It says HG20 on the bearing housing itself, but I could not find any other data.
Per Hiwin, there are 2 possible preloads: Z0 (light) and ZA (medium). (ZB is heavy pre-load, but that option is only available on matched linear rail to bearing sets).
In the application guide, Z0 preload does not list machining center as suitable application. ZA and ZB do list machining center as suitable applications:
If Avid is indeed shipping Z0 preloads, I might try replacing mine with ZA (I’ve reached the chatter limit on my machine when processing aluminum and still have additional HP on the spindle per HSMAdvisor).
Avid confirmed Z0 (zero to light). And they are “heavy” duty “C”, not “super heavy duty” / “H”, based off the measured lengths of the linear bearing.
The benefit you would get from higher preloaded bearings is lost in other factors. For instance, on my machine, just the ringing is enough to render that 0.0005" difference moot.
It is always good to remember the machine specs as defined by the factory;
Pro Series –
Repeatability: +/- 0.002" (0.05 mm)
Accuracy: +/- 0.005" (0.127 mm) or better
Benchtop Pro Series –
Repeatability: +/- 0.001" (0.025mm)
Accuracy: +/- 0.005" (0.127mm)
The cons would be higher cost, quicker wear and of course more grease so they don’t cut grooves in those expensive rails.
I’ve modified the AVID kit pretty extensively, I’m aiming for 20N/micron stiffness for the machine, I’ve FEA’d the whole thing (I’m a structural engineer by trade).
The boogy man I am chasing now is the pivoting motor mount plate in the rack and pinion assembly, it alone is a ~20N/micron hit to stiffness, been brainstorming for a while how to stiffen that part up, no good ideas yet. Here is zoom in on deflection from FEA on that part. (I replaced the servo motor with a block and simplified the parts, but it should be pretty accurate, the deflection comes from the hogged out areas where the spring adjustment double threaded nut thing mounts to the motor mount plate. If anything, I’ve modeled it stiffer than it actually is, I didn’t try to model the pinion ↔ rack mesh, it can’t be very stiff):
