Loading
mightynozzle

Customizable Linear Bushing / Bearing - LM8UU, LM8LUU, LM12UU, LM12LUU, LM16UU, ... IGUS like

by mightynozzle Mar 25, 2017
Download All Files

Thing Apps Enabled

Please Login to Comment

Can someone explain how these would be installed on the Y axis rails? I have a MonoPrice Select Mini V1. It prints quiet except for the Y axis. Do I reuse the existing mounts and press these in? Thanks

Hi, i print the linear_bushing_LM8LUU_8.15ID15.00OD same the 4 picture but with 0.20 résolution in S3d , PLA , infill 0.
I take very slowly drillbit 8mm into it...it is perfect !
after that, insert in the carriege, wtih a rod, it slides very well and it's really perfect without lubricating, better than ball bearing...but for how long time ^
^ ?
Thank you man !

what kind of infill works best?

Don't use infill. It should be printed solid with perimeters only like in the 4th picture.

Comments deleted.

i bought original igus, they have excessive free play i dont understand. rod measured 8.01... so im going to try your design, should have send back igus bearings as they have more tolerance as they should be

IGUS should be bigger when it is not in housing. If it pressed in right size housing, it is squeezed little bit and then it should have exact size.

Well, i had to go that alone. A warning to anyone considering this, after a week of fiddling, trial and error, and so much wasted time and filament, i think i would NOT recommend others do this. Printing the bearings straight up PLA is almost useless, at least on the Prusa, because even the tiniest amount of tension from the U bolts under the bed binds them up (yes i believe this was mentioned elsewhere). I ended up getting some cheapo metal bearings off Amazon that were identical to the ones shipped with the printer. Gutted them easily, but then had to trial and error my way thru getting a size that would fit the inner grooves that hold the metal balls. Finding that sweet spot where the pla would insert into the metal relatively easily, but not slide out, was not easy. Ontop of that, it seems every iteration i printed changed dimensions when it should not have, and that didnt make things any easier. Iv attached an image for how mine ended up looking.

I also thru the whole 'must print in perfect perimeters' thing out the window. Wasnt happening with how i had to shape mine to get it to fit. It would work with the standard bearings straight out of openScad, but thats it.

What i found ended up working the best was LOTS of post processing. I found a size where the bearing would slide onto the metal rod pretty snug. I printed them in batches of 4. There were always some better than others, despite each batch being 4 identical (cloned) bearings. Not sure why they varies. So id pick the best ones and process them. This requires LOTS of elbow grease. I would run each one along the bar very quickly, once it had broken in initially. Then id rotate it 90 degrees, and do it again. Then id pop it off, blow into it to cool it down, and do it again. While running the bearing along the rod, i was pulling the bearing from 1 side, so 1 inner surface of the bearing was getting the most friction, hence the rotating 90 degrees.

This sometimes took 3 or 4 passes to get the bearings into a good spot, but id go from VERY snug, to where i could flick them up the rod with my finger and theyd slide up and fall back down. I then applied a little gun oil, as its all i had for lubricant.

The 3 bearings in the X axis (behind the extruder on the mk3) were pretty forgiving. I actually printed 1 long pla bearing, and wrapped 2 metal sleeves around it, just to make sure there was no deviation between 2 seperate bearings. The 3 on the y axis/bed were more troublesome. Even with the metal sleeves, the slightest bit of misalignment or over-tensioning caused more binding than i was comfortable with. In the end, i got them to hold just enough, and still slide smoothly.

