Cooling fans inside Hub motors

[Cowardlyduck]

Offroader, your not wrong about the air flow rate making a difference. I already did some testing on this remember:
https://endless-sphere.com/forums/viewtopic.php?f=30&t=56965&start=25#p946731

The graph I made based on all my runs clearly shows the difference the air flow makes:

Granted, this is at a much lower flow rate than your EDF, so I would imagine an even greater difference would be seen with your setup.

Cheers

 

[Offroader]

cowardlyduck: where was your thermistor placed for that test? Windings or stator?

I'm not finding the same results of your test. It could be because I tested at much higher temps, 105 C vs your 52 C. Good chance your Stator was at the same temp as your windings, so both had to cool together.

Since you rode your bike for a long time to reach those temps you heated your stator.

My stator was probably much cooler and sucks in the heat from the windings.

I'm testing more peak heat cooling.

 

[Cowardlyduck]

cowardlyduck: where was your thermistor placed for that test? Windings or stator?

I'm not finding the same results of your test. It could be because I tested at much higher temps, 105 C vs your 52 C. Good chance your Stator was at the same temp as your windings, so both had to cool together.

Since you rode your bike for a long time to reach those temps you heated your stator.

My stator was probably much cooler and sucks in the heat from the windings.

I'm testing more peak heat cooling.

The thermister was placed right on the windings, wedged into a recess between the windings and the stator. The metal casing of the sensor was in direct contact with the windings.

I'm not sure about the stator temp compared to the winding temp, but all my motors have steel stator supports so it's less of a factor compared to the aluminium stator supports since the steel cannot absorb heat as quickly or easily by comparison.

To get an idea of the exact testing conditions I suggest looking back at my post as it includes a lengthy video showing exactly what I did for the testing.

In terms of peak vs non peak cooling performance, I found the air flow makes almost no difference while the motor is under heavy load...as shown by the graph above. However almost no real world riding conditions will consist of constant load, so the real benefits of fan cooling comes from removing the heat faster when the load is removed or reduced. That's why my testing includes the time to cool from peak...peak meaning the peak temperature reached.

Cheers

 

[Offroader]

Why did you only heat to 52C? I still consider my motor cold at 52C.

Cooling from 105 C to 70C takes under a minute, with or without fans running.

I've not tested it but cooling from 70C to 35C would take much much longer. Cooling really slows up at around 70C. My motor usually stays around that 70C temp also during operation.

 

[Cowardlyduck]

Others had similar comments when I did the test, which I discussed in the thread, but basically so I could quickly and easily repeat the test with fans at different speeds I kept the hill shorter which meant the temps were lower. Doesn't matter though, the test is scalable and still applicable to higher temps.

Like you Offroader, I also generally run my motor around 80C when riding hard and running the fans. I pretty much never have to stop during normal riding, or even hard off road riding. Where it struggles is on long steep uphill stuff.

Anyway, I'm hoping when I put 12 fans into my Leaf which I just opened it will be both quieter and more able to handle the steep stuff with more air flow.

Cheers

 

[SlowCo]

I'm hoping when I put 12 fans into my Leaf which I just opened it will be both quieter and more able to handle the steep stuff with more air flow.

If you could document your progress of putting the fans in your Leaf motor and post the photo's here that would be much appreciated... :wink:

 

[Cowardlyduck]

If you could document your progress of putting the fans in your Leaf motor and post the photo's here that would be much appreciated... :wink:

I'll do one better and make a video of it.

If your after photo's, there are plenty of my last 3 fan mods back through this thread. Here is the process when I did this Leaf motor for the first time:https://endless-sphere.com/forums/viewtopic.php?f=30&t=56965&start=350#p1111068
This is the same motor I will be adding 12 fans to this time.

Cheers

 

[Cowardlyduck]

Hi guy's,

So I've run into a bit of a problem. I've discovered that when I place 2 fans in a push/pull setup it actually REDUCES the air flow. I have no idea how it works, or why it does it, but it's quite noticeable.

The reason I wanted to move to 12 fans was not for an increase in air flow (although that was a nice side effect) but to absorb the higher voltage I'm now running my bike at.
So, with that in mind, does anyone have any suggestions on how I might still use all 12 fans inside the motor?

