Looks awesome!
Why not set this threaded insert into the fork in the other direction?
And related (?) question…
Why is this part shaped so funny?
According to Using Heat Set Inserts
Instead of putting the inserts on the face of your part flush with whatever you’re bolting to, put them on the other side of the part instead. This vastly improves the connection strength because when the insert is loaded, the taper of the insert meshes with the taper of the counterbore to distribute forces applied to the bolt.
EDIT: I misinterpreted your drawing and thought you were suggesting the opposite, but will leave this here to show why one might want to do what you suggested
We use the insert I have in a lot of applications, including when we have blind holes and the version you illustrated is not available. So it is easier to try and stick to one style of insert. These are press-fit inserts, they work similar to concrete anchors where they expand into the plastic when you screw into them. We have used this style insert a lot over the past year and have been really happy with them. Easy to install and super strong.
This spacer is rounded to allow the wheel to be removed. Check this cross section of the pod below:
In this picture, if you try to drop the wheel straight down out of the module, the rim of the wheel would run into the bevel gear. By rounding the spacer on the right side, and rounding the bottom edge of the fork on the left side, we are able to arc the wheel out the bottom and to the left to be able to remove the wheel without taking the fork off.
15 Likes
This is brilliant!
I also had someone ask me on Discord how I was mounting the pulley so close to the face of the UltraPlanetary on the flipped version of the module I just posted today.
The pic below shows a cross section of the Neo550 and UP and the pulley:
I am replacing the part of the UP output that actually spins and has the 5mm hex female bore with a 3d printed part. This part is able to be mounted the same way the existing output UP part does, but then has the 16T pulley integrated as part of it. This way I can have the pulley mounted to the UP output without needing space for screws etc. There is a 3/8" diameter boss at the end of the printed part that picks up a bearing so that the pulley is supported on the side opposite the UP. Below is a picture of this printed part.
9 Likes
I really like this. There’s lots of potential with custom UP outputs.
I would only be a little worried about the torque strength of the small 3DP female spline, but it’s probably OK? It’s good you have a second bearing instead of cantilevering that pulley. You might even consider stuffing in an aluminum spacer through the middle to make it extra skookum.
Why this extra bit here? It obviously makes it so you can put the NEO 550 on both sides but why not just manufacture 2 versions, removing the redundant area? It seems that would save a decent amount of weight.
If you have two versions, you need twice the number of spares.
8 Likes
Personal experience says you may have problems with the pulley. We had several fail on our robot with similar specifications. They were directly on the output shaft of a falcon driving an intake and another instance was a NEO driving our climber.
What about the application here is similar to using a pulley directly on a motor shaft in a much higher torque application?
2 points:
- I don’t know that you will have an issue, I’m just recommending that you have a back-up plan.
- The power output is much different, but not necessarily the torque.
You are driving that 550 after a 5:1 reduction. The Falcon (probably the more similar use-case on an intake roller), has a stall torque of about 4.69 N m. The 550 has a stall torque of about 0.97 N m. At the pulley, and after your reduction the teeth of the pulley could experience 5.07 N*m. Because the 550 requires so much less current (100A) its more feasible to actually stall it than a Falcon. Now, because you have the torque necessary to turn the modules, the pulley will never see these loads and will be de-rated by some amount but i’m not going to guess at what that could be.
To keep our intake pulley from Cooking it’s self we had to impose both a heavy voltage limit AND a good current limit until we were able to switch it to one of the aluminum pulleys. If you have experience using these and they have been fine then please teach me what I should do to make ours last. And yes, we are using Onyx filament.
Maybe the big game changer is that you could include fiber in the pulley and give it additional strength or that you dont have a shaft running though the middle of the part. In cases where we have had more “meat” between the pulley and the bore, we have had no issues. Those have been above 30 teeth if I remember correctly
This particular design doesn’t lend to fiber doing much at all, as the spline teeth are too small and interlayer bonds might even be weakened depending on the fiber.
What was the failure mode? Stripping the female Falcon spline in the pulley?
We had a few direct driven pulleys off Falcons on prototypes and the final robot. We used a 12T pinion gear as a larger mating spline, no failures.
1 Like
Teeth Yielded → pulley slipped relative to the belt → pulley melted.
I don’t have a photo of the Falcon/Intake Failure but I do have a photo of the Neo/Climber:
The intake system showed that yielding was the first level of failure with the melting being the second. In once case where we had it happen on the practice field, we managed to stop before everything was melted and we could see teeth rolled over. With the motor shafts in place, there were not many wall thicknesses that made up the pulley teeth and that may have contributed.
Because I don’t know exactly where the limit for a 16 tooth onyx pulley is, I thought it would be good to say something.
Dan,
Did you witness the teeth yielding? What I’m getting at is it possible that the belt was slipping/jumping profusely which then turned into melting seconds later.
Side note: We found that reducing the tooth profile by .002" makes all the difference in how well the belt grabs the pulley (reducing might not be the right CAD term).
We were using the unmodified pulley geometry from MKCad’s generator (no adjustment) and the ideal center spacing. We certainly could have had jumping cause the yielding, but the tensions seemed appropriate in both cases.
What infill % and how many wall layers were used? Also what was the tooth count on the pulley? Curious as I plan to experiment some with 3D printed Onyx pulleys.
Lets take this conversation out of this thread. I’ll PM you and show you some of our print configurations
1 Like
Any chance this could get split into a separate thread (If you’re fine with publicly sharing that information)? Many of us print pulleys and having more information on failure modes is always helpful.
Moved to this thread here:
1 Like