[rossrods]

Hello all, isnt there a video somewhere showing the correct way to true in the mill by way of setting tram. I thought I saw a video somewhere but cant find it anywhere. Thanks

[flyinchips]

NYCCNC has numerous videos on the Tormach. Here is one that starts the tramming process: <http://www.youtube.com/watch?v=lle-k_ccFAA>

As I recall, it doesn't complete the process, but I think all of the necessities are there.

[philbur]

Beware, as the video notes what is depicted is not the correct way to tram a mill. All it does is demonstrates that how you support the mill can influence the tram.

Sweeping the table with an indicator only has the capability to show how close the spindle is to being perpendicular to the table surface, it tells you nothing about the relationship of the column ways to the table, which is equally important. What procedure you use depends on where the error is.

Example: If you are indicating off a table surface that isn’t flat why would you believe that shimming the column is going to correct it.

Example: if your spindle is indicating as not being square to the table surface, is it the spindle to the column that is not correct, or is it the column to the base, or is it both.

Phil

[zephyr9900]

Phil, please do provide a link to the correct way to tram a mill. The subject has come up here several times previously, but I haven't been able to find a good, documented, reliable method that corresponds to, say, Rollie's Dad's Method of Lathe Alignment.

Thanks,

Randy

[bobeson]

I posted a while back about re-tramming my pcnc1100 column by shimming between the column and base. I'd experienced a crash that put me further out of tram than the base-shimming method could fix, so I pulled my column and inserted carefully-designed brass shims that covered the whole column-to-base interface area to compensate for the measured column lean. It worked really well, but was rather involved.

Here is my original write-up:
http://www.cnczone.com/forums/tormac...00_column.html

I have shim drawings and pictures I took of the process of lifting the column if anybody is interested. In fact, I may launch into another re-shim after yet another z crash added another bit of lean to my column last month. If so, I'll post more detailed pictures and diagrams of the process I use.

The only real trick is in constructing two precision shims (one for each side of the column) from brass sheet stock. I'd rather use a compound sine plate on a real surface grinder with proper flat stock to make the shims, but alas I have no such gear as yet. If anybody else does have such gear and wants to produce custom shims, I'll craft a parameterized solidworks part to generate shim designs given measured X and Y lean angles in exchange for a set of finished shims . Just about every pcnc1100 owner will likely need a column shim at least once, to correct for as-shipped deviance if not a crash or general wear & tear. I found that my surface finishes were noticably improved vs. the factory-shipped tram after I completed my shimming. Not that they were bad to start with, but they got really nice when it was dialed in precisely. Now my surface finishes are tolerable but nothing to be proud about, and this is after a round of base-to-stand shimming.

It'll be nice when I learn how to reliably not crash my machine. Just when I think such rookie moves are a thing of the past, I find myself humbled once again...

[philbur]

Randy, unfortunately I don't know of link to a good procedure. There are many different understandings of what tramming a mill actually means, so there are many different procedures out there.

I guess the basic definition is: to check and adjust the machine such that the spindle rotational axis is (and moves) perpendicular to the plane of the table surface (or is it more important to use the plane of the table movement), within the envelope of the x, y and z-axis travel. The precision of the adjustment will always be subject to the best compromise that is achievable within the precision errors that you cannot or do not intend to correct. A bowed column could be one such error and a table that deflects under its own weight when at the extreme of the x axis travel might be another. There are several more, all of which I am not sure I could identify.

Providing you accept all these subsystem errors then the two basic steps would be:

1) To ensure that the z axis ways are perpendicular to the table surface in both x and y planes.

2) To ensure the spindle rotational axis is parallel to the column ways in both x and y-planes.

If you use an indicator sweeping the table to check and adjust at step 2) you cannot do it before you have performed step 1).

Many people will perform step 2) without considering step one. This will potentially introduce a second error when trying to incorrectly adjust the original error.

How you perform the checks in 1) and 2) depends to some extent on what tools and measuring equipment you have available.

The proper method of correction would be to scrape the base of the column for step 1) and to scrape the head/saddle interface for step 2) The HSM’er fix would be to shim either as necessary. On the Tormach you may have to leave out the two head alignment taper pins.

PS: Not wishing to start another discussion but Rollie Dad’s method for lathes is only good for correcting an otherwise good lathe that has a twisted bed. For example it cannot differentiate between a misaligned head-stock and a twisted bed. So again you could introduce a second error when trying to incorrectly adjust the original error.

Phil

Phil, please do provide a link to the correct way to tram a mill. The subject has come up here several times previously, but I haven't been able to find a good, documented, reliable method that corresponds to, say, Rollie's Dad's Method of Lathe Alignment.

Thanks,

Randy

[philbur]

I think it is unlikely that a crash would result in a movement of the column relative to the bed, at least not one that would result in a change in vertical alignment of the spindle. The most likely source would be movement of the head on the saddle, against the taper pins, especially if the taper pins had worked lose, which they do.

Phil

I posted a while back about re-tramming my pcnc1100 column by shimming between the column and base. I'd experienced a crash that put me further out of tram than the base-shimming method could fix, so I pulled my column and inserted carefully-designed brass shims that covered the whole column-to-base interface area to compensate for the measured column lean. It worked really well, but was rather involved.

[zephyr9900]

Phil, thanks for your answer. Yes, I mentioned Rollie's Dad's method specifically because it is for setting up a well-performing lathe again after a move. And because it doens't use a lot of expensive test fixturing... I'd really like to find a corresponding technique starting from zero with the mill.

I was satisfied with my Tormach's performance in California but need to commission it properly after the move to Texas. When I first set up the mill, I had use of a long-base precision level, laboratory-grade Mitutoyo squares, etc. but probably just got lucky setting it up in the first place. It hasn't suffered any specific traumatic events during or after the move.

I have an early Series I machine, with the roughly-welded stand with the rubber pads between it and the machine base. It is sitting on a non-level or straight garage floor in an area where the ground is prone to movement if we don't soak the perimeter of the foundation every day. The machine is tight (I can interpolate short bearing bores to at least .001" size and roundness) but my flycutter shows obvious non-orthogonality of the spindle to table.

Bobeson, thank you also. I have read your thread, fairly recently in fact, but I'm wondering if I don't need to take a big step back from there. I don't even know if the stand pads are coplanar, since they weren't faced after welding as the later stands have been. So even if I would use the stand-leveling feet or shims under the machine base corners I don't have a stable connection between stand and mill. If I could borrow a precision level again and ensure the Y ways don't have any twist, how stable might that adjustment be, especially as a "foundation" for subsequent tramming steps?

Randy

[bobeson]

I think it is unlikely that a crash would result in a movement of the column relative to the bed, at least not one that would result in a change in vertical alignment of the spindle. The most likely source would be movement of the head on the saddle, against the taper pins, especially if the taper pