OwenB

This is a question, not a "How-to" - ? should be in the title.

I read the "Fly Electric" pages on DIY brushless motors, by David Theunissen, and I would like to find out more:
1) what are the "hot" cores for 550 and 2000 watt high speed motors currently?
2) how do you wind them to get the correct KV amps and watts?
--------------------------------------------------------------------------------------
I want to use 48 volts for now.
I would suppose a 550 w motor would need a 75mm drum, and a 2000 w motor would need about a 120mm drum.
Is DT still contactable?
Who knows the theory, or can point me to simple winding designs?



No-one seems to make motors for these 2 cases:

8x5 prop at 23,000 rpm, and 14x6 prop at 14,000 rpm.
This is using composite sports props.

If you have seen commercial motors that fit the specs below, please let me know.

I have a current application for the 8x5 props-using 2 of them on my current model, to get a vertical launch capability- and rather ballistic performance
requiring gyro nannies! Span is 44 inches and weight is 1.6 kgs planned. This is a belly launch VTOL, no undercarriage needed.
Good for a first RC plane! It will have all the learner nannies turned on (Ardupilot) and control sensitivity way down...

I had to shelve my current motors after I found out how to create dyno charts for them-they were just too small (BH 8208 1900KV)
-245 actual w -rated vs 550 w needed.

I had to derive a rating scheme based on W and A limits, plus their "ragged edge" benchmark test, plus motor similarity considerations.

The 14x5 is for a future proposed FW 190 lookalike stunt plane, also needing gyro stabilisation, otherwise you get lots of unplanned loops when you gun the motor!
Span is 50 inches, and it is based on 1:7.65 scale, with Lavochkin and Seversky addon mods, and a reduced wing aspect ratio.

This is to have undergear retracts, but no flaps. It should be practically impossible to stall on landing. It could also do a full vtol tail-dragger bellyflop landing (on grass!).

It should still do around 15 oz/sq ft loading. Weight may be over the 2 kg mark, due to the big motor and large batteries needed. It should generate over 5 kgs static thrust though.


I have drawn up suitable KV specs for each:


1) 8 inch x 5 inch propeller, 23,000 rpm.
This needs 16V, 1800KV, 550w, 34A rated spec.

2) 14 inch x 6 inch propeller, 14,000 rpm.
This needs 48V, 400KV, 2000w, 80A rated spec.

or 16v 1300KV 2000W , 240 A (based on 6x345w motors.
obviously, multiple batteries are needed there - maybe 3x 3000 mAh 4S 30C.
It shouldn't need to use full throttle much , as that gives 6000gf for a 2500g model.

Both motors are rated at 77% of the KV speed, and maximum load at 66% of the KV speed.

This loading scheme seems to be standard for model aeroplane brushless motors.


Cooling is from natural air flow from the prop end, from 20 m/s up to 44 m/s.

Thanks in advance for your replies.

<edit> - I have fund the sub-forum, thanks, and have a big list of resources to read. I will come back to you with more questions.

<edit>
Looking at 12S batteries-can only find 1: gens ace 5000 mAh 45C, 134.99 euro!!


Anyone heard of any 1800 KV 550 w motors?
<edit>
Here is one for consideration:
Spitz 3536-06 680w max, 50A max, 1300KV, props 8x4 and 9x6., 15V (4S)

1) that seems an odd prop selection? why not 8x5 or 9x5 ??

2) the max watts looks suspicious - usually max watts are 3 or more times the rated W.

3) the max amps also look suspicious - I estimated 6S 30 amps, max = 1.3 x normal = 39, x 24/16 = 58 amps.
now divide 50 /1.3 = 38 A
then expected w = 38 x 16 / 1.3 to allow for power factor-ac
=467w - too low - want 550w - what do you think of my calculations here?
<edit>
An alternative is tying multiple quadcopter motors together with small toothed belts. this gives an assembly 68mm wide, which is a little clunky?

-maybe suitable for the "radial" engine model. The assembly is about 120mm in diameter, and you can gang up 6 motors, giving 2000 w static load.
I can use proper 1300KV 4S quadcopter motors. It needs a center spindle with ballraces, drive collar, thread, and spinner nut.
I will spec out the little belt drives, and check the center hole diameter in the 14 inch prop.

