HeatSink that works well for smaller hub motors....

P

pullin-gs

1 kW
Joined
Jul 31, 2009
Messages
395
I have been running my 500W Q128C motor at 800W max sustained (controller set to 20-amp current ceiling....12S1P 8AH)....after about 5 miles motor was getting warm to the touch.
I tried this very simple .....and surprisingly effective...heat sink.
I use an aluminum disk screwed to the aluminum housing of the hub motor's brake rotor mount.
Heat-sink paste between the motor housing and disc provides good thermal transfer from case to disc. Heat transfer occurs by convection into the surrounding air that flows over the disc.
Result is case runs much cooler.
It also has the added benefit that the motor is shielded from view by the disc, and on the other side shielded by the 8SP cassette.
IMG_0724.JPG
IMG_0725.JPG
 
While that is a nice idea, I think just a steel brake disc would cool as well.

You really stress that motor, it won't matter any. Inside, the heat in the core transfers to the outer case very poorly and slow. You can melt the inner parts while the case is still nearly cool. If you really load one up, such as 400 pounds of total weight, and a long hill.

If you are not overloading your bike, stop worrying about overheat, its not going to happen. If you are overloading the bike, then start shopping for a heavy, powerful direct drive right now. In general, if you can't go 15 mph with full throttle, its overloaded, by the weight, or the hill, or both in combination. If you can still go 15 mph under full throttle, you are very very very unlikely to overheat it. Its just not overloaded enough.
 
dogman dan said:
While that is a nice idea, I think just a steel brake disc would cool as well.
Click to expand...
***NO....thermal conductivity of rotor steel is less than 40W/mK....Aluminum is over 230W/mK!...rotor only 1/6 of aluminum!***
***....also rotor is not solid disc....only 20% of area of solid disk on area that performs majority of cooling (outside 50% diameter of disc)***
****Steel rotor sinks only 1/20th the heat at point where it bolts to case (where heat transfer occurs) that solid aluminum plate.***
****Steel rotor is thinner.....decrease by ANOTHER 50% compared to aluminum.****
Here are facts to back up these numbers: https://www.engineeringtoolbox.com/thermal-conductivity-metals-d_858.html
dogman dan said:
You really stress that motor, it won't matter any. Inside, the heat in the core transfers to the outer case very poorly and slow.
Click to expand...
****slow yes....but much faster if case is kept cool relative to inside motor.....if case is already hot-to-the-touch, heat buildup occurs much quicker inside****
****Aluminum heat sink with 170mm diameter will sink over 40W of heat if outside temp is 75degrees-F and case is 130degrees-F.
***Motor already can run sustained 500W WITHOUT sink before thermal runaway begins.****
***I'm over-watting by 300W .....excess heat that needs to be removed is about 60 watts @80% efficiency of motor.
*** My 40W (more actually...my calculations DO NOT take into account moving air) heatsink is indeed providing a substantial heat dissipation.
For facts on how I came up with these numbers: https://www.heatsinkcalculator.com/blog/how-to-design-a-flat-plate-heat-sink/

dogman dan said:
You can melt the inner parts while the case is still nearly cool. If you really load one up, such as 400 pounds of total weight, and a long hill.
Click to expand...
Every degree cooler case is kept aids in cooling inner parts. Thermal runaway is what kills motors....which results in melted gears and toasted windings.
dogman dan said:
If you are not overloading your bike, stop worrying about overheat, its not going to happen. If you are overloading the bike, then start shopping for a heavy, powerful direct drive right now. In general, if you can't go 15 mph with full throttle, its overloaded, by the weight, or the hill, or both in combination. If you can still go 15 mph under full throttle, you are very very very unlikely to overheat it. Its just not overloaded enough.
Click to expand...
15MPH? I'm pushing just under 29MPH fresh off the charger. 20MPH on 8% hills (controller limits current at 20amps).
Show me a 1000W trike setup that weighs less that 50lbs which offers agility and performance this setup gives.
The last thing I want to do is add a heavier motor and heavier pack.
Upgrades which let me get this kind of performance from 500W motor include:
*)Manage current....never ever go over 800 watts....motor grenade due to overheat and/or overstress.
*)Reduce platform inefficiancies ....thin high-pressure tires and areo design easily free up more than 60 watts...lighter components help...especially on the few hills we have around here.
*)Control heat.....heavier welds/strappings on packs, heavier/shorter component wiring/connectors...Heatsink on motor.
*)Lightweight components....use high-C lower capacity LIPO (can dump 20C)....18650 give better range...but weighs more. 500W geared motor weighs less than half of 1000W direct drive. 20" alloy rims and stainless spokes. 1X crank setup.
PS: I'm currently have over 2000 miles on this build. Drive train issues I have run into so far include broken spokes, broken CST, and a failed 1500W controller.
 
