; I have seen spikes in excess of 2x line voltage due to standing wave phenomenon and super-position.

-Mike

 

[mike4472]

Have you thoght about a motor driven by something other than electricity. My first thoughts are an air driven motor. Obviously some thought would need to be given to venting the waste air cos it probably won't sit too well with your vacuum.

 

[itsmoked]

ryanmech; In view of Paschen's law is there any chance the dryer you're having problems with is running a different 'recipe' or pressure profile, different product? If so it could be why the 'old' motors seemed to last forever.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[zappedagain]

Any controller changes? Maybe the controller used to keep the fans off during the pressure change (i.e. so voltage wasn't applied when you passed the minimum in the curve) and someone came up with a better plan to run the fans the whole time? Or a relay is stuck closed so the fan run continuously.

Z

 

[Pythagorus]

Hi I skimmed through the posts. There seems to be a lot of wasted time over insulation and windings when towards the end the engineering design was rightly called into question. I can understand technical interest in the events however the problem needs fixing, fix it. Ok I'll add some thoughts and do the thing to death but the solution is ...fix it. I'm going to try to put this to bed.

Today with almost everything from once good companies, being made in China along with claims of 'meets spec's'I will not buy equipment from there at all however finding genuinely 'made in Sweden, Germany and the like is a mammoth task, so essentially we have become stuck with junk through our own miserable mean-ness (procrastinating euphemism 'competitiveness/level playing field') in buying it on price instead of supporting home industries...it's the lemming rush gripping the world from immorality to foreign policy and living in denial of climate change and sustainability.

In my ruminations there maybe transient spikes in your area which regularly subject them to voltages in excess of rated voltage. There could be lightning strikes on lines doing same thing. That the surface of the windings is dry when you check (you dismantled the fans?...how long did that take each time) doesn't mean there is no residual inside the windings and which may hold acids from the unknown gases and perhaps any 'preservatives' applied to them.

This equipment built by slave labour using 'just gets there' manufacturing spec's may work at some normal, under-stressed situation but doesn't make the grade thereafter. That's an unfortunate result of the change in culture. When one sees equipment from the 20's and 40's still working well having been subjected to overloading and extreme use the answer is obvious. If you tolerate junk you get junk whatever the name on it. With China in the picture the 'good old names' no longer matter.

The Italians by the way I think still locally wind motors.

If I was going to spend money in useless endeavours I'd have the three motors unwound by hand to see where the fault presented itself, looking for a commonality and testing for thickness, purity and types of contamination. I might even have their rep pick them up and let them do it. It's possible that the quality of the copper it self is poor and the problem is coming from inside. It could be that the varnish deteriorates at any change of direction. It could be that the baking process is inadequate. It could be a combination of all. The problem however is that is costing you and the people trying to assist, time and money. Fix the system...isolate the vulnerable equipment from the environment.

Other than that what's the point? They fail. They are not up to the task. Change the system if reasonable and practicable, stop worrying about the fans, take them out of the environment.

The situation itself presented the motors with poor environment and gases of which you have no analysis. TEFC may not help in the chamber as they are still fan cooled...and we are not sure where the problem lies and really, it's already wasted too much time.

Were radiation (fins) or unusual conduction (liquid cooled) the method of cooling in a vacuum that might have a chance but again...easier by far to take it all outside the environment which is more certain BUT if I was going to use the same motors or same type.. I'd be running them close to the vacuum system for some weeks to see whether the problem exists in a more normal environment. Using TEFC's in normal air flow with the extraction changed is better.. or just using an external commercial vacuum pump (or pumps)..get the motor/s out of the picture.

The information given by the 'worried' and in my answer was a bit dragged out, but essentially a vacuum is not going to provide cooling, nor is some uncertain volume of 'gas'. The only cooling will come from radiation and I think there would be hot-spots in the windings.
so put it to bed:
a) external extraction constructing it or using commercial vacuum pumps
b) using ceramic/magnetic motors with wound rotors if determined to keep motors inside
c) try never again to think about the problem...life's too short and it gains you nothing.

'Lump-hammer' Summary:
I'd get on with fixing the system and worry about the motors as a tax dodge.

The suggestion of using external fans is ideal...so long of course as not enveloped by the 'gases'. The principle of the vacuum cleaner can be employed as is done in vacuum pumps.

The suggestion of lowering voltage were fans to stay inside seems theoretically rational ...but not below 5% and really.. would that help?

Personally were I so resistant as to maintain the old system not willing to use my managerial 'nouse' and just fixing he problem I'd be looking for some suitable old 32 Volt motors and running one inside the cabinet just doing nothing and then periodically inspecting it and megger-ing it last thing evening first thing morning and anytime between...recording and ruminating in my lunch times, time -off, meal times, bedtimes and whilst asleep. If it stays good...change to 32 V but again...what a waste of time, health and capital when a more certain, environmentally isolated and a-biotic collectible system is available and practical and not expensive. ...using an external pumping system.

Redesign, fix it using the vacuum cleaner principle (nothing is inside the chamber) forget the dud fans, write them off or claim on insurance if possible and get on with life. Getting bogged down in endless loops is something I have been good-at occasionally and it gets one no where unless you were a 'concerned' scientist or CEO at the motor company but then....I think they'd be told by their CEO..'who gives a toss about failures in poor environments which would be maybe .0000000000000000001% of our sales...forget it...let some other poor mug company worry about it if they are silly enough to care.. we aren't!!'

As we of the Irish mind might say 'd'yer have me there'?...wasting your own time and the time of others on useless endeavours does not fix the problem. Fix the problem by either
i) totally wasted expenditure on scientific evaluation by a consulting engineer and seeing whether he can absolutely guarantee a suitable motor at the end of the cheque writing or
ii) rebuild the system taking into account one sensible response you received but using the vacuum cleaner principle....then sleep at nights.

 

[ryanmech]

FYI an update - we are currently doing a test in the original machine with the new fan. We are half way through a week long test. If the new fans do not survive then it indicates that the fans manufacturer has changed the insulation...if they do survive then the issue is probably a combination of several factors on the new machines setup
-Previously the old fans have been running for 10yrs with no issues.
-The new fans with the same part number (but 10 years newer) are not surviving in the new machine...we will see what happens in the old machine with the new fans.

Ideally if it is just a matter of winding insulation then we can rewind the motor with better insulation. The original question was What type of winding insulation should be used for a fan motor that is within a vacuum pressure?
thanks to Edstainless for his suggestion of Kapton insulation - double layer with Teflon fusable layer in between.
thanks itsmoked for your suggestion about using empiremagnetics
thanks to jraef for his suggestion to use supperioressex inverter spike resistant magnet wire.

thanks btrueblood, itsmoked, jdunfee5, stevenal, mikemay17
yes - we understand Paschen's law to a degree....-->better insulation would prevent voltage breakdown..

Thanks mike4472
we had thought of a fan driven by alternative means, but were hoping that the issue could be resolved by better insulation.

thanks itsmoked and Zappedagain
the tests on the new machine were simply at the worst conditions of the previous machine: 5kPa absolute running continuously - the comparative test of the new fan in the old machine will also compare the operational conditions directly.

Thanks Pythagorus
I am not sure your post provides assistance we are looking for - but I do recognise the time and effort you put into your long response. I fully recognise the justification for your post - we just don't have the space for an external motor.


thanks again and I do appreciate all responses - I recognise the efforts and interest that all responders are taking. And I hope to hear from someone who has successfully wound motors used in a vacuum environment, and the insulation specifications used in that motor....

regards
ryan


Ryanmech
Scott Technology Ltd NZ
Machinery automation

 

[jdunfee5]

ryanmech said, " we just don't have the space for an external motor. "

If you do not come up with an easier solution using the existing motors, or by re-winding, I think a motor outside of the vacuum is the next best choice.

The "external" motor does not necessarily need to be put outside the area of the chamber. It may be viable to add a can-shaped enclosure that has its open side on the wall of the chamber. If you can have air flow around the motor, then you may be able to use a smaller motor than you use now, and permit it to run hotter. Thus, the motor and its enclosure may end up the same size as the original motor.

I am just now assigned the task of researching the suitability of various materials in a vacuum. So, I am now in the same boat as you. Engineering tables on the properties of materials never include information on vacuum suitability. Does anyone know of a good go-to source for this sort of information?

-Joe

 

[MikeMay17]

Ryan,

One variable you are forgetting is the VSD; are they the same in both the "old" and "new" setup? If the "old" setup has an older, slower VSD you will not see the same DC voltage spikes at the motor as you would with the "new" setup. If this is the case, the new motor would operate just fine in the old setup, but will still fail in the new setup due to Partial Discharge insulation damage.

-Mike

 

[Dread]

Have all avenues been exhausted to identify the insulation and winding style of the older fan that seems to have provided good service?
If it aint broke don't reinvent it with exotic insulation.

 

[btrueblood]

MikeMay does have some good suggestions. Can you put an o-scope on the motor leads on the new and old motors, and compare (o-scope will need to be able to capture fast DC transients, which is tricky...hope Skogs is reading).

 

[ryanmech]

Hi Mikemay and btrueblood and dread
We had looked into the old vsd setup. It is an old pdl xtravert drive with no output filter and 4kHz switching frequency. The new vsd with sinusoidal filter was set to have the same 4kHz switching frequency.
The old machine is in a different city and is in a production environment so a little difficult to get access to for a comparative scope reading. If the new fan lasts in the old system we will have a look at the old fan.

Thanks
Ryan

Ryanmech
Scott Technology Ltd NZ
Machinery automation

 

[ChasBean1]

Such a nicely detailed post unfortunately prompts a lot of questions. For me,

1) How are the 30°C temperatures measured while the system is near a perfect vacuum? Measurement methods might be suspect without air.

2) Motors require air to cool them. If there is no air, they will not be cooled.

3) Why would ventilating/circulating fans be running to circulate air that does not exist? Can they shut off below a certain pressure?

4) What is the rough amount of air (cfm, lps, etc.) that they need to circulate? There are several vacuum compressor models designed to move/compressor air, all of which have their motors outside the medium and therefore can be cooled by ambient room air.

Sorry if some of this is repeated; just saw there are many responses above and I did not read them all.

Best,
CB

 

[mikekilroy]

couple ideas to clear up some mis-statements... Radiation is NOT the only way to get heat ou of the motor in a vacuum. In air, it is condvection, conduction, and radiation. In a vacuum it is conduction and radiation. In both cases, radiation only provides about 5% of the total cooling.

When we design motors for vacuum use, we make sure we have the proper amount of metal contact to let the conduction element take the heat away. As one stated, how do you know your 30C is max? I doubt it - I suspect you have hot spots INSIDE the windings, where you admit they are failing, where the temp may be exceeding the rating of the insulation. Something to consider.

When we designed motors for vacuum use in the past, we did NOT change the insulation material - we changed grease in bearings and things that outgassed; once the outgassing is done, it is done I believe.

I would NOT use an induction motor in a vacuum unless it is the only type applicable for some unique reason. You state your motor is just 0.8kw in size. Since you lost the 75% cooling from convection, your left with the 25% conduction, so your motor should be 75% larger than required in air or some such.

You really should be getting rid of that rotor heat that cannot get out thru convection, and change to servo motor (synchronous) motor with magnets to eliminate that internal heat source.

I can kill insulation in a 230v motor w/o inverter duty insulation, running a 460v - even limiting ac fundamental output to 0-230v max - sometimes in less than 5 minutes due to corona discharge

I do not believe you have any business running a 460v (650v bus) motor/drive in a vacuum - and you are doing very well proving it!

It was stated above lower the voltage. You do not like 40a at 40v (many drive and motors available here, but ok), but I would change to 230v system ASAP.


I would contact app engineers at

http://www.kollmorgen.com

and explain your issues to them. Their aerospace division - in VA. USA not China - routinely build motors for use in space (vacuum); they may be the source you have been looking for. I will guess your motor built for this app in your size will be around $ 4000ea. If that is too high, then don't bother.

http://www.KilroyWasHere<dot>com

 

[ryanmech]

Hi Chasbean1
notes to your questions below.
1) How are the 30°C temperatures measured while the system is near a perfect vacuum? Measurement methods might be suspect without air.
--> the temps were measured with a infrared temp sensor pointing at the hub in various locations when under a vacuum. The digital display was recorded by video camera when under vacuum. it was a bit "old Skool" a bit like taking a photo of your screen because you cant get the printer to work. When running under vacuum the temp of the hub was never above 32 deg C in various locations. When the system was stopped = vacuum dumped to atmosphere and fans stopped, the temperature creeped up to 34 deg C (this means that there was convection cooling when running in vacuum). In my mind this conclusively confirms that temperature is not the issue. Additionally when the failed windings were inspected there was never any evidence of temperature related failures. i.e. no discolouration even deep within the windings.

2) Motors require air to cool them. If there is no air, they will not be cooled.
--> umm I don't agree. We use plenty of motors under normal atmospheric conditions which are completely sealed with no air circulation fan at all. in our vacuum application there is air, just not a lot - say ~5%. therefore there is convection cooling, just not a lot. But on the other hand there is 95% less air and that means say ~90% less load on the blades = ~90% less work being done = ~90% less heat being generated.....= ~90% less heat being required to be dissipated.

3) Why would ventilating/circulating fans be running to circulate air that does not exist? Can they shut off below a certain pressure?
-->there is some commercial sensitivity which I cant elaborate on further. however there is 5% air if at a 95% vacuum or what ever the vacuum is at.

4) What is the rough amount of air (cfm, lps, etc.) that they need to circulate? There are several vacuum compressor models designed to move/compressor air, all of which have their motors outside the medium and therefore can be cooled by ambient room air.
-->sorry not the right of application for our case.


regards
ryan

 

[ryanmech]

Hi mikekilroy

its not a overtemp failing of the windings. we have inspected and found areas of where the windings insulation has been failing. there is no discolouration. there is degradation of insulation coatings. but not temp related.

insulation coatings also outgas...

yes our squirrel cage induction motor ventilation fan is the only type at a suitable cost for our application.

The load on the fan is also far less under a vacuum because it is moving far less air. there is little blade pressure = little load on the motor = little work done = little heat = little heat to be dissipated.

We completely agree that the failures are due to corona discharge.

yes USD4k is too much for a fan.


thanks for your suggestions - we may have to return to KollMorgen if we don't get the long term reliability.


best regards
ryan

 

[ryanmech]

Hi All

we had the motors rewound with standard winding wire restored to as new condition. We ran the motors at the same voltage with the same drives under the same environmental conditions. i.e. same, same, same, same...

we made a simple low cost modification which my customer is holding close to their chest with respect to intellectual property and future patents.

Whilst I would like to provide a detailed description of the solution I am prevented from doing so by our customer. I apologise as I would liked to have given back with all the responses and careful thoughts on the posts!

The machine ran for 800hrs without failure and was shipped last week. Hopefully long term testing allows the fans to last ~10 years.

regards
ryan