What's new
What's new

Squaring up New linear guides in VF2

clayton

Aluminum
Joined
Aug 19, 2011
Location
Houston Texas
Replaced the X and Y linear guides in my VF2 with the table and saddle still in the machine. A tech I am friends with told me he’s had to do it that way before. My forklift is down and couldn’t use it to remove those components.

Long story short got it done one side at a time. Now I’ve got a bow in the guides when indicating X on a T slot. Zero the indicator over a bearing pack, then move to the other one on opposite side of travel in X and it will zero there, but between it’s off 3-5 thousandths. Same result with granite square.

Has anyone had success squaring the linear guides in a machine with the table and saddle installed?

Thank you, Clayton
 
How did you clean the mounting surfaces when doing the installation? OEM replacement guides and trucks, or only guides, or non-OEM? My understanding is there's supposed to be an accurately cut raised ridge the guide is to be pushed against during mounting, along with retaining tabs to further constrain the rail, as well as the through bolts.

When you say a granite square, you used it as a straight edge or just to check perpendicularity of the two axis? I'd want to use a known-good reference edge as long as the table travel for checking, rather than a T slot or short reference.
 
How did you clean the mounting surfaces when doing the installation? OEM replacement guides and trucks, or only guides, or non-OEM? My understanding is there's supposed to be an accurately cut raised ridge the guide is to be pushed against during mounting, along with retaining tabs to further constrain the rail, as well as the through bolts.

When you say a granite square, you used it as a straight edge or just to check perpendicularity of the two axis? I'd want to use a known-good reference edge as long as the table travel for checking, rather than a T slot or short reference.

I cleaned all the surfaces completely when I had each linear guide removed.

Haas only sells the linear guides and bearing trucks together as matched sets. I replaced everything completely.

I installed the linear guide against the casting edge and lightly turned the cam bolts that hold the guide against that edge.

The local Haas tech brought over a granite square to check the perpendicularity. However, sweeping the T slot or machined surface on the table should be within .001. The granite square reflected the same result as the T slot.
 
Did you do a "pre-swap" check of the straightness and perp before changing parts? It would help to know if any bowing or perturbation was present before the changeover.

In any case, things you can do now are confirm the mounting feet are uniformly loaded, that the Z column is true, and from there drill down to what is causing the error. You may have to bite the bullet and pull the table somehow to really check the Y saddle for issues, since the X rails would be all that could cause a bow assuming the table itself hasn't warped.

Are you sure the X table isn't distorted on it's own? If I understood correctly, you have a bow in the table in the plane of the top view? That seems like an enormous amount of stress to have to induce through the linear ways. I'd think it's much more likely the table warpage is bowing the rails.

Maybe post a sketch of the how the error shows on the table to remove any uncertainty.
 
Think of it as a rail road track with a slight snake in the path of travel. Thats what it looks like when you indicate across.

Not a bow or deformation in the table.
 
Think of it as a rail road track with a slight snake in the path of travel. Thats what it looks like when you indicate across.

Not a bow or deformation in the table.

A snake or a curve? From your original description it sounds like a simple curve, not a curve-return curve (what I would imagine a "snake" looks like).

With a curve, that would be the rails or mounting guide edge on the Y saddle being distorted, and I'm still not sure how you'd check or fix that without pulling the table.
 
Did you check the level of the machine before you disassembled the machine? Now even though the machine isn't running straight I would check it. It should be level and aligned as close as you can get it using a .0002/12" level. Everyone thinks Liner guide ways are straight but new ones are like a banana. If you camming over to a solid shoulder and you got it clean.... You may have had a burr or a small piece of crud that is messing you up. If it's level you may have to pull it apart and try again. I am assuming you did it to the solid side and not the free no shoulder side. I am not familiar with Hass, but have worked on new machines built in Taiwan. Did Hass provide you with a procedure instruction sheet to follow? I'll cross my fingers it's out of level so it will be an easy fix. :-)
 
Did you check the level of the machine before you disassembled the machine? Now even though the machine isn't running straight I would check it. It should be level and aligned as close as you can get it using a .0002/12" level. Everyone thinks Liner guide ways are straight but new ones are like a banana. If you camming over to a solid shoulder and you got it clean.... You may have had a burr or a small piece of crud that is messing you up. If it's level you may have to pull it apart and try again. I am assuming you did it to the solid side and not the free no shoulder side. I am not familiar with Hass, but have worked on new machines built in Taiwan. Did Hass provide you with a procedure instruction sheet to follow? I'll cross my fingers it's out of level so it will be an easy fix. :-)

Richard,

The machine was level before procedure was performed.

Haas has excellent documentation for service and it’s the same that their techs use.

The rail has no different edge to it. It has a slight chamfered edge for relief where it sits against the shoulder for the possibility of some dust being there. I did wipe clean vigorously.

