Tramming range

Discussion on all milling machines vertical & horizontal, including but not limited to Bridgeports, Hardinge, South Bend, Clausing, Van Norman, including imports.

Moderators: Harold_V, GlennW

Mr Ron
Posts: 1914
Joined: Tue Dec 22, 2009 12:36 pm
Location: Vancleave, Mississippi

Tramming range

Post by Mr Ron » Mon Mar 30, 2020 2:02 pm

I can't seem to tram my knee mill any more closer than .001". Is this an acceptable amount, or should I keep at it? I have been tramming from the top of the vise, not the table as I hardly ever set up on the table.
Mr.Ron from South Mississippi

User avatar
Richard_W
Posts: 1933
Joined: Fri Jan 10, 2003 1:00 am
Location: Molalla, Oregon

Re: Tramming range

Post by Richard_W » Mon Mar 30, 2020 4:49 pm

I usually tram into the width of the table. If the table is 9" wide then I ream to just under a 9" circle. With in .001 inch is most likely way better then you need to be.

LIALLEGHENY
Posts: 234
Joined: Sat Jan 11, 2014 12:36 am
Location: Bohemia, NY

Re: Tramming range

Post by LIALLEGHENY » Mon Mar 30, 2020 5:03 pm

Tram it to the table, not the vise. If you get another vise, or use 123 blocks, parallel bars or some other fixture, at least you know the machine is trammed correctly. Depending on the manufacturer, it is also quite possible the top of the vise isn,t parallel with the table.

Nyle

Harold_V
Posts: 18137
Joined: Fri Dec 20, 2002 11:02 pm
Location: Onalaska, WA USA

Re: Tramming range

Post by Harold_V » Tue Mar 31, 2020 3:39 am

The overall accuracy of a mill begins with the column, knee, and way surfaces. Here's what I recommend.
Begin by checking the gib on the knee. It should be snug, but not overly tight. The knee should be able to just settle by its weight, not driven downward by the screw.
When the gib is adjusted (a snug gib ensures the knee is not leaning downward at the front), you should then dress the table of your mill. Do that by using a single cut or smooth file, one that has seen some use, so the original sharpness is gone. Note that a new file offers the opportunity for metal to be removed where it shouldn't be.

The file is placed at an angle on the table surface, no handle, and drawn by your palms parallel to the T slots, covering the entire table surface, side to side. By doing this, any slight projections (usually caused by clamping on a chip, or from dropping small objects on the mill table) will be removed, leaving the mill table surface untouched otherwise. I do not recommend using a stone for this purpose, as it isn't selective and will remove, albeit very little, metal where it shouldn't. Once you have deburred the table, it should feel silky smooth when stroked by hand. If it isn't, repeat the file operation.

With the table centered on the saddle, side to side, use an indicator mounted in the spindle to sweep a circle. I make sure my indicator crosses the front and rear T slots, so I make the circle as large as I can. I don't like using anything between the table and indicator, as that offers a slight chance of introducing error, especially if one skips the deburring of the table. When adjusting the head, the restraining bolts should be snug, but still allow the head to rotate. If they are not, when you tighten the bolts you most likely will see a change in your setting.

When you place the indicator, it is best applied such that the contact addresses the table such that you can rotate the indicator with the tip trailing. I often don't do that, but to ensure that the crossing of the T slots doesn't alter the indicator setting, I make contact by only a thou or two, so the contact moves very little. I also use the largest diameter tip for the indicator. This allows the indicator tip to address the T slots from the side without losing registration.

The larger the circle you sweep, the more precise will be your setting. New mills are made with the front edge of the table ever so slightly higher than the back, to address settling as the machine wears. If your mill is not new, ensure that you don't leave the table high at the front. I prefer to make it even, front to back, just as I do side to side.

There's no reason why you can't get a head within a half though sweeping a large circle (about 8"). That degree of precision will yield no issues when moving the knee, keeping in mind we're talking about a milling machine, not a jig borer.

Dialing the top of a vise isn't recommended. If, by chance, the vise top is not dead parallel with the table surface, you will defeat the purpose of dialing in the head. Considering the fact that the large surface is also the moveable surface on a mill vise, it most likely won't be dead parallel.