I do not want to use a DC-DC or buck converter as I don't like the added complexity. I had thought about using LED's, but can't seem to find any that would operate at the 4-8V I would need.

One idea I has was to position the fans on one side of the motor in a way to spin the air around with the wheel. That way I can bolt 6 fans like I have previously over the stator holes, and on the other side I would have to secure 6 standing up to circulate the air. The hope is that this would help spin the air out the holes faster. The question is, should I spin the air with the rotation of the wheel or in the opposite direction?

I'm open to any other simple idea's.

Cheers

 

[amberwolf]

Usually back-to-back fans need a spacing of some amount, to prevent or minimize interference between the two from eddy currents in the air between them. This spacing is probably different for different speeds, too, the faster they go the more spacing they probably need.

You'd have to determine the spacing experimentally, without an airflow model of the fans to do computer simulations of (and even then probably have to tweak it after finding the best spacing via simulation).


It might also work to change the speed of one of the fans relative to the other by a small amount, but that would add complexity you are trying to avoid, and might make it worse instead--it might require matching the speeds perfectly instead

Alternately, locking the blades together physically would also probably work, so that all the eddys are the same relative to each fan's blades, but if the motors are independently driven internally I don't know what that would do to them. You might also have to "stagger" the blades, at what angle I don't know, so that the tips are not necessarily aligned to each other directly.




A simpler way to absorb the extra voltage would be a bunch of diodes in series; depending on type the voltage drops can be something like half a volt to a volt each, since this is all pretty low-current stuff.

 

[Cowardlyduck]

Thanks AW! Some really good suggestions there.

The spacing thing is a really good point, I hadn't considered that. I do have enough room to potentially space them apart somewhat, so I'll play around and see if I can get that to work.
For air flow measuring I'm just using my face. :lol: It seems crude, but I can sense a change in air flow over my face when it's large enough, which it definitely is in this case.

I'll look into the diode option as it does sound simple enough. The only catch would be how much power they use. I guess if they get hot, I could just wire them into the wiring outside of the motor.

Cheers

 

[amberwolf]

The diodes will only dissipate a lot of heat if the fans use a lot of current. let's say you're using 8v of drop, and 1A of fan current--that's only 8W of heat. Half of a cheap soldering iron's output, or a hair more than one of those little nightlight bulbs.


The airflow thing you can also verify by sound--the fans should sound about the same "unloaded" (nothing in the way in front or behind) as they do in series, if the airflow is the same. When you hear buzzing or harmonic noises, etc., or the RPM of the fans changes (lower) then they're not moving air like the would unloaded.

I used to heavily mod my computers for the best airflow with the fewest and largest fans at the slowest RPM that would keep them cool, so I wouldn't pick up the sound with the recording studio microphones. Often this required moving fans away from components, and always required removing grates and sometimes even support braces, and ducting them with cardboard or paper (absorbs more noise than plastic sheets). Sometimes there just wasnt' enough room to do what I really wanted to without actually putting the fan(s) outside the case itself. Sometimes to slow the fans down I'd just put a potentiometer or set of diodes between it and the power source, depending on how adjustable it had to be.

But one problem with totally smooth airflow is it doesn't seem to pull as much heat out of things as turbulent airflow does, so I had to try to just get smooth airflow at the fan itself where it pulled air out (always pulled instead of pushed, was quieter for whatever reason and worked better), and actually duct things to make it *more* turbulent flwoing over components, to get the best cooling for the least noise.

 

[Cowardlyduck]

Well, this is what I ended up doing:

The fans on the disc side are normally positioned to pull air in, but the fans on the freewheel side are offset and angled to spin the air around with the rotation of the hub. I achieved this by using long bolts, which I bent using some pliers after they passed through the stator. Nuts were threaded on before bending to secure the fans on the opposite side, and nylock nuts were put on last to hold everything down.

Using the old hand next to the holes method, the flow does seem to be stronger. The noise is definitely lower than before which is nice, and one of the changes I was hoping to achieve.
I'm taking it out on a group ride with some other ES members this Friday so will give it a good thrashing and see how it holds up then.

Cheers

 

[Cowardlyduck]

So anecdotally from the Ride yesterday, It seems the cooling performance is actually slightly reduced from before.
This is a little disappointing, but I can live with it since the noise level is now about half what it used to be which is something I definitely wanted.