Thats it. If we are gettin to sciencing up some science,


Imma gonna out science all o yall.


Im gonna fill my hub up with helium and outdissapate all o yall. Alloyall. Datalog or it didnt happen. Lol.

HuckerShucks. Seal and Heliumpurge that sucka. I'll show u .

Sheeeze I'll out cool yall with pantyhose and a bottle o water, and you can have as much aluminum as you want and I'll cool more. Lol
 
DogDipstick said:
Thats it. If we are gettin to sciencing up some science,


Imma gonna out science all o yall.


Im gonna fill my hub up with helium and outdissapate all o yall. Alloyall. Datalog or it didnt happen. Lol.

HuckerShucks. Seal and Heliumpurge that sucka. I'll show u .

Sheeeze I'll out cool yall with pantyhose and water, and you can have as much aluminum as you want and I'll cool more. Lol
Click to expand...
Bring it on (the science part). I am always open to productive dialog which improve my toys.
Everything else you have posted so far is just noise. ;)
 
pullin-gs said:
DogDipstick said:
Thats it. If we are gettin to sciencing up some science,


Imma gonna out science all o yall.


Im gonna fill my hub up with helium and outdissapate all o yall. Alloyall. Datalog or it didnt happen. Lol.

HuckerShucks. Seal and Heliumpurge that sucka. I'll show u .

Sheeeze I'll out cool yall with pantyhose and water, and you can have as much aluminum as you want and I'll cool more. Lol
Click to expand...
Bring it on (the science part).
Everything else you have posted so far is just noise. ;)
Click to expand...

THATS IT IM GETIN MAH ENGINEER. U WAIT.

( I do want to see some empirical data before I do though, you know.. accurate and measurable data to some standards.... do you have any? Like temp reading with contin op?..? PullinGs? Kinda important if we fgonna have a cooloff competition.. I do have luvs for the sciencin though for real... ).. and yeah.. Pantyhose and a bottle of water... .:).. so... Empirical data? )
 
DogDipstick said:
THATS IT IM GETIN MAH ENGINEER. U WAIT.

( I do want to see some empirical data before I do though, you know.. accurate and measurable data to some standards.... do you have any? Like temp reading with contin op?..? PullinGs? Kinda important if we fgonna have a cooloff competition.. I do have luvs for the sciencin though for real... ).. and yeah.. Pantyhose and a bottle of water... .:).. so... Empirical data? )
Click to expand...

Very nice reply. Reminds me of my daughter's reply to her little brother when he tugged her ear at dinner. ;)
 
So.. how much surface area is the conduction path for your heatsink? You have the six ( 6mm?) bolts holding it on the conduction path, the six little bosses that are on the cover plate.

How much surface area is there? in the contact area that gets the thermal grease.
Also, wat is T1, the temperature at the hub cover plate surface, and the air temp? What r those? Like average, dont need exact. Just trying to figure out how much entropy goes on...


( I love that word.. Entropy.. lol...


en·tro·py
/ˈentrəpē/
Learn to pronounce
noun
noun: entropy; plural noun: entropies; symbol: S

1.
Physics
a thermodynamic quantity representing the unavailability of a system's thermal energy for conversion into mechanical work, often interpreted as the degree of disorder or randomness in the system.
2.
lack of order or predictability; gradual decline into disorder.
"a marketplace where entropy reigns supreme")
 
DogDipstick said:
( I love that word.. Entopy.. lol...
en·tro·py
/ˈentrəpē/
Learn to pronounce
noun
noun: entropy; plural noun: entropies; symbol: S
Click to expand...
4-for-4.....So every post you have made contained an insult of some sort or another. :(
Nice job!
PS: I got a really good laugh when you said a STEEL rotor cools just as effectively as a solid aluminum disk! :D :D
PSS: Your redneck-speak sealed the deal though! :D :D
 
I am not tryig to be insulting buddy, am renek, true, lol. ....

