Turns out Helical Beam Couplers aren't meant to handle that much stress . . .

Or should I say, this particular coupler from China isn’t meant to handle so much stress.

We have been trying to dampen some of the vibration caused by a huge lead screw that we use to raise and lower our ball holder. Mounting with grommets and the like helped some, but still the vibration was intense and the noise was horrendous. So I had the bright idea to get some Helical Couplers and see if they couldn’t absorb some of the movement.

Unfortunately, I was wrong:
https://drive.google.com/file/d/0B5aY7IIZS5TERk0xdTRhT05xeFE/edit?usp=sharing

You are using these incorrectly.

You are relying on them to provide bearing support to the leadscrew (via the CIMs bearings) and they are not designed to do that. The end of the leadscrew should be in bearings properly in some structure, the CIM bolting to that, and then coupled with the beam coupler. The helical portion of the coupler then just handling shaft misalignment. They’re also not really good at dampening vibration as they are intended to be stiff.

As Gregor mentioned, yes we are aware that lead screws operate less than optimally when they are not supported on both ends. It performed exactly as intended with our original coupler in competition though and the only issue we ever had when when the lead screw got a little bent. It still worked well, the vibration just drastically increased.
Unfortunately, our team does not have a Mechanical Engineer of any sort. Us computer guys learn as much as we can about mechanical operation and actuators, but the finer points are sometimes missed. While it never actually failed, it could certainly be improved, so thanks for the input.

They’re also not really good at dampening vibration as they are intended to be stiff.

Maybe they are intended to be stiff, but they are bendable by hand. If it was as stiff as I expected maybe it would not have deformed as much and taken the vibration a little better?

Generally in industrial applications the lead screw will be turned down at each end so that it can sit on a thrust bearing package. In my plant, we then drive them with a timing belt.

An easy way to do that is to get a shaft collar that fits over the threads and pin it in place. Then you don’t have to machine down the end.

http://img.directindustry.com/images_di/photo-g/ball-screw-support-bearing-units-8884-4327425.jpg

http://gallery.nibou.com/albums/LINEAR%20MOTION%20PRODUCTS/CYRUS/Ballscrew/BallScrew_Supports.jpg

Your system is putting a huge amount of axial load into the motor shaft directly. They aren’t made to support that type of load.

For FRC applications, we had success with using a coupling nut (http://www.mcmaster.com/#93023a661/=rw83fu) welded to the lead screw. The 1/2" hex interfaces well with a lot of FRC components, making it very easy to drive. We supported the lead screw on each end with thrust and ball bearings. Our lead screw was ~18" long and flexed a bit since we didn’t load it evenly, but the setup did a good enough job of isolating the screw from everything else to be very reliable.

We used this in 2010 on our kicking mechanism.

Those couplers are also not able to manage that much misalignment.

The interesting challenge with building a support system for the leadscrew is that, because the leadscrew does not stay stationary as the ball holder is raised, the support system would have to be mounted only to the motor and the leadscrew itself. It might make a good off-season project.

Yeah, we use with one very similar to that, a little shorter and different diameter and TPI. We were bounded mostly by funding and really wanted to avoid paying $14 for the nut when the leadscrew itself was so expensive.

Right, it wasn’t intended to. Like I said above, I added this in hopes that it would absorb some of the vibration and noise, not to make the system perfect.

Where does you team purchase beam couplings? I had the impression that helical beam couplings generally ran >$25, and I’d love to know where you can get them for less.

We got them for $6 from China on eBay. Shipping takes about a month (and is free), but as long as you plan ahead properly and commit to buying early then you can save a huge amount of money. In fact, I don’t think a single fastener on our robot came from outside of China (aside from KOP and Vex ones obviously). Paying Lowes or Fastenal prices for anything other than stock metal just wasn’t in the budget for us.

I’m interested to know the quality difference between the ones we got and a $45+ coupler from McMaster.

You can keep the leadscrew stationary and use a simple linkage to translate the linear motion of the lead screw to a rotational motion.







These beam couplings are really good for coupling rotational sensors with drive shafts so that small misalignments won’t damage the sensor. we use them in our swerve drives to couple the Magnepot Hall Effect Potentiometer which provides us with drive angle measurements for each of our wheels. These couplings are absolutely wonderful and work flawlessly in this application. A photo of the installed coupling is provided:

On the other hand, they won’t hold up to very high torque (use Lovejoy couplings for this), massive misalignments or huge angular displacements (use flex joints here). They also cannot be used as the device which keeps everything aligned by itself.

The movie was entertaining, though.

Oh wow, that’s clever. Thanks for sharing that.

That’s really awesome. Did you end up using that mechanism in your robot this year? I wish I had seen some like that at competition.

Yes, that entire thing was mounted onto the robot.

https://www.youtube.com/watch?v=N_gfnubBvvU

Also here is test of the tilt too.

I know you can’t see it in the video but your coupler got into a resonant mode where the parts were vibrating independent of the lead screw. When the resonance hit the right part of the coupler, the stresses increase significantly. That is why you had parts of the coupler deform more than other parts.

We used a 5/16-5/16 flex coupling on part of our winch this year. In the right application, they are very good. We purchased from McMaster-Carr.

Below is a link to the McMaster part that looks similar to what you are using. it has 5/16" and 1/2" bores. This part is rated for 0.015" parallel offset and 5 degrees angular offset. It sells for $55.

http://www.mcmaster.com/#6208k596/=ryvr6u

Another strong point to make is the connection directly to the CIM. The CIM “bearing” is just a bronze bushing and cannot take axial loading. The motor and connection really needs to be mounted in a way that limits the axial load on the shaft.

Your initial question was if the part from China was as good as a part from other suppliers. I have an opinion, but i don’t think this test can validate yes or no because the part was used well beyond what it is designed for.

Hun interesting. I just assumed that parts were more deformed because it was able to bend such a large angle and a certain ring section failed before the others.

After the fact we looked up the length of exposed threaded rod that would lead to resonance at our diameter and found out that we were WAY over the limit. This mechanism was nick named the “Garbage Disposal” shortly after due to the noise.

Yeah, the McMaster one would have to be significantly stronger in order to make it usable for this application. I’m thinking that any helical coupler like this just isn’t going to get the job done. We will have to stick with our giant solid piece of aluminum that we used during the season.

Brian,
I tell people all the time to listen to the robot. It talks to you, you just have to know what it is saying.

You really need a rigid coupling there since you are transmitting a bending moment through it. (Like you giant solid piece of aluminum?). Flexible couplings are meant to transmit torque, not to hold things in alignment.

To help with the resonant frequency and the vibration, cut off the extra length of screw since it is mass whipping around. Use a a speed controller on the motor to slow the screw down. The more it extend, the faster it can go.