H
Wise people talk because they have something to say. Fools talk because they have to say something.

Low Flyer
Posts: 13
Joined: Sat Aug 10, 2019 3:07 pm
Location: Richmond, KY
Contact:

Re: Tramming range

Post by Low Flyer » Thu Apr 09, 2020 7:50 pm

Hi Harold,

During my 30 years in this trade, I have never heard of the front of a table being higher than the back, ever. You are the only person that I have known to state this, and I think that I seen where you've stated it a couple of times. I'm not trying to sharp shoot you. It's just the skeptic in me that doesn't believe everything that I read on the internet. It seems to me that proper tramming would effectively negate any difference between the height of the front and the rear. Do you have some documentation that proves it? I'm always willing to learn something new to me, and some weeks I do.

Gil

Harold_V
Posts: 18137
Joined: Fri Dec 20, 2002 11:02 pm
Location: Onalaska, WA USA

Re: Tramming range

Post by Harold_V » Fri Apr 10, 2020 1:42 am

My comment is not based on the internet. It's based on the owner's manual for a Bridgeport mill, and makes perfect sense. By starting life a touch high, when the mill wears, it sags in instead of out. Needless to say, continued wear results in the knee sagging. It does not get higher in the front. It has no reason to do so.

The idea of the front of a mill table being slightly higher when new shouldn't surprise you. You should already understand that the cross slide of the typical engine lathe is intentionally not dead square, so that faced objects sit flat. The amount of taper is extremely small, but it does exist. Note, also, that the tailstock of the typical lathe is high as well. It, too, wears in instead of out. In all cases, the amount is very small. Bridgeport reports the amount as being a half thou higher at the front of the table than at the rear of the table.
It seems to me that proper tramming would effectively negate any difference between the height of the front and the rear.
No amount of tramming (a term I do not like) will correct for an error in a table. If the table isn't square with the column, any correction will manifest itself in an error in the part (using a bored hole as an example, it will either not be square with the part, or orientation is lost, albeit by very little, if the knee is moved). The question being is it important? We're talking tenths over several inches. You have to keep in mind that we are discussing milling machines, which are far from high precision machinery. We are not discussing jig borers or grinders.

H
Wise people talk because they have something to say. Fools talk because they have to say something.

Low Flyer
Posts: 13
Joined: Sat Aug 10, 2019 3:07 pm
Location: Richmond, KY
Contact:

Re: Tramming range

Post by Low Flyer » Fri Apr 10, 2020 1:02 pm

Of course I know we are discussing milling machines.

I'm trying to understand what you are saying. You say the front to the table is supposed to be higher than the rear.
Harold_V wrote:
Tue Mar 31, 2020 3:39 am
New mills are made with the front edge of the table ever so slightly higher than the back, to address settling as the machine wears.
I didn't say that your comment was based on the internet. I said I don't believe everything I read on the internet.
Low Flyer wrote:
Thu Apr 09, 2020 7:50 pm
It's just the skeptic in me that doesn't believe everything that I read on the internet.

Harold_V wrote:
Fri Apr 10, 2020 1:42 am
No amount of tramming (a term I do not like) will correct for an error in a table. If the table isn't square with the column, any correction will manifest itself in an error in the part (using a bored hole as an example, it will either not be square with the part, or orientation is lost, albeit by very little, if the knee is moved).
I agree with the above statement in red. That statement seems to be in direct conflict of the notion that the front of the table is higher by design than the rear.

Well my friend, I'm not going to argue about it, and an argument certainly was not my intention. We'll just have to disagree with each other.

Right now I am looking at the inspection certificate that came with my milling machine. It does not indicate that the front is higher than the rear. I pulled the same documents for the machines at work, to determine if my machine was "defective". Those machines do not show that condition either. These machines are Bridgeports, Excellos, Chaviers, Acer, Mori-Sieki, and of course my MSC machine.

I know that some may look down on my MSC, but it is my experience that it is the capability of the man that is more important than the pedigree of the machine.