I am sorry if I insulted you. Im just a goof, take anything I say online with little merit, please, nothig personal at all.

I really do have an engineer. Lol. He is a ASME certified business owning heat expert, willing to look at these things, for me, for us. With absolute 50+ year experience in just this.

I am at the point in my BLDC career that yes, imperative knowledge of heat extange would be good for me to get into.

Again, I try no insult, friend. I'll shut up if you like instead.

Also, I did not reference any steel or anything like that? I said Helium purge or a pantyhose rag wrapped around the hubsink area with a bottle of water splashed on it. I can easily show you empirical data on heat peaks ( datalogged at 5Hz/sec? or .15sec/log?) with both methods.. with many different peak draws and contin draws from my personal setup... if you would like. However, if you think I am insulting you, anyway... anyhow..


later gator.
 
Also, giving (I think) you mistook "dogmandan" for "dogdipstick".. two completely different entities. We are not the same posters, and I would like you to acknowledge this for we represent each our own separate experiences here. This is two separate people.


No insult intended.

I am 40 years old.
 
DogDipstick said:
I am not tryig to be insulting buddy, am renek, true, lol. ....

I am sorry if I insulted you. Im just a goof, take anything I say oonline with little merit, please, nothig personal at all.

I really do have an engineer. Lol. He is a ASME certified business owning heat expert, willing to look at these things, for me, for us. With absolute 50+ year experience in just this.

I am at the point in my BLDC career that yes, imperitive knowlege of heat extange would be good for me to get into.

Again, I try no insult, frined. I'll shut up if you like instead.

Also, I did not reference any steel or anything like that? I said Helium purge or a pantyhose rag wrapped around the hubsink area with a bottle of water splashed on it. I can easily show you empirical data on heat peaks with both meathods.. if you would like. However, if you think I am insulting you, anyway... anyhow..


later gator.
Click to expand...
Truce. :)
Pass this on to your engineering friend (who is he? I may have done business with him)...Its simple physics BTW...just about any engineer worth their salt can figure this out.

Here is my disk:
https://www.ebay.com/itm/1-8-125-Aluminum-Disc-x-6-Diameter-Circle-Round-5052-Aluminum/192248832433?hash=item2cc2ec61b1:g:hGgAAOSwu1VW5N~x

Other metrics:
3.9 in^2 for hub-motor mount.
140-degrees F is target max heat of motor case.
75-degrees F is ambient temp.
I use standard heat-sink paste.
Anodization of motor case removed....mounts flush to disk.
Mounting screws are stainless....and go .33 inches into mount base on hub....although this is splitting hairs.
Hub is about 1/4" at thinest point...this is splitting hairs as well.
Airflow is 20MPH.
 
Add an internal temp sensor to get the temp of the stator to check the before and after temps. The problem with a geared motor is that it doesn't transfer heat well from the stator to the case, so cooling the case doesn't help the motor much. If you fill the motor with oil so that the heat transfers to the case, then measuring the case temp would be relevant.
 
Looks/sounds like that would work good on my 9c dd. Have you thought about hub sinks. Or ferrofluid?

I could employ all three on mine I suppose. I am using about 35 watts now. I pushed it to 55a for awhile and dropped it down to 35 when I quit using lipo.
 
Sorry if the dog username led to confusion. I'll concede that the alu conducts heat better than a steel disc.

But you are NOT overloading the motor, as you stated by your speed up hills. This means you can ride years without cooking off that motor. There is some simple science for you. I'm saying its nice heatsink, but you don't need it.


And if you do overload, like you get a trailer or something, then it won't stop your motor from cooking off because the transfer of heat from the core winding to the outer shell of geared motors is very poor. It happens that fast, and can happen while the outer shell is cold if you overload it enough. Likely you don't know, I am a professional motor cooker. I would take product from Ebikekit, and go kill it deliberately. The company needed to know what the real world load limits were on its motors, so it could set limitations on warranty replacements.

What you have will be good to go for up to 300 pounds total load up very steep hills. At that weight, it should easily maintain 15 mph. Load it to 400 pounds, and it may slow to 7 mph up a steep hill, at which point it will melt in about 15 min.

If you want to overload, and get away with it, fill it with the ferrofluid. But you don't need it.

