C80-100 outrunner - Shaft keeps losening

Has anyone else had issues with the shaft loosening on large outrunners?
I have a C80-100 on a road legal moped. (link here) I have used this motor on a few bikes now and I always seem to have trouble with the locking grub screws fretting/coming loose, I have highlighted the offending item in this picture:


There are 2 grub screws, one either side. For this latest build I decided to change the shaft design.The original motor shaft has machined flats, these appeared to fret the end on the grubsrew and caused slack over time (I tried cone and flat tipped grubscrews, both wore). For the modified design I machined detents into the shaft. Bottom shaft is original, top shaft is my modified one:


I figured cone tipped grub screws would keep this puppy secure. Well, guess what - 120 miles later and I have some slop back. I use loctite 243 on the shaft to endcap interface and also for the grubscrews.

Just thought I would ask if this is a common problem with any known solution? .... Please

- Spata

i had the same prob but used an angle grinder and put 2 flat spots on it and it seems ok now.

Use green loctite

xadmx said:

i had the same prob but used an angle grinder and put 2 flat spots on it and it seems ok now.

Click to expand...

What you mean like the original shaft? (bottom shaft in the last picture)

riba2233 said:

Use green loctite

Click to expand...

Thought about using stronger loctite, but then I would need lots of heat when I eventually want to remove it.
I was worried about putting heat so close to the magnets (which are only glued on in the first place). Has anyone got any experience? is this ok?

- Spata

Yes I have the same problem on all my outrunners. The stator is also coming loose under hard acceleration.

Bobc makes really nice taper lock cogs which he puts on rc motors, no reason a taper lock flange couldn't work with the four screw holes in the end of the bell to get a bit more purchase on the axle. hope this helps spata.

spata said:

xadmx said:

i had the same prob but used an angle grinder and put 2 flat spots on it and it seems ok now.

Click to expand...

What you mean like the original shaft? (bottom shaft in the last picture)

riba2233 said:

Use green loctite

Click to expand...

Thought about using stronger loctite, but then I would need lots of heat when I eventually want to remove it.
I was worried about putting heat so close to the magnets (which are only glued on in the first place). Has anyone got any experience? is this ok?

- Spata

Click to expand...

yes i found that the flat spots on the shaft didn't line up with the grub screws so i ground some on in line with the grub screws.
but you could try taking out the grub screws and try some small m4 allen heads so you can get more bite on them.

Ok, so I did some digging around to try and work out why these shafts are coming loose. First an estimate of the motor torque. From this thread

Jeremy Harris said:

Luckily, there is a fixed relationship (for a PM motor) between the voltage constant, Kv, and the torque constant, Kt. If you know Kv you can work out Kt fairly easily. Here is the formula, for Kt in N-m per amp:

Kt (N-m/amp) = 9.5478 / Kv

so, if you have a PM motor with a Kv of 50, then it's Kt = 9.5478 / 50 = 0.191 N-m/amp

Click to expand...

So for my 130Kv motor, Kt = 0.073 N.m/Amp. At 150Amp (phase current) this gives a torque of 10.95Nm

Remember that my shaft was held with loctite 243 and the set screws, ignoring the setscrews for now and using the loctite spec we get a compression sheer strength of 7.6N/sqmm.

Now the Shaft radius of the C80-100 motor is 6mm (12mm shaft) so working out the force at the shaft we get 1000/6 * 10.95 = 1820N force =1.82kN

The shaft circumference is 12 * 3.142 = 37.7mm shaft circumference
So to work out the width of Loctite 243 we need to hold that force we get 1820/(37.7 * 7.6) = 6.35mm width of loctite 243

So by my reckoning a bead of 7mm of Loctite 243 should hold the force by itself.... without any set screws. The actual width for locking available on the shaft is around 20mm, and I think it is fair to assume that I got loctite on at least 10mm.

Now.. either my calculations are junk (most probable ).
OR My loctite application was bad (everything was clean and seemed ok).
OR The shock loads are way higher than I have calculated.
OR The controller (lyen 18 fet) is giving the motor much more phase amps than expected and hence the motor torque is higher.

Thoughts?
- Spata

when I did mine, I just drilled just shy of halfway through it. that way the screw would tighten, but it was basically a pin, instead of a grub screw.
Keep in mind that hardend steel shafts only have the outer layer hardened, so all your flatspots are going into soft untreated steel. If you can have someone case harden the shaft after you put flats on it, they wont wear out.