Wow, at last someone has bothered to read build logs and think about it before posting

Quote Originally Posted by Iwant1 View Post
The frame, which is very similar to Jonathan's, will be made from 50mm steel box section and 50mm x 100mm for the top supporting rail.
What wall thickness? If you stick with 3mm, maybe 3.6mm, it's much easier to weld. Just for reference mine is mainly 60x60x3, 100x60x3.6 (x-rails) and a small amount of 50x50x3mm. If you add some pieces in the corners of the frame at 45° it will significantly increase the rigidity of the frame. I never got round to adding them on mine, although they are on the original drawing. It's the sort of thing you can add when the rest is welded to use up any left over box section.

With regards to the Z-axis rail mounting - neither of those methods is ideal. If you had to pick one I'd put the rails on blocks as that would help stop the Z-axis plate bending, but they would have to be machined accurately - not just any old flat bar. However I'd consider that a last resort since it is important, particularly withnot having profile rails on Y, to get the cutter as close to the Y-axis bearings, when measured parallel to X (the 'overhang'), as you can. Since the Y-axis bearing blocks have a finite stiffness, when a force is applied to the cutter parallel to Y the tool will deflect and this deflection will be proportional to the aforementioned distance, therefore you should minimise it as it will directly increase rigidity.
As you're using standard BK/BF bearing blocks this presents a problem, since clearly they will collide with the Z-axis plate if the spacing is reduced. Options:
1) Make cutouts in both of the 20mm plates which the bearing and ballnut blocks 'sink' into, to effectively reduce their centre height.
2) Use a longer ballscrew and mount it to the Z-axis plate (like on mine), that way the bearing blocks wont be over the other plate. Still need cutout for ballnut.
3) Make lower profile bearing/ballnut mounts, then less deep cutouts are required.

Problem is all of these options are easiest with a milling machine, or require outsourcing but to me it would definitely be worth it.

The ballnut mount looks weak to me since it's so long. The easiest solution is to move the ballscrews up, but no doubt you didn't do that because the ballscrews aren't long enough to mount at the most convenient point. You could make the ballnut mount stronger, but even then there's quite a distance (measured vertically) between the X-ballnut and rail, which for the same reason as for Y/Z it's best to minimise. I would be inclined to raise the ballscrew so it is in line with the X-axis rails and add an aluminium plate to support the non driven end, preferably linking on to the far vertical support. The strength of the non-driven end bearing mount isn't critical as it's (roughly) only there to stop the screw whipping, so there's not much force on it.

I bolted my Y-axis rails to 80x80x3mm aluminium box section. The bolts are strong enough, but in my case the box section is a weak point as there is only one. You can get the necessary adjustment on the rails for them to run smoothly by drilling the holes in the rails slightly bigger to give the bolts a bit of clearance. It would be nice to use a greater wall thickness just from a strength point of view.