If you want to know what is going on in there, your best bet is to put a thermometer on the axle stub, then cover it with foam so it doesn't air cool as much. The axle IS just about the only heatsink that IS connected to the core of the motor. and its bolted to a large piece of aluminum too. So the temp of that axle will tell you something about how quick you are heating up that motor inside. It can't give you inside temps, but it can definitely tell you if its heating up fast or slow. If you cook one off fast, that axle does heat up fast. You can watch that axle temp, and know more or less how hard you are pushing that motor. If not overloading, you reach an equilibrium internal temp that will never reach melting off your hall sensors temp, or even darkening your winding temps. But in winter, the thermometer will cool off too much, and not be as good an indicator as it can be in summer.

But if you have that thermometer, then you have something you can compare to other rides, and have a clue when to back off if you are worried about cooking your motor. You see a temp at the axle stub you never saw before, and its still climbing, then it could be time to worry. But no worry if you just see it max out and stop getting hotter, at a temp you know is fine.

But again,, its crystal clear you are not going to melt that motor as you are running right now.
 
"Pass this on to your engineering friend (who is he? I may have done business with him)...Its simple physics BTW...just about any engineer worth their salt can figure this out."
Click to expand...

Done and done.

IDK, have you worked for any company called GE Boat? How about Raytheon? Washingtog Group International? Holtec? Aecom? I am sure there is more: He has taught a few courses in engineering and written a ton of papers. Runs a lil company called "Beckersville Steam Engineering" where we spend alot of time on powerplant stuff and rich guys boats. Absolutely a heat expert, He is my Father. He has had his hand in over 1000 nuke, coal hydro, and gas plants... Sent all over the world since I was a kid. Has been INSIDE just about every nuke/coal/oil/gas plant in the east coast.. He is the go to for heat transfer. He is my father. Question like this is what he spent his entire career on.

He sees certain materials chosen.. like copper for tube fins in boilers.. vs steel... and can quantify everything easily, unlike me. I am going to set up an account for him and have him post here so we get it right. He saw this question asked before I even got itout and he told me he runs into this all the time, some young engineer specs copper for a boiler tube thinking it would be better when steel would do.... Then he comes in when the plant has problems for him to figure out.

Yes, and we got into the fin efficiency, of our situation here.. he explained me that this transfer efficiency is very relevant here, a aluminum disk makes little difference to the steel, and the fin equations can ( and will, when he gets here).. easily show this. Pantyhose wrapped and water splashed on the hub would cool much better (UNTIL the water is gone, then you have an insulator, lol.... .)

Effectiveness of extended surfaces must be compared. For example, the radiator that cools your car.... If it was pumping air through theblock, pump and was submerged ( the radiator) in water,( opposite from normal) the effectiveness of the thermal transfer would be little..... but when you put water in the block and pump it to the rad where the transfer surface is great ( fin efficiency?) and put the radiator in airstream, the system can effectively cool... Why? You would think fins in water wold cool better than fins in air.... but.... the block cannot cool with air. The air does not let the heat reach the radiator effectively....

He compared it to a triathlon race... Where 10 miles is on foot, 10 miles in car, and 10 miles on horse. Average finish time? lets say 1 hour 30 min... ( 5 mi on foot- 1 hour ( average shoes) , 10 miles in car ( average car), 10 min, and 10 mile on horse ( average horse) , 20 min.. and the race is finished in 1 hr and 30 min) ....

Now, lets say some guy enters the race.. with a Vette' that can go 120... and the car portion is now twice as fast.. 120 mph.. and 5 min.... Goes BALLS OUT for this portion of the race, 10 mile, and completes the ten miles in 5 min instead of the normal cars 10 min. .... The race is still only 1:25 long.. not much improvement, even with a 120$K car going twice as fast.... the walking and the horse are still going to take their time.... and this is where fin effectiveness comes into play.

You would have better luck breeding a thoroughbred or buying a set of Jordan Nikes so you can speed up THAT portion of the race, and not worry about the ( negligible ) difference between a Vette and a normal car... for the portion of the race that is in a car... for a Vette only saves me 5 min, where a Thoroughbred may take the ten miles in half the time and the Nikes may make the race take half the time ( 5 miles on foot, taking 1/2 hour, running on Nikes instead of walking on "shoes") )... where the "time" is relevant, being a race... and where you can save more, means whether you can win or not.... With Nikes, you beat the guy with the Vette by 25 min... with your average car and average horse... When he had a 120$K car, didnt help much.

