Epoxy Bearing Material and Method

[JohnMcNamara]

Hi All

I made a presentation to a group in Melbourne Australia last week.

It includes the formula for the epoxy compound resulting from my experiments.

A piece of threaded rod with custom epoxy nut described:


Attached a link to a powerpoint presentation:
Bearing - CNCZ.ppt

Cheers

Macka

[ckelloug]

Dear John,

This looks like some excellent work. I find your writeup very impressive. A lot of folks including myself and jhudler have talked about making a bearing material substitute on the monstrous E/G thread but nobody besides you has done it that I am aware of.

As for mold releases, a commercial silicone mold release for epoxy is only about $10 US a can and it works wonderfully. I use the EaseRelease 200 from Mold Release Agents for Mold Making & Casting from Mann Formulated Products. Because this worked for me, I haven't investigated other alternatives though it is possible that some others might be even better.

As for teflon, it can be sourced from the internet, even directly from DuPont and because teflon parts are sintered for manufacture, there are a wide range of particle sizes available.

I can tell you that increased temperature cure generally improves the temperature resistance of a set epoxy material but the exact temperature to use and the cure cycle are specific to the epoxy and hardener. If you're really brave, call Huntsman and ask about the best cure for the epoxy you're using. The 5 minute epoxy will generally set up more brittle than the long cure stuff cured at higher temperatures.

A bit of googling turned up a Canadian company lowerfriction.com with tungsten disulphide which is said to be far superior to molybdynum disulphide in terms of coefficient of friction $50 us per pound approximately. They also have spray can versions that embed a semipermanent lubricant layer into the items that they are sprayed on. This looks interesting. Another possibility now that we are in modern times is to replace the calcium carbonate and graphite with hexagonal boron nitride which also appears to be available for about $50 US a pound.

Additionally, a titanate coupling would probably increase the strength of the material and the temperature resistance. The only way to get this though is to talk with Kenrich petrochemicals and I don't know whether they do business outside the U.S.

Drop me a PM if you want to talk. I've learned a fair amount over the last several years working on epoxy granite and have a fair reference library. Even though I am on unrelated extended travel right now, I do have a copy of the Lee and Neville Epoxy Resins handbook with me.

Regards,

Cameron

[Ken_Shea]

John,
Excellent write up, complete, concise, to the point and very useful.

Thanks for sharing.
Ken

[mactec54]

ckelloug

There have been many others do this before, but don't have the writeup like is excellent job Macka


microcarve (John a Zone member has done one for a ballscrew photo attached

I have used Moglice epoxy to make nuts like this that work very well & other materials as well

[BobWarfield]

Great write up, John!

Best Regards,

Bob Warfield

[RomanLini]

I have made quite a few composite bearing systems and low backlash leadnuts etc.

There's some pictures of the finished nuts in my router build thread;
DIY hobby small plastics mill/router
(it also uses plastic bearings - Igus brand) although I am not at liberty to discuss the materials and methods used for the leadnuts. I wish people would use more composite materials, metal is just one building material.

And congrats John on your work!

[JohnMcNamara]

Hi RomanLini

Nice Machine!

Reading through your post I noted your mention of shrinkage “The leadnuts were cast from a special slippery plastic that I work with and know its shrinkage modulus so they shrink during casting to the correct tightness and have no backlash. The end support blocks were cast from the same material.” End quote.

A search of Google using “Shrinkage epoxy” will reveal a number of threads

The example shaft and nut in my presentation also has preload. The question is was it the epoxy or the manufacturing process that caused the preload?

As the PowerPoint shows threaded rod was first coated with the metal filled bearing material (approx 5mm thick) when this had gelled off but not fully set a nut was threaded onto the shaft and with that nut at the bottom the piece of black pipe that was to become the steel outer was slipped over and the remaining void filled with plain (No metal filling) epoxy. And another nut screwed on top to clamp everything together

This was placed in an oven at 100c (212F) for one hour.

The inside of the pipe was quite rough and apart from sanding to bright with rough sandpaper was left as is. Maybe it would have been better to turn a fine thread for a better mechanical connection. If I had access to a stress testing machine it would be interesting to see how much it could take before being pressed out.

The shaft I used was commercial rolled thread 1” Whitworth form. The round crest and bottom of this form are not supposed to be used. No doubt a cause of unnecessary friction; it would be easy to remove the crests of the thread in the nut and a couple of thou off the rod may also be a good Idea?

The steel pipe will expand when heated.
The coefficient of thermal expansion for steel is 0.00000645in/in/deg F
212 minus Ambient say 72 = 140 X 00000645in x 2 for a 2’’ pipe .001806inch nearly 2 thou

Upon cooling I guess the steel will prevail over the epoxy. Suggesting the epoxy contained within the tube will be compressed. The threaded rod itself will mitigate by expanding so maybe we are back to square one or maybe not?

Engineering is an Art as well as a science.

Cheers

John

[RomanLini]

...
The example shaft and nut in my presentation also has preload. The question is was it the epoxy or the manufacturing process that caused the preload?
...

Both, I expect. Epoxy has a small percentage shrinkage due to loss of volatiles and you also used an elevated temp cure, so after curing you let it cool down and it will shrink and get tighter.

...
The shaft I used was commercial rolled thread 1” Whitworth form. The round crest and bottom of this form are not supposed to be used. No doubt a cause of unnecessary friction; it would be easy to remove the crests of the thread in the nut and a couple of thou off the rod may also be a good Idea?

I used a ballscrew so I can't speak from experience with the threaded rod. But one of the good things abut a soft nut is that it quickly gets worn (cut?) away in those fine areas and the eventual result is an "average" fit to the general shape of the screw.

It does settle in too after some work cycles and heat/cool cycles.

...
The steel pipe will expand when heated.
The coefficient of thermal expansion for steel is 0.00000645in/in/deg F
212 minus Ambient say 72 = 140 X 00000645in x 2 for a 2’’ pipe .001806inch nearly 2 thou

Upon cooling I guess the steel will prevail over the epoxy. Suggesting the epoxy contained within the tube will be compressed. The threaded rod itself will mitigate by expanding so maybe we are back to square one or maybe not?

Engineering is an Art as well as a science.
...

Sure, but the plastic usually has a higher thermal expansion than metal. And the source of heat will most likely be the screw/nut sliding junction from friction. Then the plastic is a relatively good insulator so not much of that heat will get to the outer tube. The thermal model is very complex and depends a lot on the individual motion task, so as you said it is an "art" and the best technique is always make it and see.

I've had some failures in the past but once the materials are familiar enough it's a lot easier to get successes with them.

PS. And thanks for the compliments on my machine! It looks junkyard but has very little flex or slop and I routinely machine things to a 0.01mm precision, things like moving assemblies, bores etc. Sometimes I will cut something, and before removing the part measure it with a micrometer and decide "Hmm, I'll take another 0.01mm off that"...