Gil

User avatar
SteveM
Posts: 7159
Joined: Mon Jun 27, 2005 6:18 pm
Location: Connecticut

Re: Tramming range

Post by SteveM » Fri Apr 10, 2020 1:55 pm

Low Flyer wrote:
Thu Apr 09, 2020 7:50 pm
During my 30 years in this trade, I have never heard of the front of a table being higher than the back, ever. You are the only person that I have known to state this, and I think that I seen where you've stated it a couple of times.
This is similar to the tailstock on a lathe. When new, they point slightly high so that as it wears, it goes thru center and then low. It's average error will be less than if it starts in the center and all the wear makes it worse.

Steve

Harold_V
Posts: 18137
Joined: Fri Dec 20, 2002 11:02 pm
Location: Onalaska, WA USA

Re: Tramming range

Post by Harold_V » Fri Apr 10, 2020 3:22 pm

Low Flyer wrote:
Fri Apr 10, 2020 1:02 pm
Of course I know we are discussing milling machines.

I'm trying to understand what you are saying. You say the front to the table is supposed to be higher than the rear.
Yes, that's what I'm saying. However, how much is what matters. The amount of error designed in to the machine is intended to prolong a machine's useful life. Think of it this way. I'm going to assume that a machine tool will perform daily for two years without significant change. If the machine starts out in perfect (there is no such thing) alignment, in two years it can be assumed that it will have worn enough to be concerned. By sharp contrast, assuming the machine is made slightly high (.0005"), in two years the machine will have passed through imperfection to perfection (there is no such thing), so the machine is actually in better condition than it was when built. That simple act adds years of productive life to the machine, with the degree of introduced error bordering on being insignificant. Just like the height of a tailstock.
Harold_V wrote:
Fri Apr 10, 2020 1:42 am
No amount of tramming (a term I do not like) will correct for an error in a table. If the table isn't square with the column, any correction will manifest itself in an error in the part (using a bored hole as an example, it will either not be square with the part, or orientation is lost, albeit by very little, if the knee is moved).
low flyer wrote:I agree with the above statement in red. That statement seems to be in direct conflict of the notion that the front of the table is higher by design than the rear.
Yes, it is in direct conflict, but in reality it is a good thing, which I explained above. Keep in mind, the amount of introduced error is likely smaller than the degree of precision that would be applied by many in squaring the head of a machine. Many of the people who machine appear to be intimated by anything closer than a thou.
Well my friend, I'm not going to argue about it, and an argument certainly was not my intention. We'll just have to disagree with each other.
No problem. I have no need for anyone to agree with me. If I offer a bit of advice and someone can profit, so be it. If anyone feels what I have to say is a waste of time, no problem. I'll continue to provide what little I can for those who may benefit.

One thing I have learned in my many years of moderating fora is that it makes no difference what one says, or how it is said. Someone, somewhere, is not going to agree, and often may even take umbrage to what has been said. Each of us has our own belief system, and we tend to live by what we believe, right or wrong.

H
Wise people talk because they have something to say. Fools talk because they have to say something.

Mr Ron
Posts: 1914
Joined: Tue Dec 22, 2009 12:36 pm
Location: Vancleave, Mississippi

Re: Tramming range

Post by Mr Ron » Wed Apr 22, 2020 4:23 pm

"One thing I have learned in my many years of moderating fora is that it makes no difference what one says, or how it is said. Someone, somewhere, is not going to agree, and often may even take umbrage to what has been said. Each of us has our own belief system, and we tend to live by what we believe, right or wrong."

Well said. When someones mind is made up, Arguing becomes a waste of time. I accept the words coming from someone who has proven himself from years of experience.
Mr.Ron from South Mississippi

User avatar
BadDog
Posts: 4916
Joined: Wed May 17, 2006 8:21 pm
Location: Phoenix, AZ

Re: Tramming range

Post by BadDog » Wed Apr 22, 2020 6:32 pm

Low Flyer wrote:
Thu Apr 09, 2020 7:50 pm
During my 30 years in this trade
<snip>
It seems to me that proper tramming would effectively negate any difference between the height of the front and the rear. Do you have some documentation that proves it? I'm always willing to learn something new to me, and some weeks I do.
I readily acknowledge that when it comes to comparing my ability or knowledge to an actual machinist, I'm a hack at best. I picked this up only to support my automotive custom work hobby some 18 years ago. That's starting from zero, no training, and very slowly learning only what I need to succeed at my goals, with a solid nod to developing skills likely to be needed for my interests in the future. So I have no standing to disagree relative to your experience, or Harold's (which I believe is at 40 + years in very high precision work?), or for that matter something close to 99% of the folks here...

But I understood quite some time ago that truing the head with respect to the table axis (much less referencing the surface) will do nothing to correct for errors in the slide orientation. And that's true of lathes or mills. It may be able to correct for a fixed position workpiece, but the minute you traverse the out of true slide, the game is over and you have the associated error applied to the work you are doing. Connolly, Moore, and others cover quantifying and correcting that in painful detail.

For example, consider a mill where the table table y axis ways are low near the operator (perhaps from wear) and gets higher as it goes to the far (typically column) side, and assuming all else is "perfect", then you can certainly align the spindle axis normal to that traversal. You can then bore a hole true to that travel. But the moment you traverse in Y from near to far, the work piece is going to be rising relative to that spindle, which presents a problem if you are using a face mill and have a very tight tolerance for maintaining parallel (or flat, surface will be concave) in that direction. For relatively small way alignment errors, the deviation is probably acceptable for mill work, and higher tolerance requirements would likely be delegated to some grinding or more appropriate finish operation, but the fact remains that the error in way alignment is in no way corrected by alignment of the spindle axis.

I also understand and have come across the assertion regarding "wearing into alignment" relative to "wearing out" from some degree of perfection originally achieved by manufacture. It's typically a tiny amount out "high" that is judged to minimally impacts accuracy (staying within acceptable bounds). If that initial deviation "high" is judged at/near the limit of acceptable error, then that effectively doubles the life of the tool since it wears closer to perfect, and ultimately then out on the "low" side to the same amount (limit) before it breaks out of acceptable tolerance. That allowable tolerance is very different between a low priced Grizzly and a pristine new 10ee, but any manufacturer who is concerned with providing a long life (not a guarantee with modern MBA profit maximization) is going to be shooting for "high but in tolerance" over "dead perfect as shipped".
Russ
Master Floor Sweeper

User avatar
BadDog
Posts: 4916
Joined: Wed May 17, 2006 8:21 pm
Location: Phoenix, AZ

Re: Tramming range

Post by BadDog » Thu Apr 23, 2020 2:55 am

As I was winding down for the night, I realized too long after posting to make the edit, but my attempt at a too simple example is problematic.

If the reference surface (table top, etc) is closely parallel to the Y axis travel, and the travel remains "straight", and the spindle true (normal) to that reference surface/travel, then the error I described would not occur. In that case the resulting surface would be flat within the limits of the various interacting parts. To behave as described, it would imply that the reference surface wasn't parallel to the indicated axis travel, which wasn't part of my base description. Of course with wear (or manufacturing error) to that level, the assumed parallel relationship probably is not valid either, but that's beside the point. And if accounted for, squaring the spindle to travel could still correct.

It would have been easier (more accurate) to have described that when things like worn knee travel (which tends to tip table outward) or otherwise not normal to the other axis used to adjust the spindle (tip/nod) start interacting that things get wonky. Much easier to describe without running aground, but requires more setup and description.

The various accuracy experts go to painful extreme lengths trying explain the impact of various relations and interactions, so I should know better than to post on complex hard to describe topics when not completely focused on the topic, particularly when I think I have a simple example to express without a great deal of careful thought. Perhaps I could have done a better job back when I was reading Connelly and others actively trying to understand the condition of my machines and fresh on my mind rather than trying to construct a description from memory.

Hopefully I have helped my earlier error and haven't fumbled it further. I regret posting at all and would delete my post if I could, my apologies for the noise. Now, hopefully to sleep...
Russ
Master Floor Sweeper

Post Reply