So, its doable, but its huge Fing pain in the ass. And, at least in the case of the mk3 since, since most of the noise is actually from the motors, this realy wasnt worth the huge amount of effort, and the numerous times i had to tear the printer down and build it back up.

hey, can you give a little more info on how you got them to fit in the metal sleeve from a normal bearing? i also have an I3 Mk3, got a handful of extra bearings, opened one up, but the inside of the bearing has 4 additional raised sections on the inside. im guessing this is probably the same for all these bearings. what OD size did you come up with that worked? I come up with 11.1mm at the tops of those inner ridges.

when i tried inputting this smaller OD size into the scad file, i was finding the inner teeth were basically disappearing, and i was just left with a solid plastic tube. im guessing this has to do with the fact that a minimum size needs to be maintained in order to slice properly with 4 perimeters?

just curious how you got yours working inside the metal sleeves. i printed a test bearing at the full 15mm OD, and it fits the steel rod well. but i dont want to go tearing the printer apart until i figure out which route im going... metal sleeves or just pla. im just worried about the pla version deforming, as you mentioned, when i put it in the U bolts at the bottom of the Y axis. though, honestly, with the instructions from Prusa saying to keep the tension on those U bolts to "just enough", these seem like they would be plenty strong enough to hold up.

I haven't tried pressing these bushings into anything yet. However, when I worked as a mechanic in a bike shop we installed all kinds of press fit bearings and bushings. We often used a bicycle quick release skewer. We would put the skewer through whatever we were pressing then thread up the nut on the skewer until it was firm. Then just close the handle on the skewer and the cam in the skewer would press the bearing right in. Sometimes you have to open the skewer, tighten the nut and repeat a couple of times if you are trying to press something a long ways. If your skewer is too long you can put some appropriately sized sockets on it to take up the extra space.

is this so i can use soft steel rods? :)

You can try it, but i recommend using hardend steel rods.

Mine were printed of CreateBot black PLA, enclosed in a 1 mm thick aluminum tube. After a 6 hours long ABS print in an enclosure, when they cooled down, the bearings shrunk so much that I had to remove them and put back the original metal ball bearings. Just a heads-up that you should first check if your PLA also shrinks that much. Perhaps annealing would help, but it requires a lot of trial and error, and currently I have other projects to do...

Would be nice to add a parameter to OpenSCAD that inlays the size of the configured dr into the D of the print so it is easily spotted when printing multiple files ;) Also, if you add a chamfer to the model, it does not turn out like your #4 pic no matter which way you slice it. Only if you keep it straight like your example will it look like that.

it work great on anet a8 y axis ... print this file : linear_bushing_LM8UU_chamfer_8.15ID_15.00OD.stl
noise Decrease a lot

These bushings are awsome! Thanks man.
For make it easyer for others why not export a profile in simplyfy 3d and put it to the files?

I was able to slice it corectly using cura by gathering information here from comments. In Simplify 3d i was not able to achive the right slicing for the 12mm bearings. Thanks man for your awsome work.

Hey guys. Some help if possible.

I printed and fitted these to my Cocoon Create Touch (Dup i3 Plus clone) and am finding they are binding quite a lot on my y axis. Before i installed them I made sure they moved freely and smoothly on the y axis rods. They were tight at first, too tight to move easily on the rods (at least by hand). I filed down the inside until they glide nicely but the entire printer shudders when moving in the y axis.

Any advice would be really appreciated. Thank you in advance.

Same here. Printed the LM8UU and, although they glided nicely by hand, the Y axis on my Mendel90 screams at certain frequencies and had pretty big stiction (big static friction, low dynamic friction). I overcame a big part of the stiction by modifying the bearing holders. As the Thing Details says, the bearings should not sit too tight on the holders because they deform and "grab" onto the rod.
Despite this, I´m still having problems. At certain frequencies the bushings just start vibrating and screeching very loudly. Maybe it has something to do with rod alignment.
Right now I´m printing a modified version with bigger inner diameter (8.25) to test.

As you guest it often does come down to rod alignment. If they are not perfectly straight for example. Sometimes it's the y carriage being out of alignment - on the official A8 support group one fellow found his carriage out of alignment by 2mm (so he ordered a new one).

is this possible to create LM3UU? This one is expensive

£1.53 ebay,,, as of Dec 2018, only just came across this as mine as getting nosey, but i think I'll buy some instead,, reading some of the hassle you get

I think it is possible. You should print 4 bushings at a time, because of the small size. First test print could be 3.4mm, 3.35mm, 3.3mm and 3.25mm inner diameter.

Hi, thanks for your amazing work!

I'm trying to print the file "linear_bushing_LM8UU_chamfer_8.15ID_15.00OD" for my Anet A8.

