Experimental determination of transfer coefficient

[mrpi]

First a little background:

I designed a billet aluminum forced-air cooled brushless permanent magnet electric motor housing (finned) and I also designed a sheet metal shroud to direct cooling air around the motor housing. I did hand calculations to verify that it had sufficient cooling capacity, and we also had CFD analysis performed on the shroud/motor housing as a double check (the results agreed very closely).

We sold this product to our customer and the last I heard they no longer needed us to provide the cooling shroud along with the motor. However, they did ask for the engineering drawing and solid CAD model of the shroud.

I figured that they just wanted to make the sheetmetal shroud themselves to reduce cost or whatever and didnt think twice about it.

Well, we got a motor back for RMA and the shroud is totally different. Where I had a diffuser plenum to distribute the cooling stream over the fins, the customer re-design put the inlet pipe tight against the housing!! Without the diffuser plenum, this only exposes a small number of the fins to the cooling stream! I calculated thermal resistance to be 10x greater as a result of the lost fin area and the insulating effect of the "closed off" fins.

While we haven't seen any in-field failures of motor due to overheating I'm concerned that it could happen. We have no thermal protection of the motor stator in place to monitor temperature. The highest motor loading only occurs at 99th percentile of cases, so its quite rare which is probably why we haven't seen any problems.

So, now my question: Since I have one of the cooling shrouds in my hands and I have motor housings available, I I would like to see how bad the cooling ability is. What exactly is it that I'm trying to measure? Convection transfer coefficient?

Can this be measured by seeing how much heat can be transferred INTO the motor housing? I.E. I would put a hair dryer on the shroud inlet and measure the temperature rise of the motor, the flow rate of the air, and the power draw of the hair dryer heating element? I'm not sure of all the thermal properties I need to determine.

Do I need to determine the heat capacity of the motor housing first? Do I need to preserve the direction of heat flow: put a known volume of hot water into the motor housing and blow air into the shroud, measure the air flow rate, and observe the rate of water temp change? Would I need to compare this to the control case in which I dont blow air into the shroud? (This might be the most direct measure of the rate at which the air can remove heat?)

Any suggestions on how I could go about this in as simple a manner as possible would be greatly appreciated.

Beat to fit, paint to match.

 

[CH5OH]

ultimately, provide cooling on a motor is to get rid of the heat, mainly caused by the resistance of the electric coil
so if the motor can be loaded by a constant load (generator ,pump,whatever) the electric load can be measured using a kW meter.When running the motor on full load,starting from same temperature (ambient)for a same duration (lets say an hour), when measuring the resistance with an accurate meter (milliohms, wheatstone bridge), you obtain three values:
resistance ambient
resistance cooling option 1
resistance cooling option 2
with this you should be able to work out the coil temperatures (the degradation mechanism of the motor)

 

[mrpi]

The problem is fully loading the motor requires absorbing 8kW+, which is beyond the capacity of our test equipment.

CH50H, did you mean measuring the resistance of the stator winding to determine its internal temperature?

I just found out that the motor includes a thermistor in the windings, however electrical noise during operation makes it un-readable and we do not record it with our controller. I'm going to see if they can determine the motor temperature in the field.

Beat to fit, paint to match.

 

[CH5OH]

yeah, determining resistance and calculate the temperature
if it has a thermistor and you have its characteristics,
just stop the motor to perform a measurement.though I am confused not to being able to use the thermistor when motor is running.thermistor uses DC current, may be filter out the DC current by shortening the AC noice (capacitor to ground)

 

[mrpi]

Ok, it looks like our customer is going to re-design the shroud to be like the original configuration rather than pay for testing on their re-designed shroud. Problem solved!

Beat to fit, paint to match.