Cooling hubmotors

[John in CR]

That's sort of what I had in mind, actually. Only the pipes never leave the hub.

Might have to try it someday...

Link,
If the tubing stays inside the hub, how do you expect the heat to be exchanged to the outside environment? It can't just disappear inside the hub.
John

 

[Link]

Link,
If the tubing stays inside the hub, how do you expect the heat to be exchanged to the outside environment? It can't just disappear inside the hub.

Read back a few posts. The idea would then be to get them really close to the cover plates (which would be covered in heatsinks). Obviously, the small air gap left would keep heat transference from being ideal, but it's at least doable without having to modify the axle or run extra wires (small pump could be powered by a tap on the windings).

Given those qualities, it's really the only thing like that I COULD try...

 

[Zoot Katz]

How about intermittent blasts from the CO2 fire extinguisher you crazy overclockers should be carrying anyway.

 

[RLT]

Imagine a "pie-tin" slightly larger than the sidecover:

• 
  ventilate the sidecover
  attach the pie-tin to the axle, such that it barely touches the sidecover
  attach a fan to the hole in the pie-tin
  do similar to the other side to let the air flow through

"Pie-tin" = vacuformed or stamped shroud.

You should get the general idea... the pie-tin is stationary, a flange on the sidecover perimeter could keep out bad stuff, the inlet and outlet could be baffled for same.

That seems like the most practical method I've heard or been able to imagine on my own.
Maybe a brush type 'seal' around the edges to help prevent leakage...(in or out).

Wouldn't even have to be metal... like a plastic garbage can lid, they could be manufactured for about 30 cents each, sold for five bucks each... Quite a profit.

I like it!
Sign me up for a dozen.

 

[John in CR]

Link,
If the tubing stays inside the hub, how do you expect the heat to be exchanged to the outside environment? It can't just disappear inside the hub.

Read back a few posts. The idea would then be to get them really close to the cover plates (which would be covered in heatsinks). Obviously, the small air gap left would keep heat transference from being ideal, but it's at least doable without having to modify the axle or run extra wires (small pump could be powered by a tap on the windings).

Given those qualities, it's really the only thing like that I COULD try...

Link,

That still leaves you conducting the heat from the source through a layer of air to get to the outer shell then to the environment, not to mention that 100% of the electricity used by the pump motor gets turned into heat. A simple alternative, assuming there is clearance between the spinning cover and the non-moving internal parts of the motor, might be to put some fins, paddles, whatever, on the inside of the covers. The idea would be to create more air movement and turbulence inside the motor in order to get all of the internal parts including the air to the same temperature. That might get the covers to a higher temperature, which has to be the goal in order for an unchanged exterior to exchange more heat.

Wouldn't increasing the axle size to say a 1-1.5"ID steel pipe be easier than putting a pump on the inside? Then just run your tubing to an outside pump and heat exchanger. Other than pressing the existing axle out of the hub, the most difficult part would seem to be the changes needed in the covers to fit the larger bearings required. Even if it proved more difficult, at least it would guarantee better heat exchange.

John

 

[vanilla ice]

As long as you are feeding miniature copper tubing through the axle of a brushed hubby, could you not use the copper tubing as one of the two conductors to save room? Or even plumb two separate systems with insulated tubing, and use one for each polarity. I think very small tubing could still work well as long as the speed of the water flow is adjusted to suit the cooling needs.. Besides, we're not talking about 25 or 50 horse motors here.

 

[Link]

That still leaves you conducting the heat from the source through a layer of air to get to the outer shell then to the environment, not to mention that 100% of the electricity used by the pump motor gets turned into heat. A simple alternative, assuming there is clearance between the spinning cover and the non-moving internal parts of the motor, might be to put some fins, paddles, whatever, on the inside of the covers. The idea would be to create more air movement and turbulence inside the motor in order to get all of the internal parts including the air to the same temperature. That might get the covers to a higher temperature, which has to be the goal in order for an unchanged exterior to exchange more heat.

Heh. This is starting to remind me of Safe's "Ice Kissing" argument a while back.

Anyway, like I said, the gap would hopefully be really small (like a mm or less). An important aspect is that the windings are only around the edges of the sideplates. The middle is mostly hollow. Heatwise, they'd now be right next to them with 10X or more of the exposed surface area.

A few watts from a pump will maybe amount to all of 2% of the heat output at times.

I'd imagine most sideplates already stir up the air a fair bit. Even a flat surface can churn air pretty well, and all the sideplates I've seen have like six small radial fins on the inside for mechanical strength.

Wouldn't increasing the axle size to say a 1-1.5"ID steel pipe be easier than putting a pump on the inside? Then just run your tubing to an outside pump and heat exchanger. Other than pressing the existing axle out of the hub, the most difficult part would seem to be the changes needed in the covers to fit the larger bearings required. Even if it proved more difficult, at least it would guarantee better heat exchange.

