how about this, big thin disc of aluminium, heatpaste/glue onto the coils, some holes in it also, lots of space in the middle could put some big junks of aluminium in there ( with thermal contact to teh big aluminium disc) also to soak up some heat, might not be much space between the coils and side plates could be a problem
just something of interest, to flatten down the coils ( on the sides where they are near to the side plates) they just use wooden mallets, and bash/squash it into place.
Still working on it. Only reason I didn't finish it a week or two ago is because I'm still waiting on that LiPo order from Hobby City. Got the chargers restocked thrice. Batteries?
FWIW, the specific heat capacity of water is 4186 J kg–1 K–1 (4186 joules to raise the temperature of 1kg of water 1 kelvin [aka celcius starting from absolute zero]). I'm expecting 200kg/hr through the hub. Rounding down to 4000 J kg-1 K-1 to compensate for whatever I put in it to keep bacteria from growing, that calculates out to about 222Wh/hr/K, which, assuming the radiators are able to cool the water coming from the hub by even 10°C (which I think they should; there's a total of 2m^2 of surface area on their fins), I could theoretically STALL the hub on a 2kW controller and it wouldn't be able to overheat.
Do I have that right? Even for watercooling, that sounds like a little more than I'd expect to be able to hope for...
The Stig's idea might have some merit. Even though there will be some increased friction and they don't transfer heat as quickly as water (which does it about 30 times faster than air), most alcohols and oils still conduct heat between 5 and 10 times better than air according to Wiki. Tradeoff might be worth it.
Most water-cooled computer projects used distilled water with an anti-corrosion /algae growth additive. A good source for this sort of stuff is Frozencpu.com
I'm hesitant to show my work on air cooling because it seem to instill passionate responses beyond the technical review. But I did a crude form of air cooling on my X6 while I wait for my next X6 prototypes. The air is pulled in across the winding and out the disc rotor side. The fans are on both side of the stator and the stator is drilled out to let air pass. This was a fast and dirty experiment but it definitely works. Whats even more surprising is my WH/mile went from 125 to 100 just by adding the fans. I did 2 test loops on a course I have done 100's of times, and really believe air cooling improved the motors EFF about 20% Even the AH used for the ride went down from 12 to 10 AH I was stunned by the results. The course is 9 miles (14.4Km) and the air temp was 85F. I hope to repeat the test a few more time before I have to swap out the wheel for the new prototypes.
Even though there will be some increased friction and they don't transfer heat as quickly as water (which does it about 30 times faster than air), most alcohols and oils still conduct heat between 5 and 10 times better than air according to Wiki. Tradeoff might be worth it.
Only if the media can circulate as fast, or faster than air can; plus similar or greater surface area and airflow is needed to extract the heat from the media to air at the other end of the circuit.
Since air is the final destination for the heat, any cooling system needs to convey heat and transfer it to air faster than air alone (to be an improvement).
As Reid pointed out many moons ago, a small hairdryer fan can continuously extract 1500W of heat from a hairdryer... the coils having a surface area of only a few square inches.
Markcycle's approach seems very practical. I run forced air cooling on my non-hub motors and have not had problems with contamination from road grime. The inlet screens should be good enough. A little rain water won't hurt anything.
You can actually get the cover on with those fans in there?
Even if you could get good thermal conductivity from the windings to the outside of the case, the amount of heat you can dissipate by directly blowing air on the windings would be more.
If the actual motor windings were made from tiny copper tubes with water circulating through them, you might be able to do slightly better, but that would be insanely difficult. If you try to run a copper cooling tube from one winding to the next, you would need a non-conductive section between windings to prevent the cooling line from acting like a shorted winding.
Markcycle's approach seems very practical. I run forced air cooling on my non-hub motors and have not had problems with contamination from road grime. The inlet screens should be good enough. A little rain water won't hurt anything.
You can actually get the cover on with those fans in there?
I had no problem getting the covers on with my X6 but have not tried it with an X5. hopefully I have inspired someone to try it on an X5 as my current work is with the X6. The fans are only 10mm high so I think it will work. The key is the fans need to be under the bearing and the holes in the covers need to be over the windings. I do have a X5 stator in my junk pile and covers so I'll try and do a dry fit with the fans in place later in the week.
The only thing is, you need to be very careful not to nick the winding when drilling the screw holes for the fans and tapping the holes is tricky because the aluminum is very gummy, use cutting oil when tapping. Also use locktight on the threads.
Even though there will be some increased friction and they don't transfer heat as quickly as water (which does it about 30 times faster than air), most alcohols and oils still conduct heat between 5 and 10 times better than air according to Wiki. Tradeoff might be worth it.
Only if the media can circulate as fast, or faster than air can; plus similar or greater surface area and airflow is needed to extract the heat from the media to air at the other end of the circuit.
Eh? I meant they're better than air when it comes to thermal conductivity. Their specific heat capacity is far greater than air's. If it already conducts heat faster than air, why does it need to circulate at all? :?
As for actually getting it OUT of the hub...uh...glue on some fins/heatsinks?
Eh? I meant they're better than air when it comes to thermal conductivity. Their specific heat capacity is far greater than air's. If it already conducts heat faster than air, why does it need to circulate at all?
