[nilrods]

I am working on a Z axis for a small milling machine. Already have the X-Y part done has crossed roller rails and precision ground ball screws.

I have read through many threads on the zone about linear rails. It seems that the general best recommendation is 2 rails and 4 blocks for each axis.

I understand that 2 blocks per rail increases the load/moment carrying capacity.

I have some heavy duty 35mm profile rails, I know overkill but they were a good deal. The blocks are about 6 inches long. The Hiwin 25mm blocks(for another project) I have are about 3 inches long.

Based on the size of the 35 blocks and rails I am thinking I might be ok with only 2 rails and 1 block per rail since the contact area would actually be larger than 2-25mm blocks on 1 rail. I will no where near approach the load rating of the 35mm blocks/rails.

I understand the idea behind 2 blocks per rail is to spread them apart by a specific ratio (1.6:1 or more) so the support the load better, but many projects I see don't space the 2 blocks per rail far apart at all because it takes up too much travel space and so really don't gain much then other than just increased load capacity with 2 blocks(at least how I see it).

Just want to see if I am missing some other criteria that would require the 4 blocks, even with the 5+ inches of contact from the 35mm blocks.

Any thoughts would be appreciated.

Thanks,
Chris

[mactec54]

nilrods

Your spindle Hp will be a guide as to what you need for Bearing block support, there are some Routers that are using ( 1 ) Bearing Block 25mm series per side, without any problems, your 35mm Bearing Blocks are much bigger & have a higher load rating,so should not be a problem, is this a router or a Mill for milling steel, if you plan is to mill steel you need all you can get for rigidity

[Eldon_Joh]

spacing the bearing blocks apart is to increase rigidity.

seems to me it would be common sense to make the bearings and the rails as far apart from each other as the distance from the spindle to the rails.

[nilrods]

Mactec54,
Thanks for the reply. I was thinking somewhat along those lines also. This is for a milling machine, not a router. Primarily for aluminum but also some steel.

1-2Hp R8 spindle, so not looking to be hogging off material with large cuts. The 35mm bearing are medium preload which should also help with rigidity.

The 1 block per rail is not a cost concern as I have 2 blocks per rail. The main issue with using the 35mm and 2 blocks per rail is the height needed on the Z axis. Even butted together they will take 22" for 10 travel + height of X-Y table(~7") so pushing 30 inches tall for only 10 inches travel. Of course, any spacing further apart increases the total Z axis height by the separation amount. I figure the taller the Z axis is the bigger it has to be to be stable/rigid.

I guess I am at the point of either going with 1 per rail on the 35mm or just buying some 20mm with 2 blocks per rail. I was just thinking a cantilevered Z axis might be a good place to put those beefy 35mm to use.

Always good to hear others thoughts.

Thanks,
Chris

[nilrods]

Eldon_Joh,
Yes I understand the reason behind the spacing. I was just commenting that some of the builds I see around here, even of other small milling machines being built, don't look to space the blocks very far apart at all but still seem to function fine. Maybe because they use a much larger rail than needed?

I agree in a perfect world with no other design constraints that good spacing with multiple blocks would be more rigid and probably even allow for a smaller linear rail to be used. I guess the question I was posing was could I use a much larger rail with 1 much larger block per rail to compensate and get close to the rigidity of the smaller 2 block design.

Not sure I understood your second statement correctly, are you talking about the spacing of the rails apart or the spacing of the blocks apart on each rail or both? Either was it does sound like a good rule of thumb though I will try to keep it in mind.


Thanks,
Chris

[nilrods]

Just to add some details on the large rails. The 35mm rails are NSK SH35-GL (Super High Load Type) with medium preload.

