[whimsical]

Have started a build of a 1600 x 900

I had decided on a bearing spacing of 400mm for X and 300mm for Y and Z but would like some input as to what others may think. This gives a table size of 2000 x 1200.
Z height to be 275mm Beam will be 300mm high and 200 deep.
Hopefully dmalicky will chime in as he seems to have a vast wealth of knowledge on this subject.
Have i gone too wide or not enough. The machine is targeted towards very very hard woods and ally.

Machine basics
Hiwin 20 rails for everything HGH 20CA blocks
2 of Mod 1.5 racks with 30 tooth pinions for X and 1 set for Y
THK ground ball screw for Z, 1604 2:1 belt drive
4 of Dmm servos 400W with 5:1 reductions.
ESS board
2500W 60V power supply
2.2KW water cooled spindle
Hitachi VFD

Table is made and awaiting the machine shop to make some nice true slots for the X rails. All components are in hand or in transit from the various suppliers.

I have a very rough sketch done in Corel.
The design incorporates an A axis that is up to 370mm diam by 1500. It shows a B axis but that is for the future if and when i want to play with one.


Attachment 267294

[JerryBurks]

That looks like the concept for a very capable machine. I like it.
Hard to say what the *optimal* bearing spacing is, since it is a trade-off between rigidity and motion range. The further you space it apart the lower the tilting load will be and consequently the contribution of the bearings to flex.
But at some point there is diminishing return and you just lose too much usable movement. Without deep scientific study I would say if you space them about as far apart as the maximum leverage length between the bit tip and the bearings you are probably good. Given that you are using high quality rails you can probably go a bit more compact. For the Y-axis you can also help by moving the rack/pinion as low as possible or using a double rack to eliminate the twisting load and resulting racking on Y and Z bearings. But the latter is probably a bit overkill.

[wizard]

As Jerry points out there are trade offs. For the X axis the bearings supporting the gantry ought to be spaced apart as high as the gantry is above the table. Ultimately you end up compromising in some aspect or another to keep costs under control. The trick is to avoid to closely spacing bearings while managing costs.

[whimsical]

Thanks for the comments guys.
I know the trade offs but it would be nice to get some hard numbers on the trade offs, like does going from 300 mm to 400 add enough stiffness to bother. I suppose i could search for the stiffness numbers and do the maths but not sure where to start. Thats why i was hoping dmalicky chimes in he has a good understanding. I have read lots of his posts about beam stiffness but he has never put any hard numbers on the bearings and/or the spacing that i have seen.
It is a bit of a moot point now as i have decided to go to a 6:1 ratio, speed the servos up, better inertia ratio and a bit more resolution, this has made reducing the X spacing unlikely as I am using GT2 belts and 6:1 gets big.

[whimsical]

Did some more research on this and while trying to work out some of the examples in the Hiwin book i remembered dmalicky had a spreadsheet and i just needed to plug in my numbers, lucky me.
The rails and blocks i am using are nearly twice as stiff as what he used so just plug the new numbers in and i was quite surprised. He accounts for y and z axis but not x, then i realized that when the cutter is near the table the height above the rail is not much so has little influence. As z raises the its influence is less but x becomes greater, almost cancelling each other out. To stay in proportion to the spreadsheet then x needs to be about what i guesstimated, wel probably less than that maybe 320mm but i am happy to lose 400mm of x travel.
Happy chappy now, especially with the numbers below, 0.00006" or 1.5 microns. Just need to get the beam built and test it for twist and deflection.

Attachment 267744

[JerryBurks]

.. To stay in proportion to the spreadsheet then x needs to be about what i guesstimated, wel probably less than that maybe 320mm but i am happy to lose 400mm of x travel. ........

Sounds like a plan. But don't get too hung up with the x-bearings. On my machine most of the flex is in the y and z-bearings. The distance (i.e. leverage) between the bit tip and the y-rack or screw can be substantial and some spacing between the bearings is even more important. You have much better bearings than I have with the round supported rails but I would check that out as well. Especially the lower rail bearings get higher forces if the rack is in between the rails.

[whimsical]

Jerry
Its more about understanding the numbers, i am a very curious person and like to know what is happening.
For my machine x is becomes more important as i have the a axis and it is quite large hence the 11" beam height. If malickys' numbers are correct then i am looking at numbers so small the whole question becomes a moot point but it is interesting, well to me at least. I tend to take the whole thing as a challenge, next up is to make a beam that lives up to the stiffness of the bearings so a few experiments are coming I think.
Maybe being retired i have too much time on my hands
Mike