John

LOL, undoubtedly. But it'd kill the stealth factor a bit. Though I'd imagine it could also make a pretty cool sort of accessory (no pun initially intended). :wink:

As long as you are feeding miniature copper tubing through the axle of a brushed hubby, could you not use the copper tubing as one of the two conductors to save room? Or even plumb two separate systems with insulated tubing, and use one for each polarity. I think very small tubing could still work well as long as the speed of the water flow is adjusted to suit the cooling needs.. Besides, we're not talking about 25 or 50 horse motors here.

I wouldn't wanna try it. Unless you went through some pretty painstaking procedures to completely insulate the copper from the motor, too, you'd turn the whole motor and bike frame live. The potential for a disastrous short goes way up.

And brushed hubbies aren't really the problem, since their windings are on the outside casing, like this:

And brushless ones have the coils on the inside stator, like this:

 

[Knuckles]

LIQUID COOLED: Drink beer and pee on it?
(I am talking a lot of beer and continuous peeing!)

btw .. that's got to be a goldenmotor (Link's pic).
Identical to the Grubee Black motor.

 

[John in CR]

Link,

Adding surface area to a heater core allows more heat input, but in this case the heat input is fixed. If significant turbulence already exists inside the motor, then I don't believe you're going to get that air hotter than it already is with more surface area on your heater. Higher temperature is required to get more heat out without changing the cold side. What is really required is more surface area inside and outside for the cold side heat exchanger, the motor shell. This will enable you to get the heat out at a lower temperature, so the motor stays cooler.

WRT to stealth and cooling tubes, I wouldn't use a radiator. I'd start with an aluminum frame bike and braze some aluminum tubing directly to the bike. It wouldn't take much tubing or weight to invisibly turn most of the bike frame into a cold side heat exchanger. Then you just have 2 short lengths of rubber or plastic tubing to connect the motor cooling tubes to the frame tubes, so the wheel is still removable. For someone with the right tools, it might prove to be a very worthwhile mod and permit extreme performance limited primarily by batteries.

John

 

[Link]

Hmm, you have a point. However, the core DOES have a very limited surface area to transfer heat to the surrounding air with, which in turn limits how fast it can transfer heat for a given gradient. So, simply mounting a bunch of small heatsinks directly to it might help almost as much as my original idea. All that aluminum would also improve the overall thermal conductivity of the inside of the motor as well as increasing its thermal mass.

I wonder by how much, though...but sticking a bunch of aluminum finning on the core is so simple to do that it's at least worth trying...? :?

After I get the bike working and a suitable way to measure temperature inside and outside the motor, I'll have to go grab some heatsinks and see if it makes any noticeable difference.

 

[snowcrow]

Imagine a "pie-tin" slightly larger than the sidecover:

• 
  ventilate the sidecover
  attach the pie-tin to the axle, such that it barely touches the sidecover
  attach a fan to the hole in the pie-tin
  do similar to the other side to let the air flow through

"Pie-tin" = vacuformed or stamped shroud.

You should get the general idea... the pie-tin is stationary, a flange on the sidecover perimeter could keep out bad stuff, the inlet and outlet could be baffled for same.

I'd put four holes in the tin, for four CPU fans, two blowing in and two blowing out. That way you still have air flow while stopped Blessings, Snow Crow

 

[dogman dan]

Riding my bike with normal voltage, I have no problems till the ambient temp is above 95 F. depending on the wallet size, I may try big holes in the covers next summer once the spring haboobs are over. This summer, riding the bus worked ok, if it showed up. A few times it didn't. I'd like to know if gearmotors run cooler up hills. I think so, but do they really run cool enough to climb in 105 F? Has anyone tried a personal mister on the side of the motor with no wires? Its a pressure bottle with a misting nozzle on a piece of tubing. I used to ride in a no ac car with one on the visor.

 

[fechter]

The idea is to get heat from the copper to the outside air. If the motor stays sealed, then one approach would be to increase the surface area (fins) on both the inside and outside of the housing. You still would have a high thermal resistance from the copper to the air though.

Forced-air is probably the most practical approach, but is tricky with a hub motor. Tyler's 'pie tin' concept is what I was thinking would be close. Even then, the air flowing through needs to be directed at the windings and not allowed to just pass throught without going near them. If you had air flow on one side of the windings, the heat from the other side would be carried through by the copper fairly well.

If the there was a ring of holes in the side cover that lines up with the windings, that would be the best way. Air could come in these holes and exit through holes on the other side or through holes on the same side, but closer to the axle. You'd have to maintain enough metal in the side cover so that you don't weaken it mechanically to the point where it could fail.

 

[John in CR]

Forcing air through the windings might be ideal, but the tight windings and resin will severely limit airflow. The windings have an intentionally good thermal connection to the metal of the central hub, with a thermal conductivity probably 20-30 times better than with air, so blowing more air across any of the interior surfaces is good. BTW , is that metal typically aluminum?