[Eldon_Joh]

Not sure I understood your second statement correctly, are you talking about the spacing of the rails apart or the spacing of the blocks apart on each rail or both? Either was it does sound like a good rule of thumb though I will try to keep it in mind.

both, of course.

so in the case of dovetail ways, you have to have dovetails that are about twice as long as the width so you don't get binding. if you think this through, you'll find that this is entirely proportional the frictional coefficient. the 1.61:1 ratio you mentioned is sort of arbitrary. i would say its appropriate for box ways. box ways have about half as much friction as dovetail ways for side loading because the 60 degree angle of the dovetail multiples the normal force by two.

crossed rollers or balls have a coefficient of friction of allegedly .05 or something, so somehow the self locking helix angle of a ballscrew is allegedly something like 10 degrees, or 6, i can't recall.. anyhow, i don't want to argue that but, if that is the case, then you can get away with a length to width ratio of ~5 meaning yes, you can use a 25mm long bearing block to support a spindle 125mm away from the rail.. however, you will need to multiply the spindle force by ~5 to get the ball loading in the bearing block, and almost all of that is on the balls on at the edge of the bearing block. now this frictional coefficient just dictates the minimum ratio so that with force applied to the spindle, you can still move the bearing block! otherwise you can't, its frozen.

obviously you have to read the datasheet to check the loadings, but you might have to perform your own stiffness measurements, which may/will depend heavily on preload.

as you can probably figure, two tenths of slop in the bearing block equals 2 thousanths of slop at the end of a 5:1 lever acting on a bearing block

edit:changed a bunch of numbers and reworded some arguments./

[nilrods]

Eldon_Joh,
Ok. That makes sense, especially what you said regarding dovetails and box ways. And I can definitely see the friction impacting the design. Good notes to keep for future reference.

Actually I got the 1.6:1 or > from a Slocum paper regarding machine design and linear bearings, related to Saint Venant's principle.

I can see what your saying about the force being multiplied and impact to slop, makes perfect sense.

I think I will go ahead and try the build with the single block and see how stiff it is. The numbers in the spec were very large and even with the force multiplication I don't see my small spindle creating anywhere near the forces mentioned in the specs. I believe these bearing were made for large and heavy machines 10X or larger than the one I am building.

Worse case I can always add a 2nd block and lose travel.

Thanks,
Chris

[victorofga]

instead tedeious calculations, heres something simpler..

you have a distance from block to material surface.. if it were 5 in then block or blocks need 5 in length.

if it is more then need more.. don't underestimate the forces generating within milling..
when you realize one block not sufficient, you might already damage or wear out the one block.. then you have to buy another blocks, even might need rails..

so there wouldn't be wrong to use 4 blocks on beginning

if you using ground screws, then you invest very much... so saving on two block and risk new parts.. not really worth..

[RCaffin]

I cannot imagine ever building a Z axis (or an X or Y axis) with only one block per rail.
The milling forces would overwhelm a single block.

Cheers
Roger

[hanermo]

The single blocks wont work. Period.
As Rcaffin said.

Example:
I have 35 mm hiwin rails, and I am putting in 4 rails front and 2 rails back, for a total of 6 rails, and 12 blocks.
Milling machine.
Additional rails and blocks average out errors and add rigidity in resistance-to-load (for low cost- Z rails & blocks == 1700€).

Based on my 15.000 hours experience, and industry training, and 2000 machine I have looked at over 12 years full-time in CNC stuff in 5 countries...

My theoretical capacity is 12x4000 kgf = 50 metric tons force = push.
I expect to get about 300 kgf real-world usable rigidity, which is about 2 kW spindle power.
I might get, best case, 1000 kgf, which is still a ratio of 50:1, or 2%.

1000 kgf equates to about 4-7 kW at spindle.
If I get 3 kW I will be very happy.
2 kW is fine (and I am about 97% confident of that).
1 kW will be acceptable.

I suggest your values will need an adjustment re: rigidity / spindle power.

Fwiw.
My z- axis vertical spacing will be about 500 mm with latest revision, for about 500 mm work area.
Z axis bridge is == 700 kg mass, in steel.
Spindle is ISO30 - so its about 7kW max usable power.