tramming

I started a consulting gig today at a start-up medical device. R&D, prototyping, fixturing and simple tooling. First thing I did on the mill was tram the head. It's a relatively new Sharp knee mill with low hours and looks to be well taken care of. A vise was mounted in the center of the table so I trammed the right side. It was out about .005", so I fixed it. Then I decided to check the left side, it was out about .008"!! I went back to the right and it was still zero. I closed the kurt vise and swept the top, it was out about .0015". The knee and table were locked when I checked each side and vise and when I trammed the right side.

In 20+ years of machining, I've never experienced a mill this far out on one side when the other was ok. Is this normal for Sharps (never used one before)? Suggestions?

Thanks

You're going to have to be a lot more specific in your description. Tramming one side and then the other and finding error is making no sense, at least to me. Could you describe in different, or better, terms, exactly what it is you've done and what these numbers mean? If you're referring to getting the head of the mill at right angles to the table, if you're getting different reading from one position of the table to another, it sounds like the knee is shifting as the table changes position side to side. That's as close as I can come from what you've said. On the other hand, were you trying to dial in your vise?

Harold

SOP for tramming a knee mill head. .0005" test indicator attached to quill nose. Spin the thing around with the stylus touching the table top and about .008"-.010" of travel. Sweep diameter is slightly less than table width. Quill is extended less than 2.00".

Me neither!

Could you describe in different, or better, terms,

See above for methodology. When the stylus is on the right side of table and at 3:00 position of swept area, it reads zero. When I move it 180 degrees (9:00 position) it reads +.005" I adjust the head until all four quadrants read zero. Means the quill is perpendicular to the table.

Now, I move the entire setup to the left side of the table and sweep again. I don't recall the positions. but the TIR was .008". Perhaps 3:00 position read .002 and the 9:00 position read .006", for example.

If you're

Yes. I'm getting different readings at different table positions. The knee may be shifting, but it was locked down. Could be table weight extended to one side of the knee or the other causing the shift. I hope not.

On the other

I was just checking as a quick and dirty method to confirm my other funky readings.

Thanks,

That really screws with my head! Unless the knee is shifting, I can't come close to a reason that you'd be experiencing what you are. Might be a good idea to check the gib on the saddle along with the knee, but if you're locked down in both places, can't imagine that would be the problem.

Makes me wonder if the mill table has aged and twisted, but it seems that would show up in how poorly the table moved. It might be a good idea to put an indicator on the table then move it (the table) along to see if the table is parallel with the ways. Stranger things have happened.

I recall an ExCello 4 head boring machine that could be set up dead concentric and yield eccentric bores, but it was by a couple tenths, not a few thou like you're experiencing.

I'd be keenly interested in hearing what you find, if you do. Mean time, good luck, for something like that can drive you to drink. Make mine scotch!

Harold

Put an indicator on the column and see if the knee moves relative to the column when you traverse the table. It may be shifting even though you have it locked down.

Rex the Wrench

skuke wrote:

Thanks!! I like that idea. Of course, I'm not sure what I'm gonna do if I find a lot of movement.

Hmmm, first I better do some math and figure what "acceptible movement" would be. I've never adjust gibs on the knee/column. Sounds challenging with all the weight involved.

Since I'm just a "consultant", maybe I'll just make my parts on the section of table I know to be good! They're small parts and will easily fit in the work envelope, ...for the most part.

Thanks

What you really want to do is cut the entire top of a piece of scrap with a modest sized end mill, so that the newly cut surface is actually parallel to the plane of X-Y motion. Then, you tram the head to be perpendicular to that surface.

Tramming the head to be perpendicular to the ways of the vise (I think that's what you describe above) is risky, as the vise may be twisted or sitting on a chip. Since the vise ways aren't parallel, you have no way to know which one is right (most likely, neither one is parallel to the X-Y motion.) Now, if it turns out that the top of the machine's table isn't parallel to the X-Y plane of motion, you have a real mess! If it is just that the vise was forced down on top of a chip, it may recover if you clean up and reinstall the vise. If the vise has been sprung, you could re-grind it or just replace it.

Jon

OK, I gather there's no vise involved. You need to isolate what is shifting. There seems to be only two possibilities. One is that one of the ways is loose (could be any one, knee, saddle or table) allowing the part to end up pulling up off the ways. Many machines are poorly designed, and if you run the table to the extreme ends of travel, the overhung weight tilts the short end of the table up off of the ways. You may actually be able to hear the oil film break free with a "schlurp" when this happens.

