Continuing the discussion from Pacificmeister Build Info and CAD Sources:
In this thread I would like to summarize solutions to the problem of coupling the motor or gearbox shaft with the propeller shaft. If you have additional bearings after your motor/gearbox your bearing setup is technically overdetermined. It’s not great to have an overdetermined bearing setup as it will cause excessive wear in bearings and can even cause some wobble which leads to inferior seal performance or even binding. I think with the 3d printed bearing mounts on the forum here they get away with it because the 3d printed housing is not too stiff and absorbs some minor misalignments.
Another solution could be to just extend the shaft without additional bearings. This results in a long shaft with high bending forces and thus high forces on the bearings in the drive element.
What type of coupling do you recommend or use?
I am using the spiral ones on my CNC. They also support radial missallignment. They are very strong so far. The steppers exert about 3 Nm of torque and they obviously change direction a lot which can cause fatigue. So far none have failed. I don’t know how they behave at high speeds though.
How much torque and rpms do you estimate for the time of failure? Which material is yours?
I did use this aluminium spiral coupling from ebay in my underpowered jet pump last year with about 7000rpm@1kw input power (that should be 1.4 Nm). But this did not have much shock loading, exept power rampups from idle to full power in less than a second.
It can’t work. Those are rated very low torque and rpm, furthemore you need to handle the thrust component (prop pushing toward the gearbox). I machined a 22mm diameter bar the lathe to build a solid coupler (the commercial I got were not right… Or too bulky.) the coupler should pushh on a thrust bearing (as much as possible) to avoid stress to gearbox.
You are right, thouse couplings are not designed to transmit axial loads. My design was a propshaft with two ball bearings to fix it radially and a thrust bearing to capture the axial forces. This shaft I want to couple with the motor. How did you solve overdetermining your setup with a solid coupling?
I imagine you did not need an extra radial bearing as the gearbox output is quite stiff. Thanks for your reply, if there is no problem with a solid coupling I will also machine one.
Is anyone using a flexible coupling? I’m planning on using a jaw style coupler similar to this.
The gearbox shaft will go into one side of coupler, the prop shaft will be in the other end with one ceramic bearing out near the shaft as per @pacificmeister design.
I don’t see the need for a second bearing on the prop shaft or am I wrong?
What is about using a Kardan? It should handle high torques and can be used for two directions. A jaw spider coupling will only work in forward direction. Otherwise it will dissamble itself.
Not tested yet but in my opinion a good solution is the spider coupling. It is a need to add scrub screws for a non slipping shaft.! I tried 4 different
kinds of couplings before.
They are the standard on most CNCs and they work great. The load on the grub screws is huge though. All the force is concentrated on the screw tip. I like the clamping collar style more.
Do you have 2x ball bearings (with some distance between) behind the jaw coupling? The dual setup with distance is important to prevent leaks in the shaft seal, due to wobbeling in the shaft.
A notch on the shaft between the two bearings can take thrust into the deep groove bearing (they can take quite a bit with regular service life).
See attached cross section on how we solved it with a jaw coupling.
Nikolai
Was the reason for doing this because the jaw coupling can’t take the axial thrust loading or just to remove the thrust bearing? Is a simple solution if you can get the shaft machined and the bearing can take axial load.
The jaw coupling is not made to take axial loads (at least we think so:p). So yes that was one of the reasons. And it will also take away the need for a thrust bearing.
We calculated the bearings, it should be just fine.
