Show me your Universal Joints

I think the most difficult part of my 6DoF Stewart platform so far was finding the perfect universal joint for the ends of my Linear Actuators.

I'm now on my V2 universal joint and I think it will be perfect. I designed them in Solidworks and had them manufactured in China. These are 7075 aluminium, sandblasted and anodised.

Version 1 worked ok, but there was a tiny bit of rattle and slop. This version is nice and solid.

This might be a good opportunity for others to show what solutions they are using for their Universal joints (U-Joint, Cardan Joint).

And how much did you pay for one pcs?

I paid $60usd plus my design time plus shipping, plus the trunnions bearing.

That's a lot of uj how big is your rig?

In terms of size for the 6DoF platform it is 1.4m diameter at the base. 6 actuators. Each actuator has a 490mm stroke and pushes 120kg at 500mm/sec.

They need to be strong.

Hi psionic ,
i find your u joint very interesting. I look also for this design for my 6 dof. Please can you give me the name of your manufacturer ? Or perhaps do you have the link of good seller for u joint ? I have the capacity to concept my u joint on solidworks but if someone have directly the right solution or seller, i will earn time in my project. Thanks for your advice.

Search Alibaba.com.
there are hundreds of companies that CNC machine manufacture aluminium parts.
cheers

Thanks for your answer

can you share your design? i want make ones

I can make some for you. Please PM me for costs.

Those things look gorgeous, is that 60usd for the 2 machined components for each joint? Or put another way, in your first picture for machined parts only, did you spend 60*12 = 720 or 60*24 = 1440?
Thanks for sharing.

From memory, It was $60 per piece x 24. and that was just for 6061 Aluminium. 7075 aluminium was a bit more expensive. plus there’s the trunion Cross bearing to insert.
I also paid a bit more to add a threaded hub to the end, and a frosted and anodised finish was included.
Expensive to have done but they work perfectly with zero slop.

for my seat mover i made it my self, used a alu square bar 50x50x30 made a screwhole on every side and bolted 4 12mm rodends upon the block. if you cant make screwholes this can be done with a thick steel square tube 4 holes and use bolts and nuts to secure the rodends also
regards Ad

Cool thanks for the information. They look absolutely beautiful. I like the size trunion bearings you used too. Very professional solution.

Hey

like this thread! I think U-joints are the unsung heros of motion cueing. Rarely ever do they get the attention they deserve.

I see people comparing motor power and actuation speed of their rigs, but I have yet to see anyone properly analysing their U-joints. They are a bit like the audio track of a movie: You never think much about it, but when something is not right, it becomes a real show stopper.

Also, I like the fact that yours run on ball bearings. To me that is an absolute must-have feature, but apparently we are a minority




Those are U-joints I designed a while ago, which run on 26mm ball bearings. Fully 3D printed and I would rate them for +/- 100Kgs (tested up to +/-400Kgs). The axles are 10mm shoulder bolts which are much more precise than ordinary hex screws. There is absolutely no play in them.




These are the upper U-joints I am currently using. Same strength rating.
I chose the single-sided design, because it allows for an insane range of motion (see comparison below) and many Level-D sims use it for that same reason.
I also included an internal triple axial bearing (28x12x8), because I wanted the actuator to be internally constrained to avoid a whole list of problems that externally constrained actuators would've caused.


Here's a comparison of ranges of motion:
Red: The absolute minimum range of motion I deem necessary. I would not even think about going below 30°/30°.
Blue: The older U-joints I showed. (with the red centerpiece)
Green: My current single sided U-joints.

Cost is 68€ per actuator including the internal axial bearing. So, all-in-all ~420€ for the entire rig.

...and did I mention they are absolutely rattle-free?!?

...I think I am starting to develop a fetish

Cheers,.... Dirty

@Dirty i know how you feel. I was so obsessed with getting the design right and couldn’t move forward in my rig until I was 110% satisfied that I had all the range of motion I would ever need and completely slip free. I also do a lot of 3D printing and all of my prototypes were 3D printed, but with the forces involved during power failures when the machine can slump and fall sideways, I wanted something that would withstand potentially huge amounts of bending and torque loads.

My original upper and lower UJoints were like the ones you have shown for your top joints using massive. Hardened shoulder bolts, however I used 2 sets of 45 degree thrust bearings to cover both load directions. I eventually moved away from them because of the eventual risk of full lockout and breakage if the actuator got into the wrong position.
having the traditional joint you have in the lower end obviously prevents the actuator spinning and is a great solution from what I can see. And offsetting that shaft on the top joint almost completely eliminates that arm lockout issue I mentioned for my ones above.
are they professionally SLS printed in PA12 or are you trusting FDM ABS? How did you test the strength?
Love the designs you’ve made and the angle range chart.

Did you eliminated any risk of locking, even with extreme range ?

can you show them « in action » as well as some design details ?

Hello 'Show me your Universal Joints' experts!

I'm from France and have built up an A320 6dof platform last year.

I need your expertise with 6dof joints as I face an issue with my upper rod ends heim joints that I'm not able to fixup to now.
My upper joints are doing some noise, especially on lateral movements.
Those noises break the immersion so I would like to fix that.

You may have a look at some of the photo/video on my google drive here, the last pictures/video shows my 6dof:
https://photos.google.com/share/AF1...?key=QkhtUjZ3VTRtTDZCLXpRSDRUMjI1am9valdLWDVn
or directly this one :
https://photos.google.com/share/AF1...?key=QkhtUjZ3VTRtTDZCLXpRSDRUMjI1am9valdLWDVn

The total weigh of the platform is around 500kg (so 700kg with 2 pilots), lever 20cm, 6dof range +/-15° +/-15cm. Motors are 1200W 6N/m 2000rpm, together with Thanos controller and FlyPT mover.

I don't know how to fix those noises but have noticed that on professional platforms they use U joints for the top connection like on this platform.


I could not find those kind of U-Joint on the web so maybe one of you can help me.

1) Do you think there is a way to lower those noises (have tried already silicone spray but does not improve significantly)?
2) Is one of you able to make custom UJoint for my platform ? If yes what would be the approximate cost for 6 ?
3) Do you have some links where I could find those type of U-Joints (preferably in Europe) ?

Thanks for your help,
Olivier

Wow! What an ambitious project Now that I saw the pics and videos, I understand why you have an all-up weight of 500-700Kgs.

I have no hands-on experience with these heim joints, so I can only speculate:
1. If lube doesn't make a notable difference, maybe you need ball joints with an inner teflon lining. I know these exist, but I guess they are more expensive.
2. Perhaps you are stressing those joints to the limit already? Does the squeeking sound still show up even with an empty (much lighter) platform? If it doesn't then it might indeed be an overload. If it does, maybe it is the geometry that needs a change.
3. You are using two Heim joints on each of those rods, so the connection from the ground to the platform is underdefined. That is not a bad thing in and of itself! It just means that you are giving it 7 degrees of freedom where only 6 are required. You can feel this 7th degree by twisting the rods. You should be able to twist them axially while the sim is in motion. Maybe those rods are twisted all the way in one direction and the outer body of the joint rubs on the high angle adaptors you are using. Like on this picture:


Try this:
Try to twist all 6 rods left and right a little and then leave them somewhere in the middle, so that there is still a little wiggleroom left over on either side. Then slowly and gently move the platform. If the squeeking is gone for the first few moves, this might have been the cause.

This guy seems to have a setup similar to yours. Maybe he can help you.

If you PM me your address, maybe I can send you an example of the strongest U-jonts I have before I go on vacation in August. Take a look at them and tell me what you think. But I would doubt they are able to carrry the insane loads of a 700Kg platform

Cheers :-D

Yeah, same here. And it's harder than one thinks to determine exactly how much range of motion one would need. In the end I have simulated a few extreme test cases in CAD but I don't have a mathmatically solid proof that those are indeed the most extreme angles. I am pretty confident,...but NOT certain!

I have tested every FDM material I could get my hands on just short of PEEK. In the end I came back to simple PETG. Yea, there are some higher performance materials, but the question was not: Which is the best performing material? But rather: Which is good enough?
PETG delivers ~20MPa in Z-direction, ~50MPa in XY, is virtually immune to any chemical you would ever find in a non-meth-cooking-hosehold, handles temperatures between -50°C and +65°C, and all the while printing absolutely warp free. There are better materials, but those are a nightmare to print accurately.

I have considered SLS printing PA12 through Xometry.com, but it was just prohibitively expensive. Almost the price of CNC'd Aluminum and I wasn't even sure it would provide 20MPa in Z-direction.

I have done plenty of FE analysis to find which area of the U-joint would experience the highest local stress. Then I desigined and redesigned the part with that worst case stress in mind and I generously overdesigned it with a safety factor of 6. Then I printed a few test samples, loaded them up to a factor of 4 and when I realized that there was just no way I would ever be able to generate enough force to determine the ultimate failure load, I gave up trying.
I am printing PETG at 285°C, 5 perimeters of 0,8mm each, 5 top and bottom layers of 0.3mm each. 60% adaptive cubic infill. Never had any issues.

Btw, I agree that ABS (ASA) is the better material, but PETG sure is good enough.

Cheers,....