This total assembly weight is likely to be about 400g.-comparable to any 2000w worth of 150 KV 6s motors, which are too slow!

This may not be a sensible way of getting to 550w, 1300kv, 4sS, though.
<edit>correction-50 m/s belt speed is OK, so 20,000 rpm at 1 inch dia is good.
<edit> 13-1-21 11:17 pm local-corrected the motor specs.

OwenB

Here is a drawing for the three motor array. This is probably easier for me than trying to rewind a different motor.
It is intended for a 9x7 two blade.
However, despite being a glass composite material, it is also an E-only prop with hollow-cambered blades, and I
am not sure whether it is rated for 22,000 rpm.
Otherwise I have a 3-bladed 9x6 prop lined up. - it has left-hand and right-hand versions, for a twin-prop model plane.
I will sort out loading for a 9x7 two-blade, and then de-rate the motor spec to aim at 100 A max , 50 A per 3-motor array... that may be a problem,
as these ones are 38 A each!
they are also 445 w or so each, but 200w each is probably enough.
Performance parameters desired are 105 mph, 2000 gf static thrust per prop.

OwenB

I have settled on 9x4 props, 1300 KV motors.
Peak amps seems to be 128 A per prop, up from benchmarks of 58A, 4S
I am not too sure why this is so-maybe the inefficiencies of 3 motors add up?

Best deal on motors is iFlight shop, Grumman Albatross cowls from Parkflyer plastics.
micro belt drive parts from SDP.

Here is a more developed version of the ganged motor array, including stiffeners and
air exit holes.
It is now due to be broken into layers and dimensioned up.

The layered detail is getting a bit undecipherable now.
There is no need for a fake engine front-this thing looks sufficiently impressive peeking through the cowl!

I think I will do better with this than trying to get a rewound motor to work.
I know these motors have high output , good power to weight, and good cooling.

The main development hurdle is getting the belt to run correctly. possibly 2 of the pullies need to be disassembled and rebuilt without side plates-if they come apart.
They are aluminium alloy, and look as though they are layered.

quorneng

That looks a very interesting layout but it does raise the question as to whether any 9x6 prop is good for the sort of power you are considering.
Even if you find a suitable prop the aerodynamic efficiency converting the shaft power into thrust would not be good at the rpm you mention, There will also be some additional mechanical losses of the extra bearings and the tooth belt.
Note also all the 8 bearing in your set up will be under a side load from the belt. A direct motor drive has only 2 bearings and no side load.
I have suspicion a belt ratio reduction (2:1?) would make more sense
It will be very interesting to see how it pans out.

OwenB

1) APC tables say the 9x5 is good for 16,000, and the power levels come straight from the tables. I suspect the prop tips may bend forward a bit under static load at 13,000 rpm.

Shaft power to thrust is included in the tables. - I have found that rpm levels are about 1.5 times what you would expect from basic theory for that thrust, with 2-bladed props. Power is not up 1.5 times, though.

2) I would like to find info on drag from belt and the 2 bearings.
I picked the wider ones for side and end loads, but the motor bearings will not last well with the side loading.
They do withstand quite a bit of out-of-balance (vibration) loading, but that is only milligrams.

3) The aluminium motor shafts could also snap off-due to fatigue.

There is a 20-odd kg rotating bend side load part of the time. (11 mm offset).

1 n-m acting at 7mm, = 142 N, or 14.2 kgf.

On a round 5mm shaft, this is 13.3 kPa, so not at all critical.


Putting the motor elements in a stronger housing may be required in the long run. We shall see.



4) Building a big bell-type quadcopter motor may be simpler.

I am sure doubling the diameter would make for a much more powerful and torquey motor- you double the circumference, and the radius, and the depth.

I need to find out about number of poles, number of turns, wire gauge , insulation grade, and source, etc, and a source for the barrel and stator core.

Each stator pole would be about 8 times the volume., but magnets only 4 x the area.

- making the motor over 4 x power would cause cooling problems, though.