Another Tramming Question

[ajb3015]

So I know there are hundreds if not thousands of threads and videos on tramming, but I have yet to find an answer to my question.

I have a precision matthews PM-25MV and I'm trying to tram the head. I have both a spindle square, as well as several indicators and arms to choose from. And I am aware of the typical process of setting up the square (or indicator), zeroing, sweeping side to side and adjusting the head as needed. But what I have never seen, or found in any reference, is adjusting Z-axis height and checking tram at a different height. Does anybody do this, or does everybody just assume if the tram is good at one place it's good throughout Z-travel? In theory, if your column is straight and square to the table, it SHOULD be good throughout Z-travel. Even if the column isn't square to the table, by squaring the head to the table it should stay square to the table, but translate slightly in X.

However, what I see is if I tram with the head as close to the table as possible, then raise Z ~8" (as far as my indicator arm will reach without reducing the swing radius), then tram is out ~0.004". After several attempts I finally raised Z ~4" (1/2 the available travel), and trammed there. Now when I move Z back to the low point, It's a little over 0.002" out, with it indicating that the head is rotated slightly counterclockwise. And when I move Z to the top, it's 0.002" out with it indicating that the head is rotated slightly clockwise. This matches the ~0.004" variation from high to low, and it appears as if the head is moving in an arc with the center point of that arc somewhere in space to the right side of the machine.

Throughout this entire process I have had both X and Y axes locked in place without moving them. And I lock Z at every position while tramming. I have also checked for play in the Z-axis gib, at all 3 locations mentioned above, and it's solid. I trammed first with the spindle square, but it only works with the head close to the table. Then swapped to the indicator on the arm to verify without moving Z, then moved Z. I continued to use the indicator for all following measurements.

I don't have any machinist squares (but have them ordered), so I haven't checked if the column is square to the table. And it seems to be out of tram in Y as well, but I haven't paid enough attention to see how far out it is. Shims are also on order.

Based on the above, here are my questions:

1. does anybody bother checking tram at different Z heights, or just assume it's good?
2. am I overthinking it by checking at different Z heights?
3. does my process seem sound, or am I possibly making a mistake which is distorting the measurements or giving me false readings?
4. If my process seems sound, and my measurements seem accurate, does my conclusion that the head is moving in an arc sound plausible/reasonable?
5. If my conclusion is plausible/reasonable, what are the possible causes? With no play in the gib, is it reasonable then to assume the column way was somehow ground in an arc?
6. If the column was made wrong, is there any simple solution that wouldn't require complete disassembly and re-grinding or scraping the way?
7. Am I overthinking this entire thing?

I'm sure there are many here with far more knowledge and experience than me, and I look forward to hearing your thoughts. Thanks!

 

[Alexander McGilton]

If you are going to go about scraping. You will need a method to ensure squareness at all heights of the column. This video will shed some light on some of these methods. Particularly the mid section of the video.

 

[Alexander McGilton]

What you are thinking of is vary much possible. That the column is curved or warped from factory condition. Surface grinders tend to have a horizon after years of use. Possibly one or more thou per foot. And this logitudinal wear is present in more in one axis than another. Resulting in decent looking surface finishes, but long sweeping errors.

Your machine could have been manufactured on one one of said worn machine. And the head is ridding up and down curves.

So what can you do about it. Hand scraping is an option. Though If the kind of work you are expecting from this machine is forgiving enough, you could get away with shimming the column to the mid point of error. Far less effort and expense than scraping.

Be careful of what squares and straight edges you get for that machine. A good set of reference equipment needed to scrape that machine in could be valued at more then the machine it self.

You may see a granite square and a surface plate on the web that look nice. Though may have several thou of error. Take those flatness and squareness lables with a grain of salt. I remember my master cast iron frame level form fowler had ten times the squareness error then advertised. A good scraping project to take in and self reference. But not something good out of the box.

Remember that scraping and bluing is a profile duplicating process, and not a flattening process. So you can quickly be chasing your tail if you start with a poor reference surface.

 

[ajb3015]

I have no intention of attempting to scrape it myself. I am aware that scraping is a science/art/skill that takes considerable practice to do correctly, and do well. And as you said I'm not sure the machine is worth the cost of the tools, or the effort, to scrape it in.

I'm not planning to make anything with real tight tolerances, and at this point I'm contemplating living with it as it is. After reinstalling the vice and accounting for the workpiece and tool length, the head will never be at the bottom of Z-axis, like it was when tramming with the spindle square. And I don't plan on making anything large enough, or using tools long enough, to force me to put the head at the top of the column. For 99% of what I want to do I can lock the column in the middle where it's square, and use the quill instead of Z. I'll loose some rigidity but I'm not doing any heavy cutting either.

 

[extropic]

I have no intention of attempting to scrape it myself. I am aware that scraping is a science/art/skill that takes considerable practice to do correctly, and do well. And as you said I'm not sure the machine is worth the cost of the tools, or the effort, to scrape it in.

I'm not planning to make anything with real tight tolerances, and at this point I'm contemplating living with it as it is. After reinstalling the vice and accounting for the workpiece and tool length, the head will never be at the bottom of Z-axis, like it was when tramming with the spindle square. And I don't plan on making anything large enough, or using tools long enough, to force me to put the head at the top of the column. For 99% of what I want to do I can lock the column in the middle where it's square, and use the quill instead of Z. I'll loose some rigidity but I'm not doing any heavy cutting either.

When you say "side to side" are you talking about the roll tram only? Is your machine head adjustable in nod also or is nod fixed?

 

[RJSakowski]

From my perspective,what you are doing makes sense. What I didn't see was the diameter of the swept arc. If it is 10" then the out of tram condition isn't that unreasonable. I beli3eve that what you are observing is indeed curved ways on the z axis. What to do about it is another matter. I would see if it affects my work before taking any drastic action.

 

[ajb3015]

When you say "side to side" are you talking about the roll tram only? Is your machine head adjustable in nod also or is nod fixed?

yes, "side to side" meaning roll, or tilt; setting the head perpendicular to X-axis only. I have no nod adjustment, to correct that I will have to shim either the head, or at the base of the column.

From my perspective,what you are doing makes sense. What I didn't see was the diameter of the swept arc. If it is 10" then the out of tram condition isn't that unreasonable. I beli3eve that what you are observing is indeed curved ways on the z axis. What to do about it is another matter. I would see if it affects my work before taking any drastic action.

the diameter of the swept arc is ~6". I started there as it allowed me to check a consistent arc while varying Z (low, mid, high). If my results had shown that tram was consistent regardless of Z height, I would then move the indicator out to a much larger arc diameter and refine the adjustment if needed. Unfortunately I didn't get past the first part.

I've had the machine for a few years but never checked tram before because I was just getting my feet wet and getting familiar with the machine. I've never noticed any issues with the parts I've made, the critical features were all where they needed to be, but I wasn't chasing thousandths on anything either, so "close" was always good enough.

But I just recently bought a rotary table for a new project and figured it was time to get serious about accuracy, so I pulled the vise off, cleaned everything and decided to tram it. Guess I opened a can of worms I wasn't ready for lol

 

[extropic]

My recommendation is to survey the tram at 1" increments of Z travel (13" max according to the PM specs). Unless you have a larger precision flat disk, I would swing the DTI in ~7" circle ((table is 7-1/8" wide?) and record all 4 readings (X+, Y+, X-, Y-).

If you don't think that the survey is a good idea . . . carry on and good luck.

Edit: I should have included that using a .0001" reading DTI would be important.