BEST vent hole pattern for Hub motors

[Garrick_s]

I am ready to Vent my rear hub.
I did some searching on this topic but got SO lost in all the threads.
So I'm hoping to pin point and save time by asking the pros....

Drilling a rear hub motor for ventilation - Has a hole pattern/size been determined to be the BEST ?!?!?!?

G

 

[Farfle]

I would stick with johninCR's mods. IIRC his current version is a set of holes as colse to the axle as possible on one side, and then a set of holes on the very perimeter on the other side.

 

[dogman dan]

That makes sense, Suck it in one side, whirl it out the other.

I took a simpler approach, since I believe the vent holes only do much good after you stop. Then they prevent a heat spike when you stop to cool off. My feeling was that if you can't see the winding through the holes, it's not doing much. Once you stop, air will pull in the bottom holes and vent out the top ones.

While riding, air will tend to blow into the foward facing holes imo. We are talking 40 mph or more right? So air blowing into the cover right onto the winding will be best then. A scoop that catches the air and directs the air into the holes might be a good idea. Just some simple bent tin should work.

 

[cal3thousand]

How about one of these?

http://tornadoair.com/

 

[itchynackers]

That makes sense, Suck it in one side, whirl it out the other.

I took a simpler approach, since I believe the vent holes only do much good after you stop. Then they prevent a heat spike when you stop to cool off. My feeling was that if you can't see the winding through the holes, it's not doing much. Once you stop, air will pull in the bottom holes and vent out the top ones.

While riding, air will tend to blow into the foward facing holes imo. We are talking 40 mph or more right? So air blowing into the cover right onto the winding will be best then. A scoop that catches the air and directs the air into the holes might be a good idea. Just some simple bent tin should work.

FYI, even with vent holes, there is still a temp spike. I did a lot of testing (with a temp gauge, not my hand) prior to the Pikes Peak race. For the 9c 8x8 I was testing at the time, when you stop, the temp spikes for about 2 minutes, then starts to drop (whether there are holes or not).

I'm sure John will chime in and disagree, but also I believe that more air will enter either the top holes or bottom holes (not forward facing holes) as the top holes are moving much faster relative to the wind, than the bottom holes. We all know what happens when fast wind moves over a surface (similar to an airplane wing).

 

[spinningmagnets]

If you don't have a preference as to which side of the hub you have the "near axle" holes and the "near rim" holes, then...I recommend you put the near-axle holes on the left side of the hub. This gives you the option of making a vaned impellor (similar in shape to a cars front brake disk) that can be attached to the hubs disc brake 6-hole flange.

It sounds counter-intuitive, but air will come in the rim-holes, and then flow through the interior and pass out the axle-holes. The vaned impellor will create outward flow through itself, and a low-pressure inside the hub. The inlet holes must be placed near the rim so the incoming air will flow over the stator.

I don't have a hub to prove this on, but...If someone will send me one, I will drill the holes, make the vaned impellor I have described, and perform a smoke-test to show the flow (or lack there-of). If my impellor idea is a total flop, the hub will still have holes near the axle on one side, and holes near the rim on the other. (and, besides, why put the near-axle holes behind a the gear cluster?)

 

[dogman dan]

Sounds like it would work, that is sort of what I was saying about a scoop of some kind on one side. Create a small higher pressure area on one side of the hub at the holes, and air would flow out the low pressure side. Or acutally have a blower doing the same thing, shooting air at the holes.

Yes, a hot stator winding is still going to do a temp spike, but it will be a much shorter duration with some vent holes. That's part of why I say you should see windings thorugh the holes. Sure, vent holes do help, but I just think that when you take a 500w motor, and run 6000w through it, the amount of difference vents make is a joke. But if you are running 2000-2500w, then the effect might make a significant difference.

