Idiots guide to water cooled mid motor making

My wish for this motor is to get a good system efficiency and reliability. A route to go cheaper on the controllers seems backwards to me as it lowers both.

A set of double kelly kls7230 controllers would be my first choice. That's 2/3 of the Adaptto mid cost and could give 15kW
..if you believe kelly specs 8)

I've tried them both as singles on the same setup and didn't see a big diff, the kelly seemed to run a bit smoother than Adaptto.

15kW on 85% efficiency is still 2250W of heat. I wonder how large the radiator and flow would be needed to cool this.
Would a small computer cooler like this be anywhere near enough?
https://m.aliexpress.com/s/item/32751871472.html?spm=a2g0n.search-cache.0.0.6a970d6bVKJaRh
 
Oh if you are up for the big amp controllers and can do a light spending the new ASI 8kw controller might be right up your alley. That is 8 kw continuous. Ready to ship within a few weeks or so. http://accelerated-systems.com/products/escooters-emotorcycles/

Two of those peaking @600 A each should really light up your rear tire. Iirc 6-700 $ a pop for those. Unlike Kelly these won't hold back when you crack open the throttle. I've seen some complains about the Kelly's having some sort of firmware soft launch that prevents hooligan accelerations from standstill. Those ASI 8 kw will drain as much battery amp as you can carry, if you got the battery they will do a total of 1200 A peak. :twisted: 8)

Btw those peak number are without any added cooling.....
 
The 8kw ASI has been up on the site for like a year.. I wonder if they will ever sell them in the shop, do you know anything about it?

I did a sanity check on the computer radiator size: most PSU:s are max 750-1000w and the cpu has a max temp rating of 100 degrees so
it's in the same max temp range as the motors.
it would need double the flow to do +2kW cooling so a 360x120mm radiator would likely be too small to keep 15kW continuous driving cool.
 
That's bigger than I was expecting for a computer cooler. Something that size should do the job with a good fan or sufficient airflow from forward movement of the bike. Airflow through a cooler makes a huge difference to it's thermal performance and I think the problem with computer setups is being want quiet(ish) fans, so low air velocity.

delta-T also makes a huge difference to cooling performance and I suspect a motor will happily run hotter than a CPU (I doubt anyone really wants their CPU temp over 50, maybe 60°C).
 
CPU overclockers seem to draw the line at 70-90 deg C so it's in the same region as us.

At high speed this radiator would work but that's not where the bad efficiency of the motor is. (I'd rather avoid the fans completely due to the noise and complexity.)

I think a huge radiator would look really bad on a bike so i hope you're right..after all, many motorbikes don't have fans.

I checked how much the water can "dampen" the peaks when riding hard.
1l water takes 2.5 minutes of 2kW to heat up to 90degC, still it's got nearly 5 times higher specific heat than alu. Surprising!
 
As for the ASI 8 kw, from what I've hear those 8 kw controllers will be ready to release in end of january - early february. If you would like to have a firm date, just email them. They do respond.

For the water cooling, if you are avoiding fans, you could increase the amount of coolant. You should consider an extra large reservoir, holding more water. Also you should look for faster pump.
 
I wonder if they measure cpu die or coolant temperature? There is probably a significant difference (thermal resistance)!

Yes, the specific heat capacity of water is very good (and convenient). Vehicles have radiator fans for when stationery, when moving, even at ~5mph, the fan is not required.

As a very rough guide, if you have ~10m/s (~20mph) of airflow through your radiator, it can be up to 5-10x smaller than if stationery or no fan (this is the difference between forced cooling and convection only!).
 
This cooler and fan is used on the derbi senda moped:
image.jpeg
https://m.ebay.co.uk/itm/radiador-derbi-senda-R-electro/142637496065?_trkparms=gclientid%3DTW3V3F1SSjJWPnxkl7RotQJR_7iiTZq9lezCwzAR5-gkv8s4iM3yECPEUzzT0348&_trksid=p2489528.m4983.l8656

It's roughly the same size as the CPU cooler so Mr punXor was correct :D

These mopeds can be modded to do 100km/h with their inefficient petrol engines so the capacity of the cooler will be enough mated with an OK pump.
I think maybe i'll buy one :D
 
Water has better cooling capacity than alcohol so would be more effective. A misting system of 0.5l water and a nozzle would be really light and effective for racing.

My aim is to get same reliability as a car or dirt bike so it's no option for me :D
 
larsb said:
These mopeds can be modded to do 100km/h with their inefficient petrol engines so the capacity of the cooler will be enough mated with an OK pump.
Click to expand...

That should be more than up to the job :)

As a sanity check, for the rated output power of the moped, it will put at least as much heat into its engine coolant. On a large diesel engine the ratio is about 1:1, on a moped its going to be worse, maybe 1:1.5. So a 3kW moped will be rejecting somewhere around ~4kW of heat to the coolant.
 
