Triple Rewind of Unite 500W Motor

[safe]

Ice Is A Proven Technology

The world record electric car (314.9 mph) used ice cooling.

I've used it informally by just strapping an ice pack to the motor shell and the cooling is almost instantaneous.

Solid materials are roughly 25 times better as a conductor of heat than gases. Liquids are somewhere in the middle. Generally speaking conducting materials for heat tend to be dense and non-conducting materials tend to have lot's of space in between. The tiles on the space shuttle act as an insulator and they also weigh next to nothing. (extremely low density)

Fiberglass is not as good as steel... but it's probably about equal to water and my guess about 5 to 10 times better than air.

Air is definitely the worst... but it's free and renewable.

Water is good if you recycle it through a radiator.

Ice is a one time only affair. (needs to be replaced after each ride)

-------------------------

So the correct answer is that ice is very good.

-------------------------

http://engineering.osu.edu/nie/article.php?e=793&s=2&a=3

Buckeye Bullet 2 Specifications

Designed and built by Ohio State University College of Engineering undergraduate and graduate students (including a student-designed special driver pod for safety)
Custom-built electric traction motor
Aircraft braking system, augmented by parachutes
Advanced data acquisition and wireless telemetry system to provide realtime data back to the pits during the land speed runs
Four-wheel independent suspension
Custom six-speed transmission
Student-designed, custom ice cooling system capable of dissipating 500 kilowatts of heat
Carbon fiber/honeycomb body
Steel alloy chassis

------------------------------

Thermal conductivity

http://en.wikipedia.org/wiki/Thermal_conductivity

Air --- 0.025

Epoxy (fiberglass) --- 0.30 ~ 0.59 (10 to 20 times better than air)

Water --- 0.6

Ice --- 2

Steel --- 12.11 ~ 45.0

Copper --- 380

...note: I've thought about doing it with copper and then brazing the metal, but I don't have a propane torch and I figured that a slower rate of heat transfer would actually be okay because it would allow the ice to melt more slowly, thus allowing it to deliver it's cooling effect more slowly.

 

[Fuzzo]

I got no problem with the idea of using ice to cool the motor for a very short burst. The question is how to use it, and I don't think a block of ice sitting in an insulated box statically is going to work very well. If you have it running around the motor in metal tubes, sure, it will take the heat out, until the water reaches the temperature of the engine, at which point it will work in reverse, keeping the heat in.

Yes, air is not a particularly efficient coolant, but it does have the advantage that the entire motor is completely immersed in it, and you have a constant supply. If you move the air, then you increase how much heat is being dissipated. That's why it gets colder the faster you go, on a bike, it's why sitting next to a fan cools you down. It's why the radiator is at the front of the car, and why it has a fan on it.

Air will get into every crevice in the entire motor (in fact it already has, and is sitting there, really hot, if there no airflow). If you can get more surface area into your motor, some sort of heat sinks, then you'll get good cooling.

 

[safe]

Thermal conductivity

http://en.wikipedia.org/wiki/Thermal_conductivity

Air --- 0.025

Epoxy (fiberglass) --- 0.30 ~ 0.59 (10 to 20 times better than air)

Water --- 0.6

Ice --- 2

Steel --- 12.11 ~ 45.0

Copper --- 380

------------------------------

If the coolant reaches the shell it gets transferred very, very fast. I've used ice packs all last summer on my motors and it nearly instantly gets all the way around the shell.

Assume 12 for steel and 0.025 for air so:

12 / 0.025 = 480

Steel conducts 480 times faster than air.

-----------------------------

My advice is to take an ice pack and press it against the motor and see how fast it actually cools the motor. It's frankly quite shocking. When I first tried it I could barely believe things worked like this, but the math all backs it up. The motor does a very good job of spreading the heat around evenly... so if the shell is kept cold the motors internal parts tend to not get far away from the shell. It's when the shell gets hot that everything else becomes uncontrollably hot and that's when you get a failure. Keep the shell cold as ice and the insides can't overheat.

-----------------------------

The fiberglass should help to slow the cooling process and while that means the motors average runs a little higher it should make it last longer.

Once the ice melts it then rises up to the boiling point and vaporization is another area where a lot of cooling takes place. It's only when all the ice has melted, then blown away as steam that I have no more cooling.

The real question is:

"What is the total volume of ice to completely cool a specific motor for the duration of expected power output?"

...that's where you can get an idea of how much "boost" to run because you can figure how much extra heat is going to happen (you can calculate that) and then figure how much cooling is needed to compensate.

Ideally you just accept inertial cooling (it takes energy just to get the shell hot in the first place) and a hot motor radiates energy as well.

---------------------------------

The question of "thermal conductivity" is already answered though... the heat transfer of steel is so high (480 times air) that whatever happens inside the motor happens at a rate that is an order of magnitude slower than what the shell is doing.

----------------------------------

Finally... copper is just insane... if we compare the thermal conductivity of copper to air that's:

380 / 0.025 = 15,200 times faster

...so the heat can build up much faster in the coils than the iron core. And the air is only able to dissipate that away at a slow rate which means the shell only gets heated up slowly. The hardest thing to actually keep cool is the iron core and if you saturate that with too much current (ripple currents) it very quickly overheats. The ideal approach would be with motors that used the copper on the outside and had water (ice cold) running through near the copper itself. With the permanent magnet brushed motors the design limits your cooling options to keeping the shell cold.

If the shell is colder than the outside air temperature then the internal air circulation within the motor will cool better than with pumped in outside air.

It's better to "recycle" internal motor air and cool it with an ice cold shell than to import outside air to try to lower the temperature.

But this is all dependant on the ice not running out...

