Air Cooling my x5

[Doctorbass]

Doc,

My pic is of the left side cover, not the right side. The way you're explaining can't work. Let me help the centrifugal fan thing click for you, because the angle of the holes only help the flow occur, they don't create the flow. Trying to scoop air into the motor at the perimeter and exhaust near the axle would work poorly if at all, because you would be trying to fight against the centrifugal force of the spinning air inside.

Think about a ball on the end of a string, and swing the ball in a big circle using the string. The ball pulls outward due to centrifugal force, because the ball has mass. Air has mass too, not a lot, but it has mass. When we spin that air in a circle inside our hubmotors it tries to fly to the perimeter due to centrifugal force. That's what creates the flow in all centrifugal fans. That's why intake holes need to be closer to the center where the interior air pressure is lowest, and exhaust holes at the perimeter.

The shape of the holes as presented to the outside world are that way so the wind blowing across the surface of the motor on the outside doesn't fight against the centrifugal flow created by the inside of the motor acting like a centrifugal fan. Air doesn't enter the intake holes because we scooped it in. It flows in because the air pressure is lowest at the center, since the air inside starts spinning and is flung to the perimeter and out. Fresh air naturally flows into the intake.

I hope I explained it well, because it's critical to getting the most out of our hub motor ventilation without going to some kind of blower.

John

John, I perfectly understand what you explain. That now make sense to me, but when i look at the picture you shown, i just miss interpreted the direction...

OOpss.. I realised that i just did the holes in teh wrong direction on my 5303 last night...

Now.. time to evaluate if i can swap the side cover of an X5 motor....with some... "minor" mod to the freewheel and disk brake side....... probably i'll need to order a new freewheel side cover... DOHHH !!!

[John in CR]

It's probably easier to just take a file or the drill and angle the holes the other way. It won't enlarge them too much.

Think it through first though to make sure it's not already correct. With outside air going the opposite of the direction of spin, you want it to slip by the exhaust holes, and the air will easily flow into the intake holes near the axle.

[El_Steak]

It took a while, but I finally got a chance to work on my covers in preparation for the final test run "E" (see this post)

For this run, I am adding "fan blades" inside my covers. After a lot of careful measurements, I came up with the following design. As far as size is concerned, its on the safe side, but with the phase and hall wires in there you have to be careful with your clearances.



And here are the 24 blades I made out of 1/8" thick aluminum. I made a little jig to make them all identical.



I then epoxied each blade with JB-Weld to the side of the "ribs" inside the cover:



Tomorrow when the epoxy has set, I will glue this metal disk over the blades. It will serve 2 purposes:

1- Secure all the blades together and prevent a single blade from coming loose and trashing the windings
2- "Channel" the fresh air directly to the windings instead of the stator.

[Doctorbass]

It took a while, but I finally got a chance to work on my covers in preparation for the final test run "E" (see this post)

For this run, I am adding "fan blades" inside my covers. After a lot of careful measurements, I came up with the following design. As far as size is concerned, its on the safe side, but with the phase and hall wires in there you have to be careful with your clearances.



And here are the 24 blades I made out of 1/8" thick aluminum. I made a little jig to make them all identical.



I then epoxied each blade with JB-Weld to the side of the "ribs" inside the cover:



Tomorrow when the epoxy has set, I will glue this metal disk over the blades. It will serve 2 purposes:

1- Secure all the blades together and prevent a single blade from coming loose and trashing the windings
2- "Channel" the fresh air directly to the windings instead of the stator.

it's like a vaccum cleaner blades!!

congrat for that great job!!

[El_Steak]

Forgot to mention that a metal strip will also be epoxied around the inside hole as to prevent air from being sucked from there. This will ensure that only fresh air from the outside is sucked by the fan.

With the 12 interferring "ribs" on each cover, its a biatch to make that metal strip though.

[jag]

Nice work El_Steak! I'm curious if your internal vane fan will finally give you some decrease in motor temps. It is a challenge to get much effect at the low rotation speeds of a hub motor though. Your design puts about as big vanes as can fit inside the motor so that should be promising. Regarding the disc that isolates the vanes from the stator: I'm not sure if it will help or not. Adding it makes your design resemble a squirrel cage fan. However (as you mention) it removes the airflow from the stator. Airflow around the stator may actually also help. Every surface area inside the motor that is subjected to airflow will help shed heat. For effective cooling two things have to be considered: 1/ Moving air in and out of the hub motor. 2/ Making the air inside contact a lot of hot surface area.

[El_Steak]

Those are valid points, ideally I would conduct separate test runs with just the "blades", with the blades and metal plate and with the blades, metal plate and inside ring. But at this point I'm running out of time and the backlog of stuff on the family todo list is getting way to long.

So I'm betting on the full "squirrel cage fan" approach. The source of heat in the motor is the winding and I want to throw as much fresh air as I can on it.