Well a 2 stroke will get much of it cooling from the gasoline itself actually. Thats why a laen running engine might run into problems with oversized cylinder etc so to avoid they always jet up the carburators.
 
larsb said:
Water has better cooling capacity than alcohol so would be more effective. A misting system of 0.5l water and a nozzle would be really light and effective for racing.

My aim is to get same reliability as a car or dirt bike so it's no option for me :D
Click to expand...

Yes, but ethanol boils at about 79c, water at 100c. Cooler is better in this case. And we could reclaim the resultant vapor for later ingestion. :) Build a mobile still! 2 stones, one bird! Maybe feed the flammable ethanol into a turbo unit!

I remember flying with my dad in his plane. It was a Cessna then he got a Mooney 201, and it always seemed weird to me to put extra fuel into the engine to cool it. Phase change, vaporization, etc. No extra equipment or weight, and it has worked the last 100 years. Speed cooling the engine leads to cracked blocks, and other bad things. Guess there is a reason that industry died. 1930's tech nearly 100 years later is hard to believe.

Aargh, screw it, lets just do liquid nitrogen! Too hard to get helium these days, so skip that..
 
Initial plan was to let the water pass through the stator hub to cool coils indirectly, the design evolved to cooling of all the coils directly by giving the stator assembly a shell and run water through the whole assembly.

It looks doable with clearance to stator sideplates and coils. Biggest issue might be corrosion on the laminations and i haven´t found much related to this. Epoxy dipping the heated laminations should work to seal them.
water cooling shell.jpg
I plan to move all connections outside of the stator to minimize risk for shorts so motor is again symmetric.
 
glycol like we use in ice engines have additives for anti-corrosion. should not be an issue.
 
I think you're on thin ice Macribs :D :D

It all depends on the material combinations.. i doubt putting any mild untreated steel in a car cooling system would work even with the glycol corrosion inhibitors. i'm glad if you prove me wrong!
 
well there is rust corrosion and then there is galvanic corrosion, for the latter the only way to solve is to do like on boat, sacrifice some metal/anode, boats uses sink i think. for rust anti-corrosion goes a long way, and if water/glycol fills the cavity it replaces vast part of oxygen, oxygen is needed for rust to take place. thats why old ship wrecks can keep together on the sea bed, but will dissolve rapidly if taken on shore out in the fresh air.

other then that you could do a test, use the same material as in you motor and let sit and see how it goes. and see if corrosion will occur. one test beats other mans theorem.
 
larsb said:
Initial plan was to let the water pass through the stator hub to cool coils indirectly, the design evolved to cooling of all the coils directly by giving the stator assembly a shell and run water through the whole assembly.

It looks doable with clearance to stator sideplates and coils. Biggest issue might be corrosion on the laminations and i haven´t found much related to this. Epoxy dipping the heated laminations should work to seal them.
water cooling shell.jpg
I plan to move all connections outside of the stator to minimize risk for shorts so motor is again symmetric.
Click to expand...

Corrosion of the lams will be the smallest problem :lol:

larger air gap due to the shell?
how seal it up with all the wires coming out?
conductivity of the fluid (it will come in contact with the windings)?
 
I think those points are solvable.
-Wires exit from a 30mm channel. Epoxy or silicone will seal it easily.

-Air gap will be kept, the stator will actually be mated with two end cups that are glued to the stator. Slot openings will be filled with epoxy and glass fiber filler after the winding is done and tested.

-100% winding isolation is off course needed. I could dip the whole stator and coils in thermal silicone potting if needed. If i get issues with the isolation then it's possible to switch to oil cooling.. but i guess both controller and motor might've exploded by then :D
 
Is there any way of estimating the effect of the 0.5mm-->0.3mm lamination change and how it affects optimum rpm for the motor?

If it's proportional then my guesstimation for the new rating rpms is around 5000. That's a good increase in max power!
 
larsb said:
i doubt putting any mild untreated steel in a car cooling system would work even with the glycol corrosion inhibitors.
Click to expand...

It's common practice to use (internally) unprotected mild steel for pipes in engine cooling systems. The coolant includes additives to protect ferrous metals which protects them along with cast iron cylinder blocks.

You also won't get any galvanic corrosion between the steel parts and any exposed copper. However, since you plan to fully insulate the winding this shouldn't be a problem.
 
I worked for Valeo doing design of cooling tubes and hoses some years ago. Steel pipes were surface treated with zinc and chrome passivated to stop corrosion. This was for Volvo which are premium quality parts, don't know how the competition relates.

I think on a material level the cooling systems are difficult as it's always a galvanic system electrically connected (aluminium blocks, steel fittings, coal filled rubber material hoses, brass, you name it)

I am not saying that untreated steel isn't used at other OEMs but normally the specs are pretty similar between the premium brands.
 
Fair enough Lars, now I know your experience I think there is little I can add that you won't already know :)

I know of mild steel used in cooling systems with only phosphoric acid passivation (minimal protection at best). It works OK with good coolant but can suffer bad corrosion with poor coolant (or just tap water). This solution is definitely at the lower end of the quality spectrum, though.
 
You must log in or register to reply here.
Back
Top