More Test Results

[Boondocker]

Here are some more graphs of test results which includes stator temperature.   Previously, I wrote that the efficiency of the alternator seem too good to be true.   Come to find out that the rpm read out from the drive controller was off when the alternator was loaded.  The rpm was 6% faster than the drive controller  readout compared to when I clamped an AC lead and calculated rpm based on frequency.  

Also, found out that this little alternator can really tax the 3 hp rated motor.   It is rated for 3 hp at 880 rpm.   For continuous operations, about 425 rpm is the best the motor and drive combination can do without kicking out on overload.   Was able to get one series test done at 445 rpm after resetting drive a few times.

Playing around with the boost converter, one of the interesting things is if I set the boost voltage high, the over all efficiency stays up until the boost is surpassed.   Maybe someone with the electrical background can explain why.    Granted, I would not set the boost this way because it would stall out.

Other observation:  I'm amazed on how much heat is given by the rectifiers when some sustained amperage pass through them.  What a shame to lose that captured electricity to heat.  

[SamoaPower]

Boondocker,

You derserve credit for some of the more meaningful work presented on this board pertaining to the axial flux alternator. In particular, the temperature data is most revealing and gives us an insight to the problems with the cast stator.

In looking at your data, one piece seems to be missing - the ambient air temperature. I assumed 70F in my analysis. This can be corrected for if significantly different.

Using your three power/temperature charts, the temperature rise over ambient gives gradients of 0.21, 0.24 and 0.34 degrees F per watt of stator dissipation. I used 0.3F/Watt in the following chart for the higher power levels.

The estimate for coil core temperatues includes a factor of 104F (40C) to account for the thermal resistance of the casting resin. The derivation for this is found here:

http://www.fieldlines.com/comments/2006/11/26/17291/409/27#27

Assuming 200C wire insulation rating, that estimated core temperature point is reached at 1kW output (760 stator watts).

Perhaps a more significant question is what happens to the casting structural integrity at those temperatures. Most cases of reported stator failures include reports of stator contact with the magnet rotors.

I would suggest a maximum safe sustained operating power level of 800W for your machine. This allows for higher ambient temeratures and a safety factor. It should be pointed out that this could be increased by 25% to 1kW if a non-cast stator was used.

I often see said here that the stator will run cooler when installed on a tower and exposed to the wind. In my opinion, this will not be a significant factor because the air flow is ill-defined in that area and the shielding by the air rotor hub will prevent direct wind from reaching the important parts of the stator. I wouldn't count on it for much benefit.

Yes indeed, rectifier losses are a significant issue. This is why I want to use synchronous rectifiers with less than a tenth the loss.

One final question. Were these latest tests done into a resistive DC load, which you have shown, or into a battery bank? If you really want to see the alternator efficiency itself, better to measure at AC into a resistive load. Of course the final end results is better shown with rectifiers into a battery bank or better yet, into a constant voltage battery simulator.

Many thanks for your testing and reporting. We need to see more of this done here.

[Boondocker]

SamoaPower,

Thanks. It was not my original intention to investigate stator heat.  However it has been very educational.  I have a new appreciation the influence heat has on generating electricity.  Heat is the nemesis

Neglected to measure the ambient air temperature, I would guess between 75 and 80 F.

The system voltage was controlled with a diversion controller also connected to a battery bank.  The voltage was clamped by the diversion controller at 28 vdc.  During the testing, the diversion controller was dumping energy to the resistors.

Initially, was most interested in overall system efficiency.  Did not consider measuring alternator efficiency separately.  I need to ask a basic question; when measuring the amperage of one of three phases, do you multiply that value by 1.73 for the total output?

[oztules]

Boondocker, a great many thanks for presenting this series of results.

To me it confirms my fears with the current in-vogue axial flux design.

I have deliberatly held back from putting my coils and resin to work because I had no hard evidence of the thermal mismatch between the coiltemp, and the resin casting surface and the atmosphere (heatsink), and have been scratching around to find a better way.... don't have a better way at this stage, so still feel hostage to the resin stator idea.

Your work has provided some definitive research into this. So far the resin stator is the best allround system (ease of manufacture, availability of materials, robust selfcontained stator etc etc.), but I am still very concerned with heating in 1kw and over. In my wind condition, this >1kw will be the norm rather than the unusual.

Thanks for your efforts in this.

.......oztules

[DanB]

This is great stuff - thank you for sharing!

I've gone through all your diaries again and just can't find it...  I think you've got 80 turns w/two strands of #14 gage wire in the coils - and is it wired in Star?  What magnets are you using?

[Boondocker]

Thanks,

The data is getting some interesting attention.

The magnets are 1/4" wedge obtained from Windstuffnow mounted on 12" rotor.  The coils are 51 turns of No 15 wire, 2 in-hand.  I made an error in a previous diary entry about the wire (No. 15 not No. 14) and made a post stating the correction.  The stator is wired in star configuration.

[SamoaPower]

Boondocker,

Sorry for the delay - got wound up in the other post. I hope you don't mind that I used some of your data. It's the first temperature data I've seen presented here and think it opened a few eyes. Again, many thanks.

I think the way to measure the alternator efficiency alone is to use a balanced three phase resistive load also connected in star. Assuming system balance, measuring one line current will give the total power by: Pt = 3(Irms^2*R) where R is one resistor of the load star.

Don't be too surprised if this gives you much better efficiency than the DC battery case. It points out the problem with battery charging from an AC source including those pesky rectifier losses.

Keep up the good work - you're learning things.