I will follow up with pictures of the indicator activity I am getting shortly.
 
This images are from left to right. +- .001 full travel in X.
Left side maxed out in X travel
00c40708af18758c2cf0a8414d061b18.png

Over left side truck
5589f7b875ce4148fcce62b93754d771.png

Middle of table
865ba2c81cf25725f84b823a9fbbcaee.png

Right side over truck
d89753f55b51e50fb5ecf7028d5f314d.png

Right side maxed out in X travel
0eb359189b1099668a6f19e07c86a68b.png
 
I would really like some independent readings, rather than trying to use the T slot as the reference. If you know anyone in your area with a good scraping reference like a camel back or a long granite parallel, indicate that in so it's true to the travel direction and straight up and down (to ensure no false readings). Then run the travel and see if you still get the same error path.
 
That’s what the Haas tech did. Same reading was shown.

Well, crap. One thing you can try that's some extra work but should pinpoint the fault is remove the rails and remount them 180 degrees, so the front rail L-R now becomes the back rail R-L. If you get either the same indicator readings, or the reverse it should tell you what's going on. Same readings means errors of mounting or reference surface on the casting, 180 readings means the rails are at fault.

At least, that's my logic, perhaps Rich or someone can find fault or have a better description.
 
Linear rails have a reference side that was pushed aginst a reference straight edge when grinding them. Dumb question but are you sure the rail ref side is against the ref shoulder in the machine? Same with the blocks. After replacing the rails and blocks in my mill it was straight and square, with some shimming of the blocks.
 
Replaced the X and Y linear guides in my VF2 with the table and saddle still in the machine. A tech I am friends with told me he’s had to do it that way before. My forklift is down and couldn’t use it to remove those components.

Long story short got it done one side at a time. Now I’ve got a bow in the guides when indicating X on a T slot. Zero the indicator over a bearing pack, then move to the other one on opposite side of travel in X and it will zero there, but between it’s off 3-5 thousandths. Same result with granite square.

Has anyone had success squaring the linear guides in a machine with the table and saddle installed?

Thank you, Clayton

When you say you "got it done one side at a time", I'm puzzled. The cardinal rule of assembling something like this is that nothing gets tightened down permanently until everything has first been snugged up in its 'natural condition' and the test run done. You can probably feel a tight spot as you push the table along manually before connection is made to the ball nut.

NFW would I do that without removing the table so I could work at it in comfort and without taking shortcuts.

Are the trucks preloaded or what is the case of new? I'm imagining fabbing up some kind of temporary linkage between any 3 trucks (two on one rail, one on the other) and attempt to indicate for change in the length or angle of hypotenuse as the thing is rolled along. Not sure what that would look like :D. But you'd have to have the table out of the way.
 
Blocks are preloaded on the rails like ball screws, which is why they are only sold as a set. You install the rails, then loosely attach the blocks so you can check alignment, then shim to correct any misalignment, then torque the blocks down, then the ball screw nut.
 
By Haas design, these are proprietary components made in pairs for them. They are precision ground match sets to eliminate any need of shimming. Ordinarily if this were a rail and truck that you can buy from the vendor individually it would be necessary to shim etc...

Unfortunately you can’t buy just the bearing truck or rail. You are committed to the complete axis replacement.
 
Need of shimming has nothing to do with the rails or blocks, it is how well the machine castings are machined. Shimming is normal. The machine I replaced the ways on is a fixed table so alignment was a little more important, and amazingly adjustable. My understanding is the better grades of accuraccy you can get rail and block sets only apply to the sets made, they are as accurate as stated to each other, not a set dimension, hence they are probably made in pairs for our needs no matter what.
 
A couple of thoughts. As Miland said. Im not a big fan of the tee slot tests and would like to see you use a granite straight edge the same length as the travel. How about you test the other tee slots and see if the repeat the results. Or switch sides of the slot and see if you get the opposite reading. If that is the same, then the only thing I can think of is do As David or one of the others suggested. Try the other one and see what happens. If you get the same results that the edge the rail goes against is off. If the Hass Tech was there and can't figure it out...what does he think? Maybe he needs to call the home office and see what they say. Why did you replace the rails? Have this miss-alignment problem then?

Another thing, did you see the machine leveled? Maybe you need to re-check it. Never assume, prove. A simple way to check it is set a level on the table and see if it changes as the table moves. a Slow feed.
Set another mag base and indicator on the Y and watch it to be sure it isn't moving. Maybe some control issue.

One company I worked with in Taiwan had adjustment wedges that used setscrews. The other one we scraped and stoned the edge. But when we did it the table was removed and we had the granite straight edge sitting between the ways. One last thing. You may have to rent a fork lift as the rent is probably cheaper then the down-time of the machine. Rich
 








 
Back
Top