He will be along in a few days, when I can get him signed in and posting. May take a week or two, but he is interested in these things.
 
DogDipstick said:
"Pass this on to your engineering friend (who is he? I may have done business with him)...Its simple physics BTW...just about any engineer worth their salt can figure this out."
Click to expand...

Done and done.
<SNIP>
He will be along in a few days, when I can get him signed in and posting. May take a week or two, but he is interested in these things.
Click to expand...
Thank you sir. :)
Tinkering with my trike is more fun than the ride itself.
Your dad will love participating on these boards. Every now and then some really interesting engineering topics are brought up by other forum members pertaining to their projects.
I'm a retired engineer (EE by education, telecommunications engineer in practice). I have found that implementing E-power projects very rewarding and keep my mind sharp.
 
It would seem that the bike frame is more effective dissipating heat from the stator than the heat sink, since the only direct path for heat to travel through conduction is via the axle. The path to the case is via the axle, then through the bearings, to the case, and to the heat sink. Cooling the case is only cooling the air around the cooking stator.
 
E-HP said:
It would seem that the bike frame is more effective dissipating heat from the stator than the heat sink, since the only direct path for heat to travel through conduction is via the axle. The path to the case is via the axle, then through the bearings, to the case, and to the heat sink. Cooling the case is only cooling the air around the cooking stator.
Click to expand...

Case is warm to the touch, so heat is indeed being exchanged away from heat source (stator magnets and armature coils).
It happens via three mechanisms.
1) Thermal radiation....This does not require a material medium for its propagation.
2) Heat conduction....mostly occurs via solid components which make contact with each other.
3) Heat convection....transfer of heat between an objects and its environment, due to motion of a moving medium (gasses, liquids)
 
Code:
pullin-gs said:
E-HP said:
It would seem that the bike frame is more effective dissipating heat from the stator than the heat sink, since the only direct path for heat to travel through conduction is via the axle. The path to the case is via the axle, then through the bearings, to the case, and to the heat sink. Cooling the case is only cooling the air around the cooking stator.
Click to expand...

Case is warm to the touch, so heat is indeed being exchanged away from heat source (stator magnets and armature coils).
It happens via three mechanisms.
1) Thermal radiation....This does not require a material medium for its propagation.
2) Heat conduction....mostly occurs via solid components which make contact with each other.
3) Heat convection....transfer of heat between an objects and its environment, due to motion of a moving medium (gasses, liquids)
Click to expand...

Yup, and #1 and #3 can't pull sufficient heat from the stator, which leaves you with #2. Add a temp sensor and check a before and after with and without your disc.
 
Any additional surface area will improve cooling but the thermal resistance between the windings and the disk will make the improvement pretty small. If it was a direct drive motor with ferrofluid it would make much more improvement.

You want the heat path to be as short and fat as possible.
 
I wonder if you could cut blades on your disc and fan force air at the hub.
 
fechter said:
Any additional surface area will improve cooling but the thermal resistance between the windings and the disk will make the improvement pretty small. If it was a direct drive motor with ferrofluid it would make much more improvement.

You want the heat path to be as short and fat as possible.
Click to expand...

Since it's geared, oil is an option to transfer heat to the case. Significant difference in this example. After adding the oil to transfer the heat, heat sinks make sense, otherwise barely any of the heat makes it to the case:

https://www.ebikes.ca/tools/simulator.html?motor=MG310_FST_SA&batt=cust_44_0.2_8&hp=0&axis=mph&grade=0&motor_b=MG310_FST&batt_b=cust_44_0.2_8&hp_b=0&bopen=true
 
You must log in or register to reply here.

Similar threads

CUDAcores89
Thermal paste on hub motor windings. Dumb idea or dumb idea?
Replies
10
Views
2,300
MadRhino
MadRhino
E-HP
    • Like
    • Cheers
My econo-e-bike
10 11 12 13 14
Replies
327
Views
58,146
calab
C
E
I have found a way to easily and succesfully cool hub motors
2 3 4 5
Replies
101
Views
14,103
Kyle201
K
S
Rad Power Bikes: Rad Rover Review
Replies
4
Views
6,354
Ddan2
D
S
    • Like
Rad Power Bikes: 2016 Rad Wagon Review
Replies
6
Views
5,817
Syonyk
S
Back
Top