It's been a while since I had a hub cover off, and didn't pay much attention, but I seem to remember the cover plate being smooth on the inside. It may stir the air up some, but forced air convection is far more effective with some real wind and turbulence instead of a slight breeze. That's why we blow on things to cool them off, it works. I'm with Fechter, and space permitting, some kind of fins inside and out is the easiest mod, while retaining a sealed motor. Really get the air churning, and it will create a more uniform temperature throughout the interior. More surface area for the cover plates inside and out will allow the same amount of heat to be transferred to the environment at lower temperature. That is the ultimate goal, because it then allows higher performance and more heat to be transferred while maintaining our current temperature.

With these kinds of changes, be careful of misleading data. eg If your ride ends going up relatively short hill, then the shell of the motor could easily be hotter to the touch than it was before mods, despite better heat transfer. Temperature sensors near the windings is obviously the best way to get reliable data, but a touch test can be a valid tool too. You want to become familiar with how long it takes for your motor to cool after a ride. If it cools more quickly given the same ambient temperature, then your modifications are effective.

Once I have a spare motor and time, I definitely want to give liquid cooling a shot. If these hub motors don't have some kind of inherent limitation on voltage, I may be a way to push a relatively lightweight motor to high performance. I need to leave the ventilated hub motors to the guys ride in dryer cleaner conditions.

John

 

[michaelplogue]

Why not try what they do with computers: Just dump a bunch of mineral oil into the damn thing...

http://video.google.com/videosearch?q=oil+computer#


.

 

[Miles]

Why not try what they do with computers: Just dump a bunch of mineral oil into the damn thing...

It would sound like a washing machine...

 

[HAL9000v2.0]

One idea is to fill motor with heat transfer liquid. Finally I got some to test it. It is called 3M Novec 7000 (old name was Novec 123) and it is isolator and it has kinematic viscosity 0,3 cSt. This winter when riding is over I will install two seals on shaft and fill some 95% of motor volume. It is quite expensive, 100€/lit. Plan is to messure current with and without liquid to calculate hydrodynamic resistance. Volume of air inside X5 is 1873ml (ProE says...)

 

[MidniteTweeker]

500 watts ~ 1680 BTU/hour
1 BTU = energy necessary to raise 1 pound of water 1 degree F so,
1680 BTU = raise 1 pound of water 1680 degrees F, or 1680 pounds of water 1 degree F., or
any combination thereof.

So perhaps 20 pounds of water raised 84 degrees, or 40 pounds of water raised 42 degrees, in 1 hour of time.

It sounds like a lot of heat if indeed 500W is a reasonable number.

MT

 

[Link]

BTW , is that metal typically aluminum?

The rotor is steel, the cover plates are usually aluminum. I think.

 

[John in CR]

BTW , is that metal typically aluminum?

The rotor is steel, the cover plates are usually aluminum. I think.

I'll have to pay closer attention next time I'm in there, which is hopefully never. The windings are around a thickness of steel laminates, but the spoked structure from the center out toward the windings seemed like cast aluminum to me, but I could be way off base. I do prefer the idea of aluminum sinking heat away from the laminates and windings, and it should offer greater thicknesses for more contact with the cooling tubes, not to mention it would be softer and easier to work.

John

 

[snowcrow]

Wouldn't a second stealthy motor hub eliminate the over heating problem all together ? Seems to me that this is the way to go!! Just hide one controller in the trunk bag and run one throttle (just add cord and hall sensor from standard throttle for second controller) 8)

I would rather turn the watt-hours in to horse power rather than heat, which is whats happening

Blessings, Snow Crow

 

[TylerDurden]

Agreed, heat is just a waste.

Two motors might be the best choice, albeit not the cheapest up front... over time the efficiencies may cover the costs and more.

 

[John in CR]

I'll be running dual motors within 2 weeks, a geared Bafang freewheeling rear for hill climbing and acceleration assist, and a direct drive hub on the front for speed. Split duties should keep both cool, while providing greater performance and efficiency than either alone.

This doesn't, however, eliminate heat issues with WOT running for significant distances on the flats. Once you're looking at speeds requiring more than 1hp to maintain, the 10-20% efficiency loss in that turns into heat. At 1kw that's still 100-200W of heat. Even running only on the flats my DD motor gets pretty hot to the touch with runs over a couple of miles. We'll see how much cooler it runs with acceleration assistance, and no hill duties.

John

 

[snowcrow]

I'll be running dual motors within 2 weeks, a geared Bafang freewheeling rear for hill climbing and acceleration assist, and a direct drive hub on the front for speed. Split duties should keep both cool, while providing greater performance and efficiency than either alone.

This doesn't, however, eliminate heat issues with WOT running for significant distances on the flats. Once you're looking at speeds requiring more than 1hp to maintain, the 10-20% efficiency loss in that turns into heat. At 1kw that's still 100-200W of heat. Even running only on the flats my DD motor gets pretty hot to the touch with runs over a couple of miles. We'll see how much cooler it runs with acceleration assistance, and no hill duties.

John

With split duty, % efficiency loss to heat is significantly reduced because the loss is not a linear.
It would be interesting to see before and after data for flats and climbing


Blessings, Snow Crow