Even with this 'reduced' performance, my Leaf had cooled to about 40C a few minutes after a serious climb while everyone else's motors (MXUS 3000 mostly) where measuring 70C+ or were too hot to touch still.

I did hit thermal rollback a few times where almost everyone else was able to keep going which sucked, but I was able to limp to the top at reduced power while my fans rejected all that heat. Without the fans it would have to be stop and push to the top, which is what one of our crew experienced on the same hill.

All in all, I'm happy with how the move to 12 fans turned out. Now I have to decide if I do the same to my other motor (HS4080) since something with the fans died in it on my last ride so I have to pull it apart anyway. The problem on that motor is I used epoxy, so I would have to come up with a different approach to angling the fans if I want to go a similar route.

Cheers

 

[pendragon8000]

Well, this is what I ended up doing:

The fans on the disc side are normally positioned to pull air in, but the fans on the freewheel side are offset and angled to spin the air around with the rotation of the hub. I achieved this by using long bolts, which I bent using some pliers after they passed through the stator. Nuts were threaded on before bending to secure the fans on the opposite side, and nylock nuts were put on last to hold everything down.

Using the old hand next to the holes method, the flow does seem to be stronger. The noise is definitely lower than before which is nice, and one of the changes I was hoping to achieve.
I'm taking it out on a group ride with some other ES members this Friday so will give it a good thrashing and see how it holds up then.

Cheers

SWEET JEESUS!

 

[DasDouble]

Subsribed

 

[Offroader]

I'm running into issues with my 35 mm EDF fan. Basically the plastic is failing. My first 35mm edf had the impellar break off.

The 2nd 35mm edf had the two of the three plastic arms that hold motor in place break.

The 3rd 35mm edf had the plastic that holds the bearing break.

I'm out $25 dollars each on those three EDF, so total $75.

I find this odd because I went two seasons with a 30mm EDF in my cromotor that is still working fine.

I assume the heat and possibly the slightly heavier EDF motor may be weakening the plastic on the bigger 35MM EDF. Although I am not running it at a very fast speed.

I won't be buying another 35mm EDF, although I liked these because they were a lower KV and I could run them at 12 volts with lower amps.

I was actually tempted to scrap this whole fan cooling and use FF after my 3rd one broke as I had enough of dealing with three broken EDF fans. However, after having to ride less than a mile back to my house with no fan, the motor was overheating so badly even though I was riding slowly. I realized just how much heat my motor is throwing off and how these tiny fans run at 25% speed keeps the motor cool. I am convinced that ferro fluid and heat sinks can not deal with this kind of heat for the type of riding conditions I ride in. I'm constantly pushing 7,000 watts and always fully on the throttle. Many times I'm peaking at 105C+ on hill climbs and I'd rather have the air flow blowing directly on the windings. Basically the EDF cooling works and it is hard to try something that has questionable performance. Although if the FF + Heatsinks worked as well or better, I'd be much better off just using that instead of a fan.

I think I found a solution. A 40 mm aluminum EDF. This thing is totally amazing and it is only $25 dollars also. It does weigh more at 63 grams where the 35mm was about 18 grams, but this thing is all around better. It also is 8000 KV instead of 6000 KV, since I run it at lower throttle it may be a bit harder on my ESC, but I oversize the ESC and they are cheap enough to replace. I have already got two seasons with my current ESC run constantly at about 25% throttle.

Here is the link for the 40mm EDF I ordered from hobby king.

https://hobbyking.com/en_us/dr-mad-thrust-40mm-8-blade-alloy-edf-8000kv-motor-330-watt-3s.html

This thing is just down right awesome based on the specs. I believe I could also fit it inside my motor, although it will be tight, but I fit a 35mm easily, and I can always modify the EDF or enlarge the hole in my motor. It is much more rugged with a larger motor and aluminum housing. No plastic housing which has broken twice with my cheapo 35mm edf. A review said it is much quieter than his old EDF, so it should less noisy and since it is larger at 40mm I can run it at slower speeds for the same air flow.