On a well-worn amchine, like my 1938 Bridgeport, the bottom of the table has been worn concave, so the table actually travels in a sort of arc instead of a straight line. But, you shouldn't see this on a "relatively new" machine. But, you could also have warpage of the table, due to the overhung weight and possible use of a green casting. Can you find a decent straightedge. With the amount of error you describe, I think you could easily see this with a 3' straightedge, ground machinist's square or similar tool. If you put it on the top of the table and it rocks, you have your answer.

(One way this can be caused to happen rather quickly is putting heavy fixtures, vises, rotary tables at the end of the table and just leaving them there. It may only take a couple of weeks for a rotary table to permanently warp the table this way.)

Jon

If gibs are not snug along their entire length I believe it would be possible for the knee to tilt even though locked enough to prevent movement along the axis. This would seem quite likely if there is a lock only at one end or only one is really working correctly.

Don Young

Sorry, if I understand you correctly, that's just a lot of extra work! If my vise IS sitting on a chip and I held and cut a piece of scrap then trammed to that newly cut surface, I'll still make bad parts with the vise.

I believe you should tram the reference surface, the table itself.

Also, if the head is out to begin with (it was by .005") I'd have a difficult time making a flat surface. The bottom of the endmill will be at "some" angle (~.035 degrees) to the scrap as I cut. Either the trailing edge will drag and leave witness marks or the leading edge will cut everything. Then, when I move over for the next pass, I'll get a step.

I didn't tram to the vise. I merely checked it out of curiosity because of the funky reading I got on either side of it.

Thanks,

Absolutely correct!! Anything less is a waste of time.

Yep!

Harold

This is what my "gut" is telling me. But I think the knee, not the saddle, is moving.

Yes. Assuming the table top is ground parallel to the ways.

I believe this is a machinist "wives tale". If it were true, then I should crank my Y axis all the way to the column, center the X axis, and move the ram all the way back whenever the machine is not in use. And then I'd still get warping after "a couple of weeks"! The head, table, motor... all weigh far more than a big rotary table with a vise mounted on it!

Yes, everything deflects regardless of the size, material and "beefyness". It may, for all intents and purposes be immeasurable, but I know it is deflecting. But that's why I tram in the area where I'll be working!

trailing

?? Here's what I think. ($.02) (o:

In my theory: The plate that Jon produced should be left in place on the table. That makes it, a

In real life: Now ... having said all that: I only tram the head when it's been moved and even then I tram it with ref to the existing table surface. I always check the attitude and seating of my vise when it is installed and most often, I will re-machine the seats in my soft vice jaws which gives me part holding and indexing that is true to the axis of table movement in all respects. (As per Jon's idea. )

Bill

BTW Good thread!

It almost sounds to me like if he put a simple plunger indicator in the quill, pointing straight down, and traverssed the table from side to side, he would see the 8 or whatever thou of error accumlate from one end to the other.

This seems to imply, that if he trammed the head to the way under the table, put the table back on and fly cut the top, he would be taking 8 thou at one end, and just a skim cut at the other.

This would seem to be a manufacturing error. As long as the gibs on the machine are not loose during the test, as you say. I don't think that table sag could account for what he sees, because he is never sweeping any portion of the table that is overhung.

Jim

================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================

Also note that the vise has been on the mill the whole time. When you tram the mill at one end of the table, where is the weight of the vise now acting? Off the other side. By not removing the vise, it looks to me like he is exagerating any play or flexibility in his mill...

My $0.02, --Glenn Lyford

My .02 here, If you don't lock the gibbs of the knee tight against the column all sorts of awful stuff happens. Before you tram the head to the table, raise the knee to the working height and lock it in place. I was told by someone I respect that the position of the knee is less reliable if it is lowered and locked. Works for me. Regards, Kent

While I lack specifications as I tap away at this, I'm confident the safe load capacity of that mill would be at least 500 pounds. Can't imagine that the vise would be enough of a load to effect the machine at all, let alone the amount of error being recorded. Think about it. Would a cast iron piece with the amount of mass in a mill table really flex several thou under a light load? Not likely. Otherwise, tables on mills would sag considerably by their own weight as they traversed from one extreme to the other. They don't.

Harold

The relative position of the knee as it relates to the head should play no role in the degree of precision displayed by a machine, assuming it is not a worn out piece of junk. I will concede that one would be wise to tram the table where it will be used, however, eliminating the smallest possibility of error. The entire concept of the knee is to get away from the very problem we're discussing, to have the ability to raise or lower the work surface without losing orientation. While none of us should expect the degree of precision found in jig borers and jig grinders, we should certainly expect, and under normal circumstance, receive, the degree of precision that permits movement of the knee without materially effecting perpendicularity and relative position. . It goes without saying that the gibs of the machine should be properly adjusted to prevent skewing of the moveable members, however.

Harold