Bottom line, take a motor that can handle about 2000w at best, and run 4000-6000w into it, and you've made a space heater. So now some holes is going to stop the excess 4000w from becoming heat? Of course not. Efficiency that much overwatted is still going to suck. To run 6kw, you need a motor designed for more than 500w.

But if you are running only 3000w, vent holes may well be able to get rid of the 1000w of heat you are making.

 

[spinningmagnets]

I am a big believer in air-cooling for hubs (even though I haven't done one myself, yet). Part of the reason is that the higher the temperature you are getting, the faster the BTUs will transfer to ambient air (unless, as pointed out, the hub is sealed from the outside, and the generated heat is held in). Air is a poor conductor of heat, but if you flow enough air past something, it will transfer some of it. The bigger the temp difference between the stator and the ambient air flowing over it, the more BTUs will transfer in the same amount of time.

By this I mean that, when the stator is just a little warm, the air-flow will have very little effect. When the stator is hot, the "with fan" and "without fan" temp differences will be much more significant.

A side note to this is that...if you are running near the rated wattage on a hub, and you can put your hand on it and keep it there (the hub only gets "warm" under your hardest usage)...you don't need any kind of vent holes or fan. But, as others have stated, car alternators have been run with a water-proofed/rust-proofed interior and an open case, in conditions where it can get splashed on a rainy day...with no ill effects, so...it also "couldn't hurt" to have coated interior with case-holes.

 

[liveforphysics]

Somebody needs to take some various covers, drill them with various different patterns, like JohninCR's inlet-outlet style etc, and then spin the motor up on a stand with smoke bombs billowing out some smoke around them so we can see what actually happens.

Another handy test would be putting a hubmotor with a temp sensor embedded in the windings in an oven for an hour at 120degC or whatever (to fully temp stabilize inside and outside),then set it on a stand and spin it at full speed and measure the time it takes to drop the temp to 80degC (or whatever). Then try that test with various differently drilled side plates so we can see which sheds heat fastest. Perhaps setting a fan in front to simulate a 30mph wind or something while it's spinning would make the test more useful.

 

[amberwolf]

What I would like to do is install airflow and temperature sensors all through a hubmotor, inside and out, with wireless transponder to collect and send the data from the non-stator-mounted sensors, (stator mounted ones can be wired thru the axle), then go out and ride it in realworld conditions, logging this data, then just stick the raw data up here for ES members to go thru.

I dont' have any of the necessary sensors or loggers, etc., though.

It could be done with unvented covers, then with covers drilled big and open, and then with near-axle holes on one side and near-rim holes on the other, both unangled and then angled each possible way, and in varying sizes. I don't have enough covers for all that, but I do have one 9C-type with the newer covers that is unvented, one old bell cover that is unvented and one that's drilled on outer edge with no angles, and one pair of bell covers with just big holes on both sides.


If anyone has the necessary items to do the monitoring/logging, and/or the old bell-style 9C covers, I am willing to do the tests on CrazyBike2 with it's 4KW-burst setup on the unmodified 9C 2807 stator, for real-world test riding on my work commutes, street/traffic riding, etc.

I can also test with the newer style 9C/clone covers using the same type of stator from Ohzee, which has only been modified to replace halls and coat the windings with red insulation paint, AFAIK, though I have not yet tested that motor/wheel on the bike yet.

 

[Jeremy Harris]

Doing some tests with smoke/vegetable oil mist sounds like a good idea.

The peripheral velocity of holes around even a big hub motor isn't really enough to get any appreciable "pull" from the centre towards the edge of a hub, so if it were me then I think I'd look at just trying to get cross flow from ram air, with a big scoop on one side (facing forwards) aimed at a set of holes in the hub somewhere close to the centre, and maybe a suction scoop (fact aft) on the opposite side of the wheel, drawing from holes near the periphery of the hub. My guess is that this is going to get the best flow through the motor when the bike is moving, but you'll still get the heat-soak spike when turned off. Adding a fan to one of the scoops, with an overrun timer or thermal switch (similar to that fitted to a lot of car electric fans), would probably deal with the heat soak problem.