With the settings that you provided, the slicing is like in the photo (the first and the last 3 layers are different)

Is this a problem? Thanks.

This should be no problem. The chamfered section takes no load.

Great! Can you add the 2 grooves?

Added in OpenSCAD file. You can now set D1, W and B.

Super! Thank you!

TOP!!! Neutral PLA 0.2 very very SILENT!!!

When using the customizer, I get the error: #<RuntimeError: Failed to get https://www.thingiverse.com/download:4535966 403: Forbidden>

This is a known problem of the makerbot customizer and I don't know, when it'll be fixed. You can alternatively use OpenSCAD. There you can open the scad file and change the values in the editor. The parameter are the same as in the customizer.

In Cura, when I load the LM8LUU bearing, there is a gap between the wall on the OD of the bearing and the wall on the ID of the bearing. So I tried to increase the number of shells from 2 to 3, and nothing happened. i had one of my friends try it, and he was able to get it using 3 shells. There must be some setting I am missing, anyone have any ideas?

Please scroll down in the comments. I already described the setting in Cura.

I set all of those settings. Went through all the comments yesterday and couldn't get it to work, was wondering if anyone else has this problem. I'm gonna keep trying some things, but just figured I'd ask.

Could you upload a screenshot of it?

I have tried other settings other than what is shown in the side, just curious if anyone knows of a specific one that isn't intuitive that will help.

If you want to print the example LM8UU stl file, then you have to set the wall thickness to 0.96 or wall line count to 2 and the line width under the quality section to 0.48, if you have a 0.4mm nozzle. Same problem as geiserclan had.

already tried that. I've done every combination possible of the settings that are visible. Like I said, my friend was able to get his to work, so there has to be a setting I don't have turned on that's causing this to happen. I've done lots of trouble shooting, but I'll probably just have him go through every setting and see how mine compares.

Line width may not be visible. You have to add this setting. Check out the screenshot!

I thought that the wall thickness and wall layers were working together to determine a Line width. That's my bad. Didn't know that was a setting I could change! Regardless, it works now. Thanks for the help man!

Can it be that only outer diameter has an impact on the printed busing ?
I customized in openscad to set 4 5 or 6 mm inner diameter with 15 outer diameter, but when I import those parts in the slicer they all have the same inner diameter. also tried with ID=4mm, OD=30mm, but ID is really too big.
After looking at the code, it seems, the teeth are translated of 1/2 * OD, which makes ID proportionnal to OD.

Thank you pointing me to this error. I already fixed it. The inner diameter should be correct, now.

I keep getting an error in the customizer when I want to generate the STL.

Seems to be a temporarily problem on all customizable things. Try it later again, or if it is urgent, you can use OpenSCAD.

Do these work with aluminum or carbon fiber linear rods?

I only run it on hardened steel rods. Don't know, if this will work on aluminium or carbon fiber.

Love these bushings, is there any way to customize the angle of the chamfer to be anything other than 45 degrees?

I've added an angle option!

Has anyone printed these in Cura and therefore know what the setting should be to match the 4th picture?

You should match the line width between the customizer setting and Cura. Cura calculates a line width of 0.4 with a 0.4mm nozzle. You need to set the Quality -> Line Width to 0.48 in Cura or set the Line Width to 0.4 in the Customizer.

Yup That did the trick.

Thanks for the reply :-)

This is a very good piece, no issue and awesome result on first attempt.

What is the advantage of custom layer heights?
I mean, why 0.12mm instead of 0.1mm or 0.2mm?
I ask because the last time I printed individual layer heights, I had gaps between some layers.

This is not a custom layer height. Maybe your slicer suggest these .1 steps.
A regular stepper motor has 200 full steps per revolution (1.8°/step). With a T8x8 lead screw, the Z-carriage moves 8mm per turn. So one full-step is 0.04mm (8mm/200). This is, why I only use the multiply of 0.04mm --> 0.08, 0.12, 0.16, 0.2, etc.. I've the best experience with these layer heights.