I do hope it will work but even if it doesn't it was a fun project to do. Anything that makes you less of an idiot is a good thing in my book

[docnjoj]

Wow! That is fine metal work! You could always go to a 12" wheel to make it spin faster! Good luck with the project and if it works maybe the factory could take note. They could cast this in fairly easily in the mould.
otherDoc

[John in CR]

Nice work el steak. Thanks for providing the measurements. I too have worried about a blade coming loose, so I may copy your disk idea. I wouldn't extend the disk so far though out of fear that much of the flow will bypass the stator. We want to cool the entire stator including the windings, not just the windings.

Also, I think leave the center open. The air would have to defy the centrifugal force to flow the way you are concerned about. The air will spin some on the flat side of the disk too, so you'd block what would stimulate more air intake. If I do go with a similar disk I will even go the extent of giving it some blades too.

[El_Steak]

Alright, for the sake of science, I'll do 2 more runs, 1 with just the blades and the second with the disk added, we'll see if it makes any difference.

I have the day off tomorrow. A REAL day off with no work, kids or wife. This doesn't happen often these days ! So instead of browsing p0rn on the internet all day , I'll do something more productive and run some tests

I just put the covers back on the wheel and nothing is rubbing !

I then ran the following smoke test to see if there was any "sucking" action from the centrifugal fan in the cover. See for yourselves:



When I say 20, 30, 40, 50, 60, 70, I'm reading the speed of the wheel in km/h. This is for a 24 inch wheel.

If you have a 20 inch wheel then the speeds would be 17, 25, 33, 42, 50, 58
If you have a 26 inch wheel then the speeds would be 22, 33, 43, 54, 65, 76

[texaspyro]

Your bike sucks...

On the subject of the Tao of Epoxy... proper bonding of surfaces with epoxy requires scrupulous attention to details. Epoxy hates smooth. Epoxy hates slippery. Epoxy hates oils. Epoxy hates sloppy. Epoxy hates anything on the surface of the pieces to be joined.

Epoxy hates YOU! Use it enough without hand condoms and you WILL become allergic/sensitized. Wear non-latex, nitrile gloves... nasty epoxy cooties pass through latex like it was not there.

First remove any paint and surface contamination. Then clean the surfaces to be joined with pure alcohol. Remove EVERY trace of surface contamination, and then some.

Then rough up the surfaces in all possible directions and a few impossible ones with coarse grit sandpaper/grinder wheel/etc. The idea is to provide a microscopic surface texture for the epoxy to "key" into. You clean with alcohol first so that you do not sand oils, etc into the metal surface. Clean everything again with alcohol. Do not touch the surfaces with your greasy, grimy, grubby, oil oozing, festering skin!

Mix your epoxy by weight in the proper ratio (don't assume it is the same as the volume ratio). Use an accurate scale. Don't mix by eyeball/volume. Don't weigh into separate containers and then combine them (you always leave something behind and your mix ratio will be wrong). Mix your epoxy well, then mix some more. Frequently scrape the side. Scrape your mixing stick. Mix some more. Scrape well. Mix, scrape, mix, scrape, mix.

Beware of your epoxy "pot life". Don't mix in a insulted/foam container. Be careful mixing large quantities. It will go into thermal runway and set up faster than you can say WTF!

[John in CR]

Once you finalize your ventilation mods you might want to try messing with the current limits, since Justin ran some tests on the 9c and came up a stator saturation at 90A. You're probably exceeding that quite a lot with your phase amps. Dialing down your battery and phase currents a little may not impact performance much, but could reduce heat generation by quite a bit.

How were you planning to attach the disk?...a few bolts running to the outside of the cover?

John

[El_Steak]

How were you planning to attach the disk?...a few bolts running to the outside of the cover?

The disk is thin sheet metal and I'll epoxy it to the blades. (everything is clean and properly roughed up, thanks texaspyro).

I did some tests with aluminum pieces and JB Weld and this stuff is really strong. For the disk to come loose inside the motor, all of the 12 epoxy bonded blades would have to break off.

[texaspyro]

Another little recognized aspect of epoxy is its tendency to have a memory of the highest temperature it has seen in the past. Above that temp, it begins to soften. Above its "glass transition temperature" it begins to fail. Proper use of epoxy in high temperature applications involves thermal conditioning and annealing. And you want to avoid rapid thermal shock when conditioning the epoxy.

I have epoxied carbon fiber fins to the surface of linen-phenolic tubing and had them survive a 0 to mach 2.9 acceleration in under 1 second followed by a few minutes of contact with a 500+ degree aluminum tube. Surface prep and temperature conditioning makes all the difference in the world...

[El_Steak]

Run E is done and numbers are logged !

This latest run is with the fan blades added, but not the metal plate.

Outside temperature was pretty much identical to my last run... so were the results

Go see the results for yourself in my test post:
viewtopic.php?f=2&t=9791&p=278852#p278852

[John in CR]

I didn't look at your wh/km before, and you seem to be running consistently higher consumption than my heavier bike and load at higher speeds. This tells me you must be up into stator saturation levels with the phase currents. Justin came up with 90A as the saturation point for the 9c stator, so dialing your current limits down a bit may not impact performance a lot while decreasing heat generation significantly.

