Regreasing a Linear Bearing done easily

There has been a lot of weird tips for cleaning and regreasing a Linear Ball Bearing on the web, and I hope to briefly clean (hah) that up here.
Before I go any further though, I'm not sure how much difference this will even make. On a Voron 1.8 (which this tip is exclusive to), there are four smooth shafts an aluminium plate is moved up and down with. It moves very slowly.
This video shows that cleaning the grease off (isopropyl), then regreasing it with White Lithium grease makes a big difference - they are quieter in operation, and have less resistance (you can see this @9:21 ) .
I done this by using an applicator, and packing the grease inside by using two rods together to help push the grease on to the balls.
In this video, the parts used are an 8MM smooth shaft and a LM8LUU (tip applies to a LM8UU as well).
If you are an Aussie, this is the grease:
core-electronics.com.au/mg-ch...

[@oleurgast730]

Actually I am about to change most of my 3d printers from linear bearings to a combination of brass and polymer bushings. As I first started that 6 years ago, I got a lot of problems, as sometimes the movement stopped or sensorless homing triggered at a false position. The reason simply was bushings being much more prezise. Printers like the Prusa mk3s+ (my main printers) and the Prusa mini and many others are staticly overconstrained. Due to tolerances in production and veriation temperatures two rods are never ever 100% in parallel (I talk about micrometers). Ball bearings actually have more tolerance than bushings. So they still work.
Two weeks ago I got the solution to this problem. As example the x-axis of a Prusa mini.
The printhead uses two ball bearings and two parallel rods. Only one bearing and one rod would be enough staticly, exept the printhead could rotate around the rod. So the second rod actually only have to prevent this by restricting the movement in y-direction, so the head can not rotate arount the x-axis. But instead of a construction only preventing this motion, a normal bearing is used. So both bearings restrict in z. This is called over constraining. Normal ball bearings have some play - and thats needed to compensate for the slight missalignments if the rods are not perfectly in parallel (wich they never are). Simply replacing both ball bearings by Igus bushings or (much more precise) brass bushings can not work well, as the over constrainment leads to more friction (wrong triggering sensorless homeing) or even blocking.
The solution was quite simple: I downloaded a stl of an Igus-style bushing, used two cylinders with 8mm diameter and their centerpoint 1mm appart. Also I added a cube with 8by1mm² area in the middle. I subtracted this from the Igus style bearing and got an asymetric bushing, having 8mm in one axis but 9mm in the other. So I now use a very precise brass bushing on the upper rod, but my asymetric bushing (with the 9mm diameter oriented vertically) on the lower rod. Now it is very quite, self lubricating and works absolutly flawless.
For my experiments I used PETG for the asymetric bushing, so still some grease needed and not a permanent solution, but I will reprint it from Iglidur i150 soon.
For the next weeks I plan to try this solution on the other axises also.
For example the bed uses 3 bearings. The side with two bearings I will use Igus bushings instead (I already do this), as there is an optimised pearing holder on printebles for the prusa mini, wich alows to give the bushings the right amound of pretension (and it is very hard to align two brass bushings precise enough). On the over rod I will use my asymetric bushing, longer axis oriented horizontaly (at the moment I use a ruthex bushing wich has some play).
As I upgraded my Prusa mini to second z-axis, leaving the right side as is, I even have 3 rods with LM10LUU on z. I will also construct an Asymetric bushing for two of the rods (using 10/11mm) and one brass bushing.
These brass bushings are available with embedded graphite, so being self lubricating (but you can add aditional graphite lubricant to make the movement even smoother). Also Iglidur is self lubricating. So altogether much less maintanance required. Also much less noise (and as noise simply is vibration transfered to the air - much less vibrations)
This method of reducing overconstrainment and therefore using much more precise bushings insted of bearings actually should work on all cartesian printers (not only bedshakers, but also crossing rods or ender5-style). Sadly, coreXY haevely depends on over-constrainment, as the belts do a lot of turns. Also brass bushings are heavy, so their use is limited if you want to go lightweight.

[@amelliamendel2227]

I believe the grease goes inside the channel Sir