Bending of the rail should obey the calculations, but youngs modulus is what matters, not yield strength.
The real issue is how the bearings ride on the rails. They create a very concentrated point load and this can mash flat spots into the rails. I think Ryan did some tests in the early days and it was a problem.
Quick, back-of-fag-packet, calculations suggest to me that, given a Young’s modulus of about 70Mpa for Al and about 180Mpa for Stainless steel (SS), the Al is only slightly less stiff with a wall thickness of 3.2mm vs 1.5 for SS.
As a follow up, if stiffness is a problem I will fill the pipes with “builder’s foam”. Not only will that increase stiffness but it will also help with dampening vibration. I feel like Al tubing is the way to go for me but I will also buy mild steel pipes as they’re really cheap and I can do some comparison tests. To prevent corrosion I will lightly coat the steel with light machine oil - standard practice in machine shops.
The steel bearings can run flats in thick walled stainless if over tightened. Aluminum will get flats with regular tension.
I waxed my steel tubes, so far no rust, even on the bearing tracks.
Wow. The loads are higher than I imagined. Sounds like ordinary mild steel is the way to go.
It isn’t the loads bending the Al as I thought was being implied (and thus denting the rails) but the abrasion of the steel bearings on the surface causing problems. This abrasion wears a flat onto the rails themselves (Different Z axis tubes). Hence the need to avoid soft materials like Aluminium.
If you’re getting abrasion from your bearings, you need to change them, they’re no longer rolling. (it’s not abrasion)
bearing surfaces always abrade each other don’t they? Hopefully by a tiny amount…
No, bearings are there to keep abrasion from happening in the first place. Unless it’s a bushing, then you’ll get some wear. The roller bearings we use will deform the steel tubes they roll over. For steel it’s just a little amount, you’ll see flats happen after a few months to a year. Aluminum is much softer than steel, it will happen faster. In theory it should only deform until there isn’t enough force, but since these guys are flexing a little when they’re moving, it will deform the tubes to the point the bearings rattle.
I see what you’re saying, Barry. I am not sure it isn’t some kind of abrasion. That would be different than an application like a bearing in a router where the outside of the bearing is always firmly in contact with the bearing. Rolling has funny stuff happen because it can’t possibly touch at only one point, but the intersection of a line and a circle is one line. So there is funny stretching/compression that happens when a bearing rolls on a surface.
I would expect the bigger affect is more like a steel hammer on an aluminum bar stock though. The bearing is applying pressure and moving around, the leading edge is like a little hammer smashing the aluminum out of the way. I don’t know the ME terms for that.
I wouldn’t be surprised if the way to characterize it was as an abrasion or as an impact. But it definitely will deform and CF tubes delaminate.
It would likely depend on whether it’s anodized or hard-anodized, two distinctly different processes.
Now I am really confused. Here’s what I think people have observed:
After a while flats appear on the tubes where the roller bearings touch the tubes.
I can think of only 2 ways that this can happen:
This should be easy enough to prove. Section the tube and measure the thickness of material at the “flat”. If it’s thinner then it has been abraded, if not then it has yielded.
For plastic deformation I can also see that the contact is initially just a point (tangent of the circle) and that a flat will be bent into the tube until there is enough material to support the, now enlarged, contact surface. If this is the case, it seems to me that some kind of adjustment mechanism might be helpful to cope with the bending (or just increase the tube thickness).
Steel becomes significantly stronger as it work hardens whereas aluminium doesn’t. Hence a tube formed by direct extrusion (ODM) should be stronger than one formed by bending and welding the seam (conduit).
As for carbon fibre tubes I am only concerned about abrasion as the yield point is pretty damn close to the Ultimate Tensile Strength (UTS). Thicker tubes would solve that. Delamination of the composite is normally associated with a chemical process which should not happen in this case.
Even with a solid bar you could mash or plastically deform flat spots with a strong point load. The radius could decrease slightly even without abrasion or with very thick walls.
I would agree that it should “settle in” to a nonzero but limited amount of deformation since the deformation causes it to spread the load so it’s no longer at a single point. It would effectively have a reduced OD at that point but if the tube started out slightly oversize it might settle in to a proper fit instead of starting at the correct fit and ending up undersize.
It would probably be good to run a job that rolls along the entire working length of both axes many times because if the rails have flats in some areas and not in others it could introduce a bump when the roller crosses the discontinuity.
@jchidley #2 is what we’re seeing. Depending on what you want to replace more often is what you want to be the softer material. With our tubes, we can just rotate them slightly to get a new nonflat spot. Bearings we’d have to disassemble things.
I can see that it’s possible to overtighten the bearings onto the tube (pretty easy to do I’d think, even just using a hand-held screw driver) and therefore crush it, leading to flat spots. However, the load (< 2 kg per tube, I estimate) strikes me as way to little to cause problems by itself.
This does suggest, to me, that I will have to be very careful when tighening up the bearings - just enough to eliminate slop and no more.
One the the advantages of the current design, I’d say. But it would be good to know why the deformation is happening and eliminate it.
Point load of less than a half millimeter square. Realistically these are the wrong bearings to be using on round surfaces. The reason they’re used is because it’s a cheap system to build, and readily available parts in a medium sized city.
what bearings should be used? will the fit if we can source them?
Doubt you could find the proper bearings. This machine was designed to be built from hardware store parts. Big machines use linear bearings.
I have a local bearing manufacturer, so I know I could get better bearings. Ones with a v or u profile on the outer raceway would be better wouldn’t they? The big issue is I don’t think they would fit inside of Ryan’s trucks.
I know big machines use linear bearings im just wondering if there are bearings that would be better in that they would be less likely to put flat spots in the steel.