Linear actuator using ClearPath Integrated Servo System @ 72V DC

Haha,...

this is actually the part I am almost ashamed of

I wanted a double upper bearing, because I was anxious that a single upper bearing would not hold the buckling load. I know that others have built very similar designs with a single upper bearing and found nothing to complain about, but it just didn't feel right.

In theory, the only load on those upper bearings should come from the weight of the actuator itself (about 9Kg) but at the angle it sits at, only about ~2Kg are the effective load (9Kg * cosine of elevation angle * CG_ratio... bla bla bla.... divided by bla bla bla...). That really does not sound like much. But those ~20N act over a rather long arm (half actuator length = 820mm @ max extension) and it acts twice. Once from the upper half, once from the lower half. That alone is actually not a problem.

Here's where it gets dicey: Imagine the load of those meagre 20N flex the actuator by only 10mm off its unloaded axis (keep in mind: This thing can be 1650mm long!). Now you can potentially have all of the 1000N of nominal axial push force act across that 10mm off-axis-arm which results in another 17N of force perpendicular to the axis. Makes a total of 37N of perpendicular load all together, wich means more flex-->more arm-->more flex-->more arm-->more flex---> BROKEN ACTUATOR :-(

It just felt it was the safer bet to put the double upper bearing in.

The 98mm size of the thing was also a bit of a gut decision. Here's the trade-off: Larger will give more stability, but smaller will give a smaller overall build length. Hmmm,.... tough choice.

This is not a linear thing, however. It's more of a a hyperbolic function (or something similar, I believe). The cost (build length) is linear, but the benefit is non linear. Initially every bit of additional overlap will increase strength by a lot, but after some inches, you run into a "problem of marginal utility". I started with 80mm and increased to to 98mm, because it felt much stiffer and I found those 18mm to be a price I was willing to accept. And also, the double upper bearing will reduce forces in the bearing that generate friction significantly. Much more than just the factor of 1/2. You will get twice the sliding surface area AND much less leverage to exert load on them.

The embarrassing thing is this silly curvy shape. I did this to save some print time, but I think it didn't even make a minute of difference. If I did them again, I would go with a more classic, straight, almost cubic design that would surely underline this actuators timeless beauty and elegance

Feel free to make modifications! Just let me know why and how it worked out. I'm curious.

Cheers,... Dirty

Hey Dirty!

At all makes sense! This is what answers my question:

"The 98mm size of the thing was also a bit of a gut decision"

I was wondering if such a precise number has anything to do with any specifics. I DO LIKE YOUR CURVY SURFACE! It looks elegant, matching the curved surface of the joints and contrasting with the rails... and I like it seems to save some material! As I normally print parts during research multiple times (low/fast printing, then slow printing not enough density, then good density and so on) so I can test tolerances and stuff, saving material/time makes a difference... The top head is right now printing it's "final" test and it's taking 8.5 hours!

I found not tons of changes are needed. I will post the finished parts once done. Here the 3 parts so far I have been working on:

Head:

-Added some guides to help the insertion of the hot threads... Just a 1mm bigger hole to sit the thread while heating with the iron.

-Made bigger bearing hole (from your 13.05 to 13.15) as it was too tight to insert the bearings.



Middle Part:

Did a bit more changes, mostly to help printing without supports, and some tolerances.


Rod head:

Changed to a 1 inch tube, as 28mm was hard to find here. Also added a "rod screw" socket to secure the rod as I am not doing the inner threading



And that is where I am so far! Next part to work is the upper block... that is why I was asking details

Let me know what you think and if you have any comment.

Thanks so much Dirty!

Hey ,

Nice to see how another brain goes through the same considerations. And comes to slightly different conclusions, but that is the fun of this hobby: Making decisions and then seeing the results.

The wrench access crossed my mind as well, but I have opted for "less hole, more strength", so I had to print my own custom slotted sorta-sleeve-nuts:

This nut carries absolutely no axial load, and the PETG will not vibrate itself loose.

On the other side I used a hex screw. Also following the "less hole, more strength" paradigm

...but those are minor details. No major strength concerns there.

The 13.05 mm was a result of the tolerance of my printer. I tried different sizes and I ended up with the size that was so tight, that I had to press the bearings in with about 20Kg of load. But that was what I chose.

On one test part I put the plastic part in 60°C hot water and the bearing in a -20°C freezer. Then it could be pressed in place with minimal force and as the parts equalised temperature they developed a nice tight fit.

The "center body" could be printed without supports already, but admittedly it has some borderline steep overhangs

keep posting what you do,.... I find that super interesting. Or better: Start a build thread, I will surely follow.

Cheers,...

Ohh! So you printed those sleeve nuts! I was looking for those for a couple of hours without finding them! I thought printing them would not give enough hold for the axis! Will test that..!

That hot water/freezer is awesome! hehehe!!

I thought that starting a new thread on this design that is fully based on yours might be a waste, but you are right that it might complement and give some info to other builders! Might do that once I start officially building more than one actuator. For now I am still in the single setup process and enjoying it. Thanks so much for your time helping!

That being said...! One more question!

I saw that the depth of the bearings on the star of the show (that is what I call the top head of the actuator!... love that piece!) is shallower than the 2x8mm bearings. Is this intentional (I am sure!!??)

Would you leave a gap when pushing the aluminum piece or press it so is completely flat when screwing it. Wondering if it's better to have the entire part flushed to the head or leaving space.





Thanks again!

Charlie

My design is also based off of someone elses

Yes, that gap is intentional. There are two bearings inside that cavity, and the only thing holding them on the other side is a little "lip".


If I had made them perfectly flush with the surface, I was anxious that eventuall they would develop some slack, if the PETG lip would have given just a litttle bit. With the gap, I can now put the bearings in, and they will initiall protrude out by a few 10th of a mm. Then I can tighten the screws, which will press the bearings firmly against that lip. The gap will disappear then and the aluminum will act as a very firm "kinda-spring". So I can be reasonalby sure that the load changes of the actuator will not make them wiggle around, because they are under a constant positive load anyways.

Might be seeing ghosts here, but it really didn't have any negative side effect, so I kept the gap

You will see, it will dissapear when tightening the screws.

I used these Inserts, they hold like 3-4 times the load of cheaper ones,because of the diagonal knurling.

See the videos in this post:
https://www.xsimulator.net/communit...servo-system-72v-dc.11873/page-11#post-206076

Copy that! Thanks so much! Very informative and clear as always!

I have been looking everywhere for those INSERTS! Crazy enough I can't find any for metric or imperial that are reasonable close. Might end up getting them from Amazon.de, but they would arrive until February!!!

Meanwhile I ordered this... will get them today or tomorrow and see if it works!

Thanks again!

Oh, they look sweet!

I will give those a try as well. With these huge horizontal surfaces they should be insanely strong!

I guess they need a certain minimum distance from edges to cater for the huge threads. But other than that I'd expect them to be perfect.

I heard before that the Rutex inserts are hard to get outside Europe. I wonder why that is!?!

Yeah some things are weird... SFU1610 is impossible to get here in the USA, only 1605. FK12 is only available after March! So hoping on Aliexpress mercy!

Will report on the the inserts, and might change the hole position based on the wall distance... we thought the same!

Charlie

Just to confirm that the threads inserts work great.

Had to re-design some parts of the top head joint to allow the extra space of the threads. Also needed to add the thread in the hole as there is so much area that the thread takes. Also the trick was to add a "tapper helix" with a 4 degree in angle so the thread would get tight enough and at the right depth.






And here the top part finished with a little bit more "fatty" waist.



I 3d printed the metal plates (in black) for now as the shop that I am going to use is close for the holidays... and also it helps to make sure sizes are correct and all.

Charlie

Nice,.... they look rock solid!

Absolutely no strength concerns there, that's for sure I think a single one of the ones I used could hold about 200Kg and yours should be even stronger. And then there's six of them!!! This should be able to lift a small car (European car, not US)

This is not Fusion360, is it? I will have to see how to model tapered threads in F360. Never done that. I think sooner or later I will need the strength of those threaded inserts somewhere.
What I probably would have done is try to model those inserts as good as possible (with an uninterupred thread), and then subtract the shape from the parent body.
Did you have to heat them up and"melt" them into place, or were you able to screw them in cold?

Hey Dirty!

Its Solidworks, not Fusion360, but I am sure is doable. And yes, I guess doing a normal thread and substracting/doing a Boolean of a conic shape would work!

No need to heat at all. I was testing different options, including heating, and the best was cold with the thread.

Definitely that thing will stay there!

Hey guys,

Nice to see that there are more dirty-followers that are busy building. I am also working on the first actuator. Apart from the spacer, I have all parts ready for the first actuator.

@Dirty : I just think the inner diameter of my Upper block is a bit too small as the Igus lbearing will not go in without force. Or does it have to be pressed with a certain force?



Does this Igus bearing still need to be processed with Locktite or will it remain?

Tomorrow I will not be online because I am busy making traditional food for New Years, so I wish all readers a good transition to 2021.

Builder.

You will have to press them in with some force. If I had to guess, I'd say maybe about 20 Kg or so. It was definitely some work to get those in place. That, like all other parts with tight tolerances, depend a lot on the tolerances of the printer. I did some test prints to get quite a tight fit, but that might be different on your printer. If it's too tight, you can sand away a few 1/100th of a mm. If it's too loose, you can apply a thin strip of adhesive tape on the outside of the igus bearing. Either way, it should require some force to get it in place.

The upside is, that you will not need any adhesive if it's a nice tight fit. I have to admit, I don't think that there is an adhesive that would give good adhesion on these plastic bearings anyway. They are slippery as hell.

I had doubts if the "just-press-it-in" method would work, but so far it works flawless.

Hey Dirty,

Did you removed the "Key" from the shaft in the 80ST motors? I am still wondering if my pulleys are coming with or without keyhole... but want to make sure the shaft is removable before I start beating the Sh@t out of it trying to remove it hehehe.

I remember reading somewhere that there was a way to remove it... but maybe I dreamt about it!

Thanks so much!

Charlie

The key is easily removable with just a pair of pliers.
I found that out after I spent an hour online searching for key'd pulleys. :-D

I glued the pulleys onto the shaft with Loctite 638. Can be removed by heating up with a heatgun for 30 seconds.

Yes! hehehe, I am just discussing with the vendor from Aliexpress about it... that would make things easier!

So I would need to pull "normal" to the cylinder, or it slides out towards the end of the shaft...

I don't want to break my precious motors... they became the most expensive part of the setup hehehe

Thanks Dirty!

IIRC, I worked one or two of them out with pliers but eventually used a small vice to clamp the key into. Then wiggle the motor body around gently and it will become loose.

My plan B: Drilling into the key until it has so many holes it just falls apart.
My plan C: Grinding it away with an angle grinder and then filing it flush.

Still, the pliers/vice method should work and has the advantage of being reversible.

The key normally has a threaded hole in it so it can easily be removed with a screw without damaging anything.

Thanks Seattime!

Indeed, the key gets removed super smooth and easily with a M3 bolt.