Our team have saw many incredible telescoping arm and elevator during 2023 season, and we like the lower weight of telescoping arm although it would be more difficult to be built.
However, we can’t find the methos to rotate the telescoping arm by a shift such as 1538, only understanding the measure like 2910 did.
I would appreciate it if anyone could share their experience.
There are basically two systems by which an arm can be rotated on a center pivot, like 1538 (or my own team) did.
One is with a live axle, which is what 1538 seems to have had, judging by the videos of their robot. Essentially, the axle is fixed to the arm and rotates in bushings or bearings in the mounts on either side. 1538’s arm is typical of this type and this does let you do what they did, which is align the axle with the center of the arm tube. If you watch their highlights video closely, you can see that the axle is fixed to a big set of clamping blocks that surround and grip the arm, so it’s really two half-axles, one on each side of the arm. Then there’s a chain sprocket attached to the axle that runs down to a motor/gearbox on the base of the robot to drive the arm rotation.
The other way to rotate an arm is by using a dead axle system, which is what my team did this year. The axle itself is fixed to the supports, then the arm mounts have bushings or bearings that allow it to rotate on the axle. The chain sprocket is attached to the arm mounts and driven the same way as above. The axle extends across the robot, but this means that the arm must rotate off-center on the axle, since passing it through the arm would obviously interfere with the arm’s own operation. You can see this type of arm on our robot in the picture below (note the axle extends all the way out one side to connect to the camera boom):
If you look closely, you can see one of the bronze bushings that the arm rotates on just to the left of the visible chain sprocket. The axle itself is a 7/8" aluminum tube from McMaster-Carr, with flanged bushings to fit (which also happen to fit nicely into the interior holes of the chain sprockets) and machined aluminum mounts. Here’s a picture of Charles just after it dropped a cone on the high grid but before the arm retracts. You can see how the arm is offset from the axle:
Hope this helps explain things a bit.
You can also consider a method that does not use a chain + sprocket (or belt + pulley) for rotation of the telescoping arm. 1756, 6919, and our team went with the linear actuator style approach where you mount one linear actuator underneath the telescoping arm you want to rotate and you achieve a limited in range rotation around a dead axle pivot. While not the lightest weight method, it was successful for us and basically just involved building 2 of the same style telescoping systems.
We lifted our telescope arm using an elevator and some surgical tubing. 3970 Robot The telescope was on a dead axle in the back for pivot. The elevator had rollers which rolled underneath the arm when the elevator went up.
Thanks!
However, is there any problem about telescope heavier than axle can support because it is two half instead of one?
Wow, It is amazing, and I have not heard this method before. Nevertheless, is it hard to program if the angle of two linear actuator are different (for rising and failing telescope angle)?
It was very simple to program.
Really? Does it need to use a trigonometric functions, doesn’t it?
If you’re going to try and do it perfectly it would require use of inverse kinematics.
We simply used a collision avoidance threshold value for the stow, and then the telescoping arm was limited to only traveling once the rotation arm was near it’s final setpoint. It was rudimentary but it worked fine.
That depends on how you make the axle and arm mounts for it. The one on 1538’s robot is pretty hefty (it looks like 3/4" rod at least) and the mounts are massive. As in all things with engineering, it’s best to overbuild rather than underbuild something like this. If the mounts are of sufficient size and the connection to the axle is solid, there shouldn’t be a problem. 1538’s mounts look to be fairly thick at the axle attachment point, maybe with two separate plates with space between them, which would give enough support to the axle to take a lot of force. As an example of that, they flip over at one point in their highlights video and use the arm to flip themselves back upright. Not something they could do without a very strong and secure joint between the axle and mounting plates.
If you don’t mind, I still have a question.
You use a circle tube instead of hex shift, and which shift clamping shift collar do you use?
Can I also know the clamping blocks you use? Is that made by yourself or you buy?
This was a relatively common design this year that was quite successful.
4028, and 1756 looked very similar using a linear actuators to pivot and another to telescope.
Okay, to answer both your questions I had to get back to the robot and take some closeup pics. I’ll use those to illustrate how we did this.
For the axle, we don’t use shaft collars at all. Instead, we bolt through the axle at each of the three points where it goes through a support. That secures it in place. Everything that rotates on the axle is spaced out by either the bushings or by 3D printed spacers between those. There’s also an encoder pulley on the camera mast side that turns a belt to a lower pulley and shaft for the cancoder that monitors the arm position.
The mounting plates are machined on our CNC (an Omio X8) and have through holes for the axle bushings and for the hex shaft that attaches the chain sprockets. There is also a slightly recessed center section that allows it to wrap around the arm on each side to hold it securely and prevent the arm from twisting in the mounts. The plates are then secured to the arm itself by four very short bolts on each side that go into threaded holes in the arm The hex shaft and bolts front and back take most of the force and the short bolts into the arm make sure nothing moves around or slides up and down the arm.
You can see from that third picture how the arm/bracket/sprocket assembly rides on the axle using the bronze bushings. It’s a remarkably solid and secure system. We actually had an accident happen this year where our end effector got stuck in the single loading station chute and we lifted the whole unit off the ground, but it didn’t damage the arm at all (or the loading station either, fortunately.)
I hope this helps answer some of your questions.
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.
