I'm doing some testing on these hub heatsinks and ill be able to post some meaningful conclusive results shortly, but so far adding a heatsink/radiator been amazingly effective, in ride tests on the prototypes we've been seeing up to 70-80% improvement in thermal handling & even greater shedding at higher temps due to a bigger Delta T and high thermal diffusivity when attached with thermal paste/epoxy, when they're ready ill have some excess for cost soon in 15 and 30mm widths.
I have the magnet ring flange clearance for the MXUS, Leaf 1500w, 3540, and QS205 i'm always open to collecting more data if anyone's doesn't have at least 15mm of flat between the flanges let me know
While i don’t have much experience in thermal analysis, the law of conservation of energy dictates that the heat generated by the heat source must be equal to the heat dissipated by the heatsink to remain neutral under steady state conditions, so while we can’t dissipate all of the excess heat from running a hub motor at 14-20kw peaks and treating them like a room heater, we can do our best to improve the thermal path from stator to the outside air especially during varying loads to bring the mean temp down as much and as quickly as possible.
In my prototype testing with 4X smaller sinks around the hub, each one had 69 fins, with a length of 30mm and width of 20mm & 1mm fin thickness (it varies form bottom to top but 1mm is the mean) and a total length per sink of 145mm *4 Sinks which gives us a fin surface area of 1300 mm^2 times a total of 264 fins
They’re made of aluminium which has a general thermal conductivity of 205 W/m*c and glued on to the ring with Arctic Silver epoxy with approx the same conductivity, we’ll use a figure of 195 to allow for an imperfect thermal connection.
At 100 DegC hub temp and 21DegC outside we get an alpha T, temperature difference of 78 DegC
The convective heat transfer coefficient of air is HTC = 10.45 - v + 10 v1/2 (2)
Where v = the relative speed of the object through the air (m/s)
(50kph is 13.8889m/s) and we’ll have turbulent flow so the approx HTC across a wider range of speeds is about 20 W/m2*c
Using the formula for convection Q= (K*Ab*nΛT) It looks like the heatsinks, at around 50kph cruising speed with the hub at 100c on a 20 degree day, will be shedding an additional 500-600 Watts which is really quite efficient for an unpowered addition weighing only 300 grams.
With a thermal conductivity of 0.195 W / (mm C) and Convection coefficient during rotation of 13W /mm2 and and ambient of 20c the final versions could shed up to 1000W of heat at speed.
In real-world testing we achieved an even better result than we projected, reducing the temps on identical rides repeatedly, from 127C to 48C at the same points on the ride under identical conditions, effetively solving any thermal issues even on intense rides, so i thought it would be a good idea to design these and maybe make it a bit easier for everyone else to solve their heat issues permanently.
in short, radiating heatsinks work well,especially adding almost 3sq ft of surface area, so i'm building some better ones
