16mm supported rail how many linear bearings

[progomez]

Im thinking of buying some 16mm supported rails from zapp. Would it be advisable to use 2 bearings per length of rail, even on the z axis, which will be 200mm?

Would it twist with only two?

[Lee Roberts]

I think given the choice, i would go with x2 bearing blocks per rail. I would think this would give a better distribution of all things considered.

[Gary]

Yes it is, but also this type of rail is not the best to use as a Z axis, we suggest either round rail or profiled rail for vertical axis.

Im thinking of buying some 16mm supported rails from zapp. Would it be advisable to use 2 bearings per length of rail, even on the z axis, which will be 200mm?

Would it twist with only two?

[irving2008]

Gary, whats your reasoning on this? and some links to your recommendations?

[progomez]

Yes it is, but also this type of rail is not the best to use as a Z axis, we suggest either round rail or profiled rail for vertical axis.

Thanks for the reply.

So supported rail can be used on the x and y? Would 12mm z rails + carriages be sufficient to lift a standard full size wood router?

[Lee Roberts]

Thanks for the reply.

So supported rail can be used on the x and y? Would 12mm z rails + carriages be sufficient to lift a standard full size wood router?

Thats all im using on my rouckcliff machine and everything seems ok so far.

FOR SALE: x2 12mm Round Rail 1000mm - CHEAP XYZ (http://www.mycncuk.com/forums/showthread.php?t=430)

or

FOR SALE: x2 16mm Round Rail 1000mm - CHEAP XYZ (http://www.mycncuk.com/forums/showthread.php?t=429)

I'v also got the SMA units as well, the prices would be the same as Zapp's. :)

[irving2008]

The issue isnt the weight... its the amount of flex and that depends on the accuracy you want to achieve.

Remember that the cutter, router and mounting at the lowest part of the Z motion is acting as a lever trying to bend the z-rails. Assuming you are cutting wood with a typical router and a 6mm cutter you will be pushing something like 100N sideways force at a distance of say 200mm which is a moment of 20Nm (torque). That will flex 12mm unsupported rail by about .06mm and will result in a deviation of the cutter point about 0.6mm. If you are OK with that worst case inaccuracy then fine but even for wood thats high. 16mm rail will give a worst case deviation of 0.15mm, which is better.

[progomez]

The issue isnt the weight... its the amount of flex and that depends on the accuracy you want to achieve.

Remember that the cutter, router and mounting at the lowest part of the Z motion is acting as a lever trying to bend the z-rails. Assuming you are cutting wood with a typical router and a 6mm cutter you will be pushing something like 100N sideways force at a distance of say 200mm which is a moment of 20Nm (torque). That will flex 12mm unsupported rail by about .06mm and will result in a deviation of the cutter point about 0.6mm. If you are OK with that worst case inaccuracy then fine but even for wood thats high. 16mm rail will give a worst case deviation of 0.15mm, which is better.

Thanks for the reply.

That is why I was going to buy supported rails but zapp advised against. Maybe because they are opened backed? Just need a simple and quick way of making an x axis so that I can used the machine to cut a joe cnc.

[irving2008]

Thanks for the reply.

That is why I was going to buy supported rails but zapp advised against. Maybe because they are opened backed? Just need a simple and quick way of making an x axis so that I can used the machine to cut a joe cnc.
Supported for X and Y is preferable because the weight of the load (router + vertical forces) causes significant negative deflection in the vertical plane which the support effectively resists. However supported rail is strong to resist loads into the support but poor when being pulled away from the support which is why Gary recommends something stronger such solid profiled rails where the glides run on the edge and better resist forces equally into and out of the plane of the rail as well as turning moments.

Thats doesnt mean you cant use unsupported rail in thr Z direction. In reality the deflection will be much smaller than what I suggested and for a small machine 12 or 16mm rail may well be adequate if the depth of cut is kept small enough. As Lee says, his works fine...

[Gary]

Yes this is why.
I have had customers use supported (TBR) rail for Z, and it worked, but i dont suggest it.
If the Z is a short stroke, standard 16mm should be sufficient, for light wood and plastic, or 20mm if you can stretch to it.
I dont suggest using 12, 13 or 16 mm rail for X and Y, unless doing very light work or engraving.


Supported for X and Y is preferable because the weight of the load (router + vertical forces) causes significant negative deflection in the vertical plane which the support effectively resists. However supported rail is strong to resist loads into the support but poor when being pulled away from the support which is why Gary recommends something stronger such solid profiled rails where the glides run on the edge and better resist forces equally into and out of the plane of the rail as well as turning moments.

Thats doesnt mean you cant use unsupported rail in thr Z direction. In reality the deflection will be much smaller than what I suggested and for a small machine 12 or 16mm rail may well be adequate if the depth of cut is kept small enough. As Lee says, his works fine...

[Jeremy]

I've been toying with the idea of using supported 16mm rail for a Z axis, but was thinking of fitting the rails back-to-back, with the rails facing outwards. The idea is similar to the angled version already shown, but would be easier to fabricate from standard parts. The Z carriage would be channel shaped, to keep the linear bearings loaded in the right direction.

I had thought about doing it the other way around, too, with the rails facing inwards and a block carriage sliding up between them. This ends up a bit wider though, so might limit travel in the Y axis.

Jeremy

[irving2008]

I've been toying with the idea of using supported 16mm rail for a Z axis, but was thinking of fitting the rails back-to-back, with the rails facing outwards. The idea is similar to the angled version already shown, but would be easier to fabricate from standard parts. The Z carriage would be channel shaped, to keep the linear bearings loaded in the right direction.

I had thought about doing it the other way around, too, with the rails facing inwards and a block carriage sliding up between them. This ends up a bit wider though, so might limit travel in the Y axis.

JeremyThe problem with the back to back is that the main forces on the z-axis would then be perpendicular to the support which isnt really going to help much. Better to mount the supports on a bit of angle or a square like so:

[Jeremy]

Thanks for that idea. I've spent an hour or two looking at ways to make such an angled system, but concluded that it's too tricky. I think I will give the back to back option a try, but as a "face to face" configuration, with the rails fitted either side.

The rails will be fitted inside a fabricated channel section, capped at top and bottom to stiffen it up. The slides will be bolted either side of the "tongue" of a hefty bit of T section, which in turn will be bolted to the carriage.

The advantages this arrangement has is that it's relatively easy to make and align, plus it's quite narrow, which gives a smidgen more useful travel.

I will try and post a sketch later.

Jeremy

[Jeremy]

Hopefully the top view rough sketch below should give an idea of what I'm currently thinking of. My slim spindle (the one with the brushless motor) fits into the 38mm hole (I've not bothered to sketch the clamp) in the green part.

The grey parts are 1/2" alloy plate, the blue parts are 2" X 2" X 1/4" alloy angle. The rails will move up and down, with the carriages fixed to the cross slide. I just need to work out where best to fit the 12mm x 3mm leadscrew.

I think I can probably trim the edges of the angle behind the spindle mounting plate, and mill a slot in the back of that plate, and "just" get enough room to run the leadscrew down the gap. Then I can fit a plate at the top and bottom of the slide, to take the stepper motor and leadscrew bearing respectively.

The sketch isn't too clear, but there is about a mm of clearance between the carriages and the angle that carries the rails.

Jeremy



http://farm4.static.flickr.com/3467/3199211921_35e77d0387.jpg