We experimented with some drop in bearing replacements for 12mm rod today. Brian machined them on the lathe them reamed them to perfect size. They seem to be perfect on our 12mm rods and oh, sooo smooth. Much less friction than linear bearings. Of course we need to torture test them but are there any materials experts out there with reasons why this is a bad idea? Will they wear well over the long haul? If these results are good, we may consider injection molding some in Delrin.
Opinions? Experience?
I love the idea of making our own Delrin bushings in the USofA to eliminate Chinese bearings from out BOM. I gotta believe that they will print better but I’ll try that tomorrow.
Did Brian machine them out of metal, delrin or another material? If metal, were they oil and friction smoothed? Just curious? I heard that was something you did for smooth metal parts. I need to play around with metal working techniques some day.
Have you checked out Igus bushings? A lot of material science work to come up with materials better than Delrin. I’m working on printing ABS adapters for 2 Igus J bushings that provides the correct compression for smooth no wiggle motion. Lighter than steel backed or oilite and not very expensive. Only a thin bushings needs to be exotic material. If you wear through that it’s not useful anyway.
Igus Drylin is formulated to be low-wearing, not necessarily low-friction. I personally find them to be quite sticky – any meaningful preload and they bind up. So you have to get the fit almost exactly right. A plain delrin, nylon, or PTFE glide will be much smoother.
The main issue I have with delrin/etc plastic bushings is that they’ll wear over time if the rods get dust/dirt/debris on them. In extreme cases, grit can impregnate the plastic surface and act as an abrasive on the rods, too. (Look into the Makerbot Replicator 1/2/2x X-idler pulley design – they used GFR-nylon and the spinning pulleys gradually cut into the steel shaft.
The service loads we see in 3DP are pretty mild, but we’re talking about millions of cycles and unknown end-user environments. Not saying LMxUU is necessarily the way to go, they have downsides too, but you need to consider what your target longevity is and whether end-users will tolerate the part having a finite effective live.
We machined them from Delrin. I guess dust is the biggest worry. But dust affects lm12uus a little. Hard to believe it wouldn’t work for a few years, and if they are replaceable, it might be worth it.
The loads are so light for these bushings… The X and y would get a lot more wear than the z… Maybe it’s worth looking at for the z axis.
While the figment test today was drop in replacement, it was only around a half inch long. Seems like it wouldn’t take much material to equal performance of a much more complicated bearing with lots of moving parts.
The igus and other materials are very expensive compared to Chinese bearings… I think I pay 50 cents or something (off the top of my head). But they would be pennies if injection molded. And sourcing would be so much easier. Just being able to make my own would be a dream.
I will set up a test rig to torture test and count the millions of cycles
@Brook_Drumm Something random that you may already have thought of… I think you may need/want different rods for plastic bushings vs linear bearings. Bearings need hardened rods and dry bushings need chromed rods. Chromed/hardened works for everything, but you can potentially go cheaper.
The biggest issue I’ve seen with the LM’s used is that we don’t load them ENOUGH with light weight structures. Most of the time there’s not enough weight on them to get the balls rolling, so you’re dealing with steel on steel sliding instead of rolling. I’ve experienced this with much larger systems too. The bearing seemed to stick when just on the rail, then I stepped on it and it was like skating on ice (no I didn’t try to skate on it).
There are lm bearings with adjustable preload, but are more expensive.
Brook, easy test, take that bearing, put a single motor on it and run it at high speeds. Add a 500g load on it and you’ll get a good estimate of wear.
A good bronze bearing can last for ages. I have a 60 year old sewing machine that runs circles around a new machine, all with original parts. The higher the accuracy of the bearing and rod, the longer it will last.
There’s a couple of ways you could pre-load the LM’s to ensure rolling. Use a third LM between the main load bearing ones, and spring-load it sideways to pre-load the others.
Alternatively, If you’re using steel rods, a Nd magnet mounted with a small air-gap can do your pre-load without adding friction. Added benefit with the magnets would be getting some vibration damping from eddy currents generated in the rail during movement.
Then there’s doing a proper precision press-fit housing or a split block where you can dial in the pre-load.
I like brass bushings but searching for a sweet spot of 1)cheap to manufacture 2)good performance 3)fast to make… I thinking just broke the “pick two” rule… Maybe I don’t get three ;). But not made in China really appeals to me. I’d love to make every single part here in America… But steppers seem impossible. Bushings would be a step forward at least. All things are possible but not cheaply.
I’ll preload that test for sure.
So how many days of 24/7 testing under load would be reasonable? I’ll run two or three rigs.
@Brook_Drumm , you can also test vesconite. Its specially formulated to be high wear resistant and low friction. Also, its easier to machine than nylon or delrin.
@Brook_Drumm there is a man named regpye that has been using 8mm delrin bushings for thousands of hours now and claims to have no wear issues. I ran across him on the reprap forum a while back and just the other day I saw some on his website. But if you can injection mold them and get the price to be super low that would be amazing!! They are so much more quiet than linear bearings.
I know your goal is to have everything made on America, but global trade and economy make it quite difficult to have everything manufactured in the US competitively. There is nothing wrong with using components from other countries, the issue is the labor supply chain, are the companies compliant with labor and safety law, are the employees sufficiently compensated. How are the working conditions, etc. You have a tremendous amount of power to improve the lives of people across the world.
And for test rig, the time is dependent on speed. If you calculated a 20% usage for an entire year (overestimate) at 40mm/s,how many mm would be travelled.
@Brook_Drumm , I built a small CNC machine about 15 years ago and turned my own Delrin bushings for it… they’ve been in use ever since and I’ve never had to replace them… and that’s against a simple stainless rod.
Anecdata here on the willingness of linear bearings to scratch rod, even with only a few hundred hours of use.
The other thing about a delrin (or equivalent) bushing is that it could be easier to replace. So people could run some kind of torture print every x months and pop in new bearings if needed. Also nice just not to have to ship stuff around the world multiple times.