Ah, okay. My Anet A8 uses T8x1.5 lead screws. So the optimized layer high is 0.1mm or 0.2mm

No, the Anet A8 uses a T8x8 --> one turn = 8mm travel. You can simply check it, by marking the screw or rod, travel 8mm very slow in the z-axis and count how many times the mark passes (=1 time).

i'm going to try and get this to fit 3/8" and 1/2" tube for my hevo build, has anyone done this already? I've just looked at the openscad file, this is amazing, thanks for this it'll make my hevo build much easier now.

Hey guys which bushings (Filename) does I need for my anet a8?

Thanks :)

You can start with linear_bushing_LM8UU_chamfer_8.15ID_15.00OD.stl
If it is too tight or loose, you have to use the customizer or openscad to change the inner diameter, until it fits your tolerance of your printer. Then you can print a whole bunch of bushings.

do you know where i can find the bushings that fit linear_bushing_LM8UU_chamfer_8.15ID_15.00OD.stl? i dont want the heavy alu thats stock
and what is the correct bushing for the Z axiz for the anet a8?

It should be this one: https://www.thingiverse.com/thing:1769943
For Z, you can use 2x lm8uu or 1x lm8luu. But I made the experience, that clamps of the Z carriage makes the ID of plastic bushings (Igus or printed ones) a bit tighter and they didn't move freely. That's why I only use the metal bearing for the Z axis.

LM8UU Bed & Extruder Cariage holder
by stasxp

ty for the fast replay and help, maybe i should bu some quality bushings for the z carriage, im missing some ball bearings in them.

edit: the clamps have the bushings allready inside

Thanks a lot I will try it!!

tnx mightynozzle

the petg should not consume less than the pla?

pla lubricated or not?

I've only tested it with PLA and a drop of PTFE oil or PTFE dry spray. Don't know if it will work with PETG. But you can try it and share your experience!

Hello, I tried petg, it feels sticky (high static friction, low dynamic friction). i think pla is better

Thanks for testing! My first PLA bushings have been in use for almost a year. Still no wear. So, PLA is a good choice.

Hello! Would you suggest using these to replace Anet A8's stock bushings? I want to get rid of them (NOISE) really badly, but I'm not sure if it's wise, as I would be printing with bq's blue PLA instead of IGUS's... Thank you!

I think the most people uses Igus or printed Igus style bushings to get rid of the noise. Needs a bit of time to get the right setting for the printed one, but after that, you can print a whole bunch of bushings in one step.
My bushings moves every day since march this year and I can't recognize any wear out. If you are not happy with the result, you can always fall back to the stock bearings.

How long do the bushings last say compared to nylon ones you purchase?

Okey, I'll give it a try, see if I can get the right dimensions with a good quality and have them working ASAP. Thanks!

Hi,
I can't get the gcode make look like the model in your 4th picture. I matched the extrusion_width_in_millimeter = 0.48; and number_of_perimeters = 4; to the corresponding settings in Simplyfy3D 4.0. But the outcome looks like in the 5th picture.
Maybe you can give me a hint?
Thanks in advance!

Activate the option "Allow gap Fill" in your FFF-Setting --> Advanced --> Internal Thin Wall Type
If it doesn't help, upload a screenshot of the Advanced settings or upload the FFF setting file on dropbox and send me the link as PM or as a comment.

I already had the setting "Allow gap Fill", so I've set "Internal Thin Wall Type" to "Perimeters only" and then it looks like the model in your 4th picture.... Thanks for your help!

Can you post the link about the oil you used?

I bought the spray in a hardware store. It is Nigrin 72247 PTFE-Spray 100 ml. I think any dry PTFE spray will work similar. There might be also WD-40 dry PTFE spray in your area available.

Are you guys using this one the CR-10 if so where?

The stock CR-10 uses a different linear system, without LMxUU.

I really like the idea of your customizable bushings but would like to request adding a chamfer to the ends of the bushings for the inner and outer. Also adding customizable number of internal rings to limit the amount surface area to reduce friction on the smooth rods. Openscad is way to complicated for me to write such a script like you did.

I've added a chamfer setting in the customizer.