 

[safe]

Doing (Redoing) My Homework

My Ice Cooler holds 250 ml of water.

When that water is frozen as ice it will hold:

250 ml * 1 g/ml (for water) * 80 cal/g (heat of fusion) = 20,000 cal

Which means that just to melt the ice the Ice Cooler will be able to extract 20,000 calories of excess heat away from the motor.

Is that good?

The "rated heat" for the motor I'm going to use it on is about 100 watts. This means that when I'm running the motor within it's design specification that it can deal with this much energy "forever". (that's what the "rated load" means... that you can run like that without burning the motor up)

So let's say that I want to run my motor above the "rated heat" (in my case running the motor at 60 amps current limit) for the duration of my ride. Let's assume that the "extra" heat is an additional 100 watts that needs to be extracted by the Ice Cooler.

Does it work out?

I figure that full throttle running will give me a runtime of about 30 minutes. So the total "excess" energy is:

100 watts = 100 joules per second * 30 minutes * 60 seconds / minute = 180,000 joules

4.1858 joule / cal * 180,000 joules = 753,444 cal

...so it only takes (20,000 / 750,000 * 30 = 0.8) about one minute before the ice melts if we assume instantaneous heat transfer.

Taking the motor from zero to boiling it takes:

250 ml * 1 g/ml (for water) * 100 cal/g (heat of fusion) = 25,000 cal

Then boiling to steam:

250 ml * 1 g/ml (for water) * 540 cal/g (heat of fusion) = 135,000 cal

...so this gets us to:

20,000 cal + 25,000 cal + 135,000 cal = 180,000 cal

...so it only takes (180,000 / 750,000 * 30 = 7.2) or just about seven minutes of full throttle before the ice is all steam.

--------------------------------

The "bottom line" is that the motor is going to be able to handle extra power that would normally burn the motor up, but the time limit is going to be about enough for one high speed run and that's about it.

Which was the goal from the beginning...

Normally I'll run the motor within the "rated heat" by using an "Armature Current Limiting" circuit and that will place a cap on the heating that will keep it under the "rated heat".

--------------------------------

Ice is used when you want intense cooling for a short period.

--------------------------------

In order to fully cool the motor (at freezing) at full throttle when in a seriously high overcurrent limit situation (high levels of boost) it would take:

753,444 cal / 80 cal/g = 9418 ml = 9 liters of ice

 

[duivendyk]

You are both misinformed,what really matters is not the conductivity of ice but the amount of heat it takes to melt it.It takes 93 Wh to melt 1kg of ice,or 42.5 Wh to melt
1 lbs.To illustrate,if you had a motor with 240W input ,75% efficiency,and 60W (0.25x240) dissipation you could keep the shell at 0 deg C for (42.5/60)x60 =42.5 minutes (if that was the only cooling) with your 1 lbs of ic.It turns out that 1kcal is 1.17 Wh.So if you also let it heat that 1 lbs of water to 40 deg C,it could absorb another 0.455x40x1.17=21.3 Wh and that would take another 20min.If you did not know it 1kcal is the heat it takes to heat 1kg of water by 1deg. C.

 

[safe]

There are two issues.

The first is the ability of the shell to transfer cooling to the air inside the motor and that requires that the cold moves quickly through the mediums of fiberglass and steel which they do.

The second issue is the quantity of ice needed.

Maybe I got it wrong... I'm "attempting" the homework.

1kg = 1000g = 1000ml of water.

My Ice Cooler is 0.25kg.

Assuming a heat dissipation rate of 100W how long will that take in your calculation?

------------------------------------

This might be my mistake:

"4.1858 joule / cal * 180,000 joules = 753,444 cal"

...should have been:

180,000 joules / 4.1858 joule per cal = 43,002 cal

...which means it takes about 15 minutes before the ice even melts.

(much better)

Based on this I should get home with water that is near the boiling point, but not yet actually boiling.

20,000 cal to melt

25,000 cal more to reach boiling

135,000 cal more to actually boil it all off

----------------------------------

So 250 ml looks like a good size after all?

It would take about half a liter to do the whole thing as ice all the way...

.

 

[Fuzzo]

Trying it is the only way to find out for sure. Then comparing it with a scoop. It surely takes an awful lot of cals to heat up the thousands of litres of air you can pour across the engine in 7 mins at full throttle.

With your ice cooling, the iciness of the water only adds a bit to the cooling. Room temperature water would also have quite an effect, and you can get that pretty much any time. Then you don't have to store your pack in the freezer, indeed, it could be a permanently attached metal tank, and you just fill it with a bottle.

 

[safe]

I like to do both. Sometimes in reality things turn out far different than you expect and that can be good or bad. Maybe the fiberglass acts like an insulator to such a degree that I get moderate cooling for the whole half hour and that might actually be better than if it was super cooled for the first 15 minutes and then became hot.

Only time and testing will tell...

 

[Fuzzo]

A side note: Having a tank of boiling water venting steam on a bike does require some safety precautions. You wouldn't want to be able to accidentally touch the tank with your skin, and you certainly don't want anything impeding the steam coming out - if your engine gets really hot, you want to be ****ed sure that the vent is big enough to vent the steam fast enough, otherwise the tank could explode. On your bike the motor is right below the seat, so this tank of boiling water is going to be right between your legs.

How hot does your motor get?

 

[Fuzzo]

If you just use water, rather than ice, the only limit to how long you can go on full throttle is how much water you can carry. It's a pretty easy substance to acquire along the way, too. Has the added advantage that if you have to stop for a while, it will sink back to ambient temperature and be good to go straight away.

But it's way less tech-cool than ice.