Mill ways alignment problem

[Janger]

Hi,
A buddy of mine -Marius - is working on a CNC mill conversion. He needed to enlarge the lead screw area to fit ball screws in. He gently used an angle grinder to open it up over a few days to minimize heat. However the front of the ways are now 0.004" wider than the rear. The casting moved on him. See pictures. How should he fix this? I had a brute force idea to simply run a dove tail cutter and straighten it out by taking off 4 thou from front to back - on both sides I suppose. This is a hobby sized desktop mill - it's a king but very similar to the the usual CX602 from craftex or the grizzly models. The ways don't appear to be hardened. See the red lines - the front one measures wider than the rear one. The other issue is the gib no longer fits correctly due to the changed dimension on the ways.

Advice please?

 

[Tom Kitta]

Problem is did it stop moving. Maybe wait some time to make sure it is somewhat "stable" and then cut to make it even.

He picked too big of a ball screw - for my machine RF-31 clone I went with 16mm same as many others even through it looks wimpy - but pp did calculations and it is more then good enough. So if his machine is smaller maybe 12mm would be totally fine. Then again what is done is done.

 

[RobinHood]

What Tom said.

Here is what I think happened. The center support rib (red arrow) was cut out causing the Center of the ways to “collapse” inwards (blue arrows). The ends are most likely still the original width because they are supported by the end casting walls.

So, since it is already bad, I would set up in a mill based on the parallel side dovetail indicating off the front & rear only (ignore the center for alignment purposes). Use a dovetail cutter to mill that side back to straight, using very shallow DoC and see how things behave. I think you will be able to get it straight again. Then do the same on the tapered (gib) dovetail side. Again ignore the center section and only indicate near the ends for alignment.

If the gib had enough “meat” on it to begin with, it should be reusable (it will just be deeper in the slot). If it was marginal from the start, you need a new gib. Scrape everything in, and the machine should be good to go.

Did the saddle move, because I see grinding marks on it? If yes, it is time to rebuild it as well.

 

[Mcgyver]

Ouch! its clear a lot of grinding has been done on the saddle, its likely got issues as well.

Is he sure it moved? Some of these can be horrendous.....its not beyond possibility that all or part that is from the factory. Although I suppose that doesn't change the current reality

I doubt there will be further movement. for example, afaik current engineering knowledge is that the need for aging castings to stop them moving has proved untrue; They (or any material) can move when you change the equilibrium of internal stresses (by heat or machining), but they don't keep moving.

Incidentally, the temperature from machining (in this case grinding) is predominately a function of cutting speed. It doesn't matter if you use flood coolant or waiting for it to cool (that just stops heat build up). The temp where molecule of material meets a molecule of abrasive depends on speed...and its that temp that can mess things up. Just ask the guys who tried to grind their planes instead of scrape them

If it did move, I'd be a betting man that it didn't move in one direction. Therefore the fix is not likely to be running a cutter along one plane. If it move there will be warp, I'd bet.

View this as a recondition exercise; you've got mating bearing surfaces that don't mate and need treatment. First thing to do in any reconditioning is to survey. Quantify all of the geometry. Then you work on a plan, a sequence to bring the geometry back to where it needs to be relative to each other, done via scraping. There is just no way of knowing what the problem is or what the solution is without doing this. It could be a big job correcting it, but that’s somewhat unsurprising as he's got himself into a mess if the castings have warped.

Cripes its even a big job typing out how to survey it.

If you assume the base is good (rule 1, never assume), as a starting point, trying printing the bottom horizontal surfaces of the dovetail on a surface plate. if its flat, try bluing it and printing the based to see if the two horizontal surfaces are co planer. You really need an angled reference flat to get in there and check if. Then try measuring across pins placed in the dovetail. No calipers here, tenths micrometers or indicators. In my scraping article (ran for 2 years in HSM) I covered in detail reconditioning a mini mill x/y/table assembly, that would give him all the info he needs.

The mating of a machines bearing surface is fundamental the accuracy and function of the machine. If they've been messed up there is a path back, but its onerous. Is there a shade tree solution? don't know, maybe, but I doubt it without big compromise/risk. I don't think that way, not because of being a complete stickler (maybe I am a bit, but if so, its my way making sure projects come out right), just because there is only way to insure success with this and many to make it worse

 

[Janger]

Awesome replies thanks.