[Mwmx54]

I am just finishing up my zx45 cnc conversion, I have another thread about it, but my question is if anyone has a solution to the flex of the dovetail ways in the z axis, in about the middle of the travel, the dovetails flex in quite a lot, which as a manual machine wasn’t a huge issue, as I could just tighten up the locks on the gibs and take up the slack where they flexed inward, but as a cnc it is going to be a bit of a problem. Has anyone else run into this issue? Do you have a solution, like bolting some thick steel plate the outsid of the column down each side? Or filling it with some epoxy granite? I’ve read into filling it quite a lot and there seems to be some mixed opinions, would this be strong enough to add any structural integrity to the column?
Here’s a pic of the mill currently. The flex is the worst right behind the head saddle in the pic, right in the center where there is no support like the top and bottom and has.
Thanks

[ninefinger]

Lots of solutions, just picking the right one for your situation is the hard part. Everyone has an opinion on which solution is right, I'll list the ones I know of:

1. Linear rails - mounted to the flats on the column and the dovetails milled away for clearance. A few have done this and claimed great success, but requires a large mill for modifying the column, and maybe even a break for the head as it reduces friction to the point it can fall under its own weight, also fairly costly in materials
2. external mounted ballscrew and close up the column face (and optionally fill the column with granite epoxy or whatever). Probably the least difficult option on a new build if Z parts hasn't been made yet.
3. build an "exoskeleton" for the column to stiffen it up, support the opening from closing in - never seen it done but in theory could help, would be relatively easy if access to plate material and plasma cutter or waterjet was reasonable (very rough, non engineered image attached). I'm kind of tempted to try somethin like this on mine from the point of view of improving what is existing, without blowing it all up for option 1, or rebuilding for option 2.
4. Use a counter balance. This reduces the need to have the gibbs as tight, thereby reducing the amount of flex introduced due to clamping forces and the head weight pivoted off the dovetails. Also relatively non-invasive and can be added on to any build.
   

[ninefinger]

Further to above point 3 - I would want to measure the deflection at the sides before doing this - if all or most of the flex is in the dovetails this won't help much (but it looks neat )
.

[Mwmx54]

External mounted ball screw would have been a good idea, That never crossed my mind, I have already built and installed the motor and ball screw though, not that I couldn’t do it over, just that I don’t want to, something about redoing stuff is just hard for me, it sounds dumb but I’ll do ten times the work just to not have to redo something. I did consider the exo skeleton thing, as well as linear rails, but I don’t have access to a second mill which I think would be necessary to do correctly. I measured the deflection a few years ago, and it is significant, I don’t remember exactly, but I think upwards of .03-.035” it’s a lot. I just don’t know that building a mold around the ball screw and filling it with epoxy granite or cement would add much stiffness, however an exoskeleton like what you designed probably could, maybe 5-6 ribs of 3/4 steel then welded together with more 3/4” steel that fits against the column and bolted on between each rib. I like it plus more could be added if needed. But I do like the EG or cement idea as it would be the easiest. I’ll have to think.
Thanks for the ideas!

[Stigoe]

I did consider the exo skeleton thing, as well as linear rails, but I don’t have access to a second mill which I think would be necessary to do correctly.

If you have a technical school nearby, check if you can get them to do the milling for you. That's what I did when I installed linear bearings on the Z-axis of my BF25 (Weiss WMD25).

[ninefinger]

I may try 1 or 2 of the exoskeleton members (think C clamp working in reverse) to see if it creates any noticeable improvement. If it does then I'll add more. I just need to relocate my home/limit switch to do it..
Those pieces don't need to be from plate - that was my 3 minute drawing. They can be fabricated up as nothing about them needs to be precise. An inside and outside bar with joining girder or thin sheet structure can work just as well. We're just aiming to add some structure to support the column. Then screw a stud into the column and pull against it to support / spread the column open.

[Mwmx54]

You are probably right, I bet one exoskeleton brace would probably fix 90% of the problem, the other 10% would probably need another 3-4 braces, but I bet one placed right in the middle would do the trick, I’ll probably go that route for now.

[Mwmx54]

I made up a brace, quite a bit of work cutting out a chunk of 1” steel plate, I cut it out with a torch, then had to go back and forth getting the draft angle of the casting just right, and test fitting the piece over and over, I ended up with about 1* of angle being pretty close, then fine tuned it with a grinder, and I had to chisel out about 1/4” of filler to get down to the base metal, there must be 20 lbs of filler on this machine. Adding 2 more of these braces, and attaching them together would probably stiffen up the machine quite a lot, but I think I’ll stick with the one for now. I still have a little more adjusting to do with a grinder to get it just right, but column flex is far better than it was.

[ninefinger]

Thanks for the update, great to hear it worked.
Now I just need to make mine. I started, and cut up some 1" square tubing to weld into a brace, but ran into my poor welding skills and decided to stop before I ruined all the parts, I'll bring them to a buddy to weld.

[bluestratus]

Thread is a few years old but I am interested to hear how it works and see any more pics to understand it

[Mwmx54]

Well I kinda forgot about this thread, but if anyone is interested, I made a very amateur video showing some of the things I’ve done to the mill, I only did one brace on the column, and it did indeed fix most of the problem, I used I think 7/8” plate steel, scraped all the paint and bondo off the column, and just worked the brace with a file until I had a good fit. Here’s the vid, I also made a new base that moved the column back about an inch. Here’s a pic and a link to the video. Still considering the linear rail for the z axis, building a whole new column is also not out of the question. I really need to get the rpm's up though, so that's next in line. either a 3ph motor swap, or one of those Chinese atc spindles.
Video: https://youtu.be/Lr62hf2b81c
Don't make fun, I was kinda rushing to get it all together so I could start on my auto knife build.

[Enforced227]

My attempt to make Z-axis stiffer:

Base plate 40mm steel,
Back plate 40mm steel.
Back support ribs 15mm hardox 500 tied together with 3 20mm plates in the back.

It’s locked and adjusted with 8 - M24 and 8 - M10 bolts

The M24 bolt is drilled through the center to allow the M10 bolt to go trough and thread to the column.

The M24 bolt is adjusted to seat against the column and measured with a dial indicator between the support and column to prevent tram issues/movement , and then the M10 gets tightened to lock the assembly as seen on picture 3.









Skickat från min iPhone med Tapatalk

[joeavaerage]

Hi,
is there any reason that you cant unbolt the column off the base and make a new column?

My machine is made of a steel frame, being mainly two 'L' shaped 32mm steel plates with separators bolted between them.

Craig