Here is a similar 35mm cheapo edf I was using. The one I got was 6,000 KV from aliexpress but basically the same motor with a different KV. This is totally cheap junk that just falls apart quickly. Although I had a cheapo 30mm edf that held up for two riding seasons and many times run for hours each ride.

https://hobbyking.com/en_us/edf35-with-11000kv-motor-assembled.html

 

[ninepointeight]

Since it seems like you 2 are the only ones on the forum that have really executed this cooling solution, I come to y'all for advice. I am running a qs273 at 550 phase amps max and wonder what your advice is, edf cooling, or ferrofluid. I don't really have machining capability beyond a drill press. Since I would have to open the motor for either method, I am going to spray varnish inside. I really like the way you had the the air flowing between the stator and the magnet ring. While it is certainly easier, it seems that you have found ferrofluid to be very inferior?

 

[Cowardlyduck]

I would suggest trying FF first. It's not that hard to clean off if you decide to go fan cooling later, whereas you cannot easily undo vent holes in your side covers.
Ferro Fluids effectiveness depends a lot on riding style, speeds, motor load dynamic and the type of motor you use.
I think a 273 with it's large diameter might benefit from FF more than our smaller hubs would.

Try FF first.

Cheers

 

[Offroader]

Updates on my 40mm aluminum EDF. Finally got around to Silicone it in. This thing is a beast of an EDF in every way. It is ridiculously more powerful than my old 35mm edf. Pulls so much air at much less power and speed. It is built very strong with an aluminum construction instead of the cheap plastic that fell apart in my 35mm EDF. It also is extremely smooth as it is balanced at factory.

It just fit in also, both width and height just fit with not much more room to spare. The motor is very large.

What still seems crazy is this cost $25 while that cheap plastic one that fell apart in 2 hours cost $25 also.

I should have no issues turning this one up at a maximum of 25 amps at 12 volts if I needed a quick cooldown and this wouldn't fall apart. Most of the time it will be run about 1 or 2 amps as 12 volts, which is lots of air flow.

 

[Cowardlyduck]

Nice stuff Offroader.

If E-Bike racing ever becomes a thing, I have no doubts that your EDF method will be copied to keep things cool. 8)

Cheers

 

[DasDouble]

Hey thats some high information sort of poste right there.. Thank you for taking your time to tell us about that. How do you managed to keep that thing in place? Only with silicone?!? I always thought silicone would loos its stability with time.. And I dont think that having a wobbeling fan inside a hubby at high speeds would be a nice thing... Right?

 

[Offroader]

Hey thats some high information sort of poste right there.. Thank you for taking your time to tell us about that. How do you managed to keep that thing in place? Only with silicone?!? I always thought silicone would loos its stability with time.. And I dont think that having a wobbeling fan inside a hubby at high speeds would be a nice thing... Right?

Silicone makes a very strong bond. It would never come loose. However, if I had to remove the fan I could do so also.

 

[Offroader]

Here is a video of my 40mm EDF aluminum fan running at different throttle speeds. I can only go about 1/4th throttle because my power supply is only limited to about 8 amps as 12 volts.

This thing is really powerful and can run really silent at lower speeds yet still move a lot of air. At high throttle the noise does pick up but it will blow an extreme amount of air. I can feel the air flow coming out the sealed bearing on the other side.

This aluminum EDF is really high quality and runs really smooth, it is balanced at the factory. Much better than those cheap plastic EDF ones. Although I used a 30mm plastic EDF for two riding seasons in my cromotor and it has yet given me any issue. The 35mm ones broke in two rides, all three of them I purchased. I believe this is because the 35mm EDF uses a larger motor and the plastic isn't strong enough.

Hopefully this 40mm EDF with its aluminum construction will last a real long time. It only costs $25 dollars as much as the cheap plastic ones so if it breaks after a few years not a big deal. It is easily serviceable also and can be taken apart, it isn't glued together like the plastic ones. The plastic ones glue the bearing and it can't be changed.


[youtube]VchIApATnH4[/youtube]

 

[Cowardlyduck]

That thing certainly moves some air!!!
Given how powerful it is, I would be a bit worried it would break free if it's only siliconed in. Or have you also bolted it somehow?

Cheers

 

[Offroader]

That thing certainly moves some air!!!
Given how powerful it is, I would be a bit worried it would break free if it's only siliconed in. Or have you also bolted it somehow?

Cheers

It is only siliconed. However, it would never come loose. I can remove it, but it takes a lot of work to peel it out of the silicone. You would be surprised how well silicone sticks to plastics and aluminum.

I bet I could put 30KG+ of weight on it and it wouldn't push the fan out.