The blades don't seem to stimulate more flow, which is a surprise. The original vanes must be enough to get the air spinning, or the tapered exit vents are the dominant factor of the centrifugal fan effect, not the intake holes and interior blades. I know it's a disappointment that the extra work didn't net any gain, but for other modders it's good news because the major benefit can be obtained with the relatively simple effort of properly placed and shaped holes.

The stator is still warming up in the first segment, and the 2nd and 3rd segment reflect 10-20°C cooler, but why are you running so hot on the slow speed flat trail? You must be doing a lot of accelerations during that segment.

If dialing back the current a little doesn't help out a lot with the trail riding temps, then you trail riding guys may need to active like Arlo1 uses, because the rpms are too low to do enough good. On my bigger motor, I've got a lot more air inside so the blades may help more, plus it's a 19.5" OD tire with no trail riding at all, so I'm getting higher and more sustained rpms. I should get a temp sensor. I haven't worried about it since the motor never runs hot at all. Once I do I'll do some runs similar to yours with the intake holes taped over and without.

[El_Steak]

The stator is still warming up in the first segment, and the 2nd and 3rd segment reflect 10-20°C cooler, but why are you running so hot on the slow speed flat trail? You must be doing a lot of accelerations during that segment.

Yes, last segment is low speed in the first part (narrow bike path) then acceleration - brake - acceleration - brake - etc.

[John in CR]

The stator is still warming up in the first segment, and the 2nd and 3rd segment reflect 10-20°C cooler, but why are you running so hot on the slow speed flat trail? You must be doing a lot of accelerations during that segment.

Yes, last segment is low speed in the first part (narrow bike path) then acceleration - brake - acceleration - brake - etc.

I've been messing with fans for cooling controllers, and small centrifugal blowers is what you need instead of the regular axial fans if you go the active route.

[itchynackers]

I hate to resurrect this thread, but it seems to me that the flow characteristics indicated in some of these videos (stationary spinning wheel) aren't representative of what is going on under real world conditions (the bike/wind is moving laterally). My intuition is that larger holes drilled in both covers near the outer rim may have better cooling effects. I'm thinking of doing this, but even if I do, I don't have a thermal probe to collect data with. Hopefully some brave soul will test this out for us. Any thoughts? Am I thinking about this all wrong? What do the fluid dynamics guys have to say?

Edit: This video may give an idea of what I'm getting at... http://www.youtube.com/watch?v=5Idmf55L8FM

[icecube57]

Unless you are doing 5kw or more continuous I wouldnt bother doing air cooling. Im running my 5303 at 45-50A and 66v and the amount of heat created in the windings is retarded but i never felt the need to air cool it. Being that i ressurected this motor from being frozen and rusted out they are bullet proof for most insane riders. Even the Stealth Bomber and Figher arent air cooled and they push the motors to the limit.

[vanilla ice]

But cooler = more efficient right?

[itchynackers]

I plan to monitor the motor case heat via the hand method for a bit. If it feels too hot (whatever that is) then I'll air cool. Just trying to decide on the best configuration for the holes. I'm not convinced the previous discussions/placements are optimal since the placement seems to be based on non-representative air movement over the outside of the motor.

[El_Steak]

I compiled some test data with different hole patterns here:

viewtopic.php?f=2&t=9791&start=405#p278852

I also did an additional run (after RUN "E") for which I double the size (surface area) of all the holes around the perimeter. The results where pretty close to RUN E (no improvement).

If you have other ideas for hole patterns, try it out and share with us, but if you want to end up with a meaningfull data set, you definately need to install a temp probe inside.

Like Icecube, I stopped worrying about it. On my commute, the motor will reach 130-135C but will not climb higher, I guess it reaches some sort of equilibrium at that point. I have 3000km on it and it still runs great. Should

[itchynackers]

Yes, your data seemed to be the most meaningful, as it used real conditions, not smoke on a bench conditions. Also, your hole patterns were symmetrical (cover to cover). I just had a hard time believing that asymmetrical patterns (holes near axle on right, holes near perimeter on left) had any meaningful purpose.

You seemed to have fixed the cooling. Do you think there is anything that can be done to improve exposure to the elements (sand/water entering holes)? Based on your data, would placing bits of screen over the holes be a good comprimise between cooling/exposure?

[El_Steak]

I opened up the motor after about 1500km with the holes in the cover and didn't see any dirt or rust inside. I don't go offroad with my bike, but I rode through a few nasty storms.

Having all those holes in the covers can let some water in the motor but it also offers good ventilation for it to quickly evaporate (especially when the inside temps reach 130C !)

With the type of riding I do (commute on the road), I don't feel that its necessary to install any screen or mesh on my covers.

Those 9c motors are so cheap that I consider them semi-expendable. If I end up killing it after a season or two, I'll just buy a new one