Be interesting to see the results of any airflow testing.

 

[amberwolf]

Would we really get any appreciable airflow forced into a motor thru a ram-air duct, at bicycle speeds, without a ramscoop opening so large as to be impractical?

I've never even tried to comprehend how to run those numbers, but intuitively it just doesnt' seem like it could move enough air to be worth doing, if it moved any at all.

Then again, I also figured at hubmotor RPMs that centrifugal venting wouldn't help much, but some poeple seem to have shown that it does (and it appears to move air thru my own motor, based on dust-settling patterns, whereas with just big holes I didn't even really get dust in there, implying not much air was moving thru either)

 

[Sancho's Horse]

I may have already posted this, or just had some sort of dream that I did, but if anyone does the smoke test: use a large box that you can fill with smoke from the bottom, and a small vent at the top that you can cover with tape. This will give you greater smoke density, and allow you to "place" the smoke at different sites, so you can get an idea just what is happening. If you videotape it, you will probably need some special lighting effects to be worthwhile. Perhaps someone knows a roadee for AC/DC?

 

[dogman dan]

The idea for the ram duct I was thinking of was also thinking in terms of speeds well above 40 mph. Then a small scoop will push a lot of air. In a drag race, mabye just blow it with AC freon.

Parked between heats at a kart track race, mabye make a dry ice container with a blower on it.

My opinion about holes being useless at some point if you are pushing too much wattage came from the flaming dick video. Where the too much line actually is will be argued forever here.

 

[megacycle]

Anyone used smoke spray? http://mobile.maplin.co.uk/module.aspx?moduleno=20746

 

[John in CR]

Somebody needs to take some various covers, drill them with various different patterns, like JohninCR's inlet-outlet style etc, and then spin the motor up on a stand with smoke bombs billowing out some smoke around them so we can see what actually happens.

Another handy test would be putting a hubmotor with a temp sensor embedded in the windings in an oven for an hour at 120degC or whatever (to fully temp stabilize inside and outside),then set it on a stand and spin it at full speed and measure the time it takes to drop the temp to 80degC (or whatever). Then try that test with various differently drilled side plates so we can see which sheds heat fastest. Perhaps setting a fan in front to simulate a 30mph wind or something while it's spinning would make the test more useful.

I still say the best test is putting a fat guy on an ebike and climb a steep curvy mountain road riding hard enough into every curve that you need to use regen braking that nets you 6% regen going up the mountain, coming to a full stop before accelerating up the steepest section (30% grade), and having a nice cool motor at the top. Next time I do it I'll bring my digital thermometer and heat the motor up first with some repetitive full stop launches before the climb, check the temperature of the stator at the bottom and top of the mountain to prove it's cooler at the top.

As far as I'm concerned I'm done with the topic. I've proved how well it works and shared the hows and whys of my approach here http://endless-sphere.com/forums/viewtopic.php?f=2&t=39773. It's not just holes and their shape and placement either. The blades I put inside the covers plays a significant role in my results.

The bottom line is that no one runs the kind of power I do through hubmotors pushing heavier loads up steep hills, yet none of my ventilated motors has ever had any heat problems whatsoever. They aren't magic, so where does the heat go?....Right out the exhaust holes, that's where, since the covers barely get above ambient so they're not conducting the heat away. I've yet to ventilate any of my newer high efficiency hubmotors, but once I ventilate one I'll bring it across town to my friends shop to do a demonstration video for my customers. He's got a dense smoke machine used in the car repair business. Once I get Hubmonster back on a bike I want to test it too, because I have a question about what happens at it's high end rpm operation and whether I need to increase the size of the intake area.

John

 

[John in CR]

...The peripheral velocity of holes around even a big hub motor isn't really enough to get any appreciable "pull" from the centre towards the edge of a hub, so if it were me then I think I'd look at just trying to get cross flow from ram air, with a big scoop on one side (facing forwards) aimed at a set of holes in the hub somewhere close to the centre, and maybe a suction scoop (fact aft) on the opposite side of the wheel, drawing from holes near the periphery of the hub. My guess is that this is going to get the best flow through the motor when the bike is moving, but you'll still get the heat-soak spike when turned off. Adding a fan to one of the scoops, with an overrun timer or thermal switch (similar to that fitted to a lot of car electric fans), would probably deal with the heat soak problem....

Jeremy,

The air inside a hubmotor is spinning whether it's ventilated or still sealed. The exhaust holes at the perimeter don't need to draw the air out, though building a reverse curve blade structure outside them would certainly increase flow. Instead, as long as you have intake area closer to the center like any centrifugal fan, the perimeter holes will naturally allow the air to flow out that is forced there by centrifugal force. Because our motors are moving through the environment with no housing that a centrifugal blower typically has, the shape of those exhaust holes becomes important because at the top part of the rotation they are moving fast in the direction of motion and the outside influence can easily stuff up the flow. DoctorBass proved this when he accidentally cut his holes angled in the wrong direction and ended up with a hot motor as a result.

FWIW, my vented motors have no heat soak spike problem. They're no warmer 1, 2 or 5 minutes after stopping than they are immediately after I stop.

John

 

[megacycle]

Sounds like it can too easily be over simplified, air pressures, turbulence, vortex, velocity yada a yada
Nasa might be up to it, Im going to go to explore your link John, empirical sounds the way for me thanks.

 

[John in CR]

Sounds like it can too easily be over simplified, air pressures, turbulence, vortex, velocity yada a yada
Nasa might be up to it, Im going to go to explore your link John, empirical sounds the way for me thanks.

If you're planning to vent your 5404 and want to flow air through the gap, be sure to take into account those channels between the teeth of the angled stator. Since they are angled, there's at least some cross flow created even with a sealed motor, and you don't want to fight that natural flow direction. You want to compliment it by enhancing flow in that direction.

Also, if you use a temperature sensor, I disagree with the common method of burying it at the windings. The factory of my high efficiency motors puts 95°C thermistors used to trigger a temperature warning light on their scooters epoxied squarely on the stator steel some distance from the windings. That location seems like a much better gauge of the motor temperature as a whole, because the windings temperature can change so much in a matter of seconds, and if the stator iron isn't hot yet it will rapidly sink the heat away from the windings. eg While 130°C is a pretty safe temp for the magnet wire's dielectric, the back iron could be 60-70°C and the motor is perfectly safe, but it could also be over 100°C and the motor likely to suffer failure with the next hill or hard takeoff. Burnt windings mostly happen from pushing a motor too far into saturation, which should be avoided anyway, but the magnets, their epoxy, hall sensor wire insulation, etc. often have a lower temp limit.

I hope you come up with a new twist I can copy.

John

 

[megacycle]

It sounds from what your saying its better to monitor the heat sink in iron laden motors like 5404 towards monitoring the rise in temp well before it overshoots into the red.
I expect the 4060 Ive got then being stamped, to have a much shorter lag/lead profile then.

Thanks for the advice regarding the 5404s internals as i Havnt had the lid off yet.
Though it confuses the heck out of me regarding air flow in a sealed motor :?

Just a note industrial motors usually have a sensor in the windings,in the stator and on the bearings too.

 

[megacycle]

I had a thought, more like a brain strain these days.
Apologies if someone has suggested it before.
If there was an inlet scoop and outlet exhaust mounted on the seat stays, directed at the hub cover holes as they pass.
Could either just scoop or force the air in, have a small compressor whose output was temperature controlled.
Plausible or raving?

 

[John in CR]

It sounds from what your saying its better to monitor the heat sink in iron laden motors like 5404 towards monitoring the rise in temp well before it overshoots into the red.
I expect the 4060 Ive got then being stamped, to have a much shorter lag/lead profile then.

I'm talking about the stack of laminations, not the spoke structure in whatever form that centers the stator on the axle, and the temp of that stator iron sure seems like a better indicator of overall motor temperature especially since the factory that makes arguably the best hubmotors in the world installs theirs there, and not in the windings.

Thanks for the advice regarding the 5404s internals as i Havnt had the lid off yet.
Though it confuses the heck out of me regarding air flow in a sealed motor :?

The spinning rotor ring and side covers cause the air inside to spin. On the X4's and X5's the laminations are played, so the resulting channels between the stator teeth are at an angle, not perpendicular to the uneven magnet surface and moving air. When the moving air in the magnetic gap encounters these channels, some of the moving air will follow the channel. This creates a higher pressure on one side of the motor than the other, and no matter how small the pressure difference, air will naturally flow from the higher pressure side to the lower through the spaces in the stator spokes. I'm sure it's not much flow, but you always want to enhance natural flow, not fight against it.

Just a note industrial motors usually have a sensor in the windings,in the stator and on the bearings too.

Our motors are used very differently. We don't really need them on the bearings, because we can typically tell when a bearing is going bad. Ideally we should also run 2, one buried in the windings and one on the stator iron. If given a choice of only one, I believe on the iron is better. Like in my example before, windings at 130° doesn't tell you much unless you've been running at that temp fairly steadily or you know the stator iron's temp too. otoh the stator iron temp alone gives you pretty good info on the overall temp of the motor, and if it's approaching 100°C you're in dangerous territory. If you have effective ventilation and that stator iron is getting that hot, then you need a different motor for those current settings, because you're obviously pushing it well into saturation.

John

 

[Ypedal]

2 pages of discussion.. no pictures.. damn..

 

[nicobie]

Big holes pass more gas

 

[megacycle]

It sounds from what your saying its better to monitor the heat sink in iron laden motors like 5404 towards monitoring the rise in temp well before it overshoots into the red.
I expect the 4060 Ive got then being stamped, to have a much shorter lag/lead profile then.

I'm talking about the stack of laminations, not the spoke structure in whatever form that centers the stator on the axle, and the temp of that stator iron sure seems like a better indicator of overall motor temperature especially since the factory that makes arguably the best hubmotors in the world installs theirs there, and not in the windings.

I meant the laminations temp lead/lag.

Thanks for the advice regarding the 5404s internals as i Havnt had the lid off yet.
Though it confuses the heck out of me regarding air flow in a sealed motor :?

The spinning rotor ring and side covers cause the air inside to spin. On the X4's and X5's the laminations are played, so the resulting channels between the stator teeth are at an angle, not perpendicular to the uneven magnet surface and moving air. When the moving air in the magnetic gap encounters these channels, some of the moving air will follow the channel. This creates a higher pressure on one side of the motor than the other, and no matter how small the pressure difference, air will naturally flow from the higher pressure side to the lower through the spaces in the stator spokes. I'm sure it's not much flow, but you always want to enhance natural flow, not fight against it.

I think that's just a lot of hot air and no pun intended.

Just a note industrial motors usually have a sensor in the windings,in the stator and on the bearings too.

Our motors are used very differently. We don't really need them on the bearings, because we can typically tell when a bearing is going bad. Ideally we should also run 2, one buried in the windings and one on the stator iron. If given a choice of only one, I believe on the iron is better. Like in my example before, windings at 130° doesn't tell you much unless you've been running at that temp fairly steadily or you know the stator iron's temp too. otoh the stator iron temp alone gives you pretty good info on the overall temp of the motor, and if it's approaching 100°C you're in dangerous territory. If you have effective ventilation and that stator iron is getting that hot, then you need a different motor for those current settings, because you're obviously pushing it well into saturation.
John

I think it would be overkill too in a few hundred dollar motor.
But i think it still would indicate a faulty bearing before you felt it