So you don't have a motor full of hot oil. (It's just more hot mass unless you can cool the oil.) The oil will absorb less and less heat from the windings the hotter it gets.
Eh? I meant they're better than air when it comes to thermal conductivity. Their specific heat capacity is far greater than air's. If it already conducts heat faster than air, why does it need to circulate at all?
So you don't have a motor full of hot oil. (It's just more hot mass unless you can cool the oil.) The oil will absorb less and less heat from the windings the hotter it gets.
In a stock X5 most of the heat transfer from the windings to the casing goes through the air trapped inside the hub. The air gets the heat from the rather large surface area of the jumble of copper windings but then there isn't much surface area around the casing for the heat to get from the air to the casing. The outside temperature air hitting the larger area of the outside of the casing and the spokes should take heat away from the casing faster than the hot air on the inside can transfer heat to the casing. If this is so there is a significant temperature gradient between the stator and the casing. We could be sure of this if someone with a thermometer in their motor could measure the temperature of their casing when the windings are at 100C+ after a quick acceleration.
Adding oil around the circumference of the stator, first of all, will not interrupt the heat transfer from the windings to the air inside the hub! Since oil transfers heat 10 times better than air and the distance the heat is being(thickness of the yellow line in first picture) transferred across is so small compared to the area around the outside of the stator(green area second picture)
Please don't discourage people from even testing this idea. There is potential, even if its a little difficult to quantify.
Besides this is by far the easiest mod in the thread. You just undo the screws to open the side of the casing and pour a bit of oil in. I'm not sure if you can open the side of the casing when the hub is bolted onto the bike but you should be able too. Plus its reversible! If you don't like it or it doesn't work well you can just poor the oil out again, wipe it up and you'll have layer of corrosion protection left in your hub. Much more reversible than drilling holes in the casing.
I can say that with my temp monitor placed on the stator frame (in the middle radius of the axel and the winding), the side cover are at 105 degree when the motor stator frame temp is to around 150.. the gradient is 45 degree C (and it change if you stop the ebike or if you ride with! This is normal.. when i stop , if i measure immediatly the temp of the side cover, it start to increase and stabilize when it get an equilibrium with heat transfered to ext compare to heat absorbed inside..
I can do some more test for you with my IR hermometer and some black tape i'll put in the side cover to get a grat black body effect.. otherwise the aluminium side cober will reflect any IR... and i need it to only emitt.
I can say that with my temp monitor placed on the stator frame (in the middle radius of the axel and the winding), the side cover are at 105 degree when the motor stator frame temp is to around 150.. the gradient is 45 degree C (and it change if you stop the ebike or if you ride with! This is normal.. when i stop , if i measure immediatly the temp of the side cover, it start to increase and stabilize when it get an equilibrium with heat transfered to ext compare to heat absorbed inside..
I can do some more test for you with my IR hermometer and some black tape i'll put in the side cover to get a grat black body effect.. otherwise the aluminium side cober will reflect any IR... and i need it to only emitt.
Thanks Doc. Thats a decent temperature gradient. Was this measurement taken when you stopped after you had been cruising at speed? How fast would the temperature gradient stabilize once you had stopped the e-bike?
According to my theory in my last post, if the problem is the lack of heat transfer from the air inside the hub to the casing then maybe we could just increase the surface area of the inside of the side covers :? sticking fins or machining grooves to the inside of the side covers may be riskier and harder than pouring a bit of oil into the hubbie.
I pulled an X5 stator out and looked to see if the fans I have indeed fit here are the pictures.
The fans fit and have a few MM clearance as shown in the last picture. These fans are 40mm X 10mm fans, there are even smaller fans 30mm if needed. The key is to keep the fans under the bearing
i'll search for some of them at my local electronic surplus stores. They have like 35 different fans... i'll search for some with ball bearing that draw more mA.. they usually are more powerfull...
I might also see for some small dc-dc to put inside the motor that could be powered thru a six diode bridge rectifier direct on the 3 phases wires :wink: .. or the pcb of a small AC-DC adaptor (100-240V) that can accept lower voltage.. and thai i could put indide.
i'll search for some of them at my local electronic surplus stores. They have like 35 different fans... i'll search for some with ball bearing that draw more mA.. they usually are more powerfull...
I might also see for some small dc-dc to put inside the motor that could be powered thru a six diode bridge rectifier direct on the 3 phases wires :wink: .. or the pcb of a small AC-DC adaptor (100-240V) that can accept lower voltage.. and thai i could put indide.
i'll search for some of them at my local electronic surplus stores. They have like 35 different fans... i'll search for some with ball bearing that draw more mA.. they usually are more powerfull...
I might also see for some small dc-dc to put inside the motor that could be powered thru a six diode bridge rectifier direct on the 3 phases wires :wink: .. or the pcb of a small AC-DC adaptor (100-240V) that can accept lower voltage.. and thai i could put indide.
I might also see for some small dc-dc to put inside the motor that could be powered thru a six diode bridge rectifier direct on the 3 phases wires :wink: .. or the pcb of a small AC-DC adaptor (100-240V) that can accept lower voltage.. and thai i could put indide.
You can power it from the hall sensor +5V supply. You have to provide beefier external 5V source rather than use controller 5V but this is easier than internal rectifier/DC converter. Just add some mlcc shunting cap for 5V inside.