I'll add an chamfer option in the next update. Thank you!
The ring option will be to difficult to print without support. But you can already set the number of teeth of the bushing to reduce or add friction.

How to do this for Anet A8? The ones I did became great!

what is the best setting for Cura software to print the bearing

I don't have Cura installed. But if your slicer has a preview option, you should check the perimeter as you can see in the 4th picture. It is important, that the extrusion width and perimeter is the same as in your slicer. You don't want any infill, because this can destroy the round shape of the bushing. Only use perimeter without gaps between it. For layer height, you should choose the smallest possible, s.th. between 0.08 and 0.20mm. Temperature and speed depends on the filament and hotend/extruder.

Printed a pair of LM8LUU
Working as a charm, very silent.
THANK U!

Your prints look amazing. Nice work.

Looks awesome! This may be exactly what I've been looking for! Been wanting to build a hyper cube but don't really want to use 10mm rods and super expensive bushings... This would allow people to still use the normal hypercube parts and just make the printed bushing have an 8mm inner diameter instead of 10! Awesome!

Actually I have replaced all the china LM8UU bearing to the expensive IGUS RJ4JP-01-08, but they are to loose on the rod. Thats why I programmed this customizable design to print my own bushings that perfect fits. And a printed LM8UU costs only about 3g of PLA.

The Igus are designed to be used with the right sized smooth rod eg. 8mm h7 and for the low profile one to be press fitted. In that case there is virtually no play. Igus clerly document the way it must be mounted for best results.

As for PLA, there are tens of similar bearings published here. In fact is easy, 5 minutes, to do. The real issue is they won't exhibit the qualities of Igus polymer, will wear out quickly because of non flat surface (round filament), may bind because of incorrect alignment.

On the other hand, properly made bearings using Igus filament (careful adjustment and finish using for eg. a good reamer) will have minimal friction, play, no binding, long life without lubrication similar to the factory ones.

Bronze bearings sinterd iron etc...are also suitable but requires lubrication.

My rods are h6 smoothed and hardened steel rods from https://www.dold-mechatronik.de/Praezisionswelle-8mm-h6-geschliffen-und-gehaertet-ZUSCHNITT-bis-1200mm-1300-EUR-m-025-EUR-pro-Schnitt
But the RJ4JP-01-08 are to loose + they stuck on the rod (they don't move alone throw gravity). The whole X-Idler segment will wobble against the bushing and will cause more tolerance in the x and y axis. This might be no problem in the y axis, or maybe not in the x axis, but the problem will be the z-axis of a Prusa i3-like printer. But the printed one flies over the rod and have no noticable play.
About the wearing I will use some PTFE or machine oil. This should prevent some heat through friction. I see, that you also have some IGUS-style designs. Do you use any kind of lubrication?
I will also try print some bushings with iglidur filament from IGUS. I have three samples from them to test. But I don't think that they are better than PLA. These are advertised to be stronger as ABS and they should be printed like ABS.

I don't use any lubrication.
I don't have any wobble because first I eliminated its cause. All my bearings are of my own made with Igus filament. I kept the Z linear ball bearings but replaced the SS smooth rod by hardened one of the correct size. The play then is minimal (0.05) thus vibrations due to the X carriage high acceleration have no effect and no binding at all while with polymer bushings because of the impossibility to perfectly align the two Z smooth rods on a Prusa frame, there could lead to some binding. In fact to validate some ideas I had, I removed the two smooth rods and just used the lead screw (TR8 whose quality surprised me) as lead screw and guide using specially designed Igus nuts (no play, no backlash). It was working, but I had to drive my Z motor to the max because of this lack of // leading to extra friction. Would work with a better frame, more rigid, box type. I am looking at it but for an other kind of set up than the Prusa.
Even the bearings of the idler are made with Igus filament and running on Aluminum 4mm bar, same for a bearing in my extruder. Indeed Igus bearings run well on anodized Al. In fact too smooth a rod is not advisable.
The advantage in this hobby is that by making one's own bearings we can perfectly match them like in the old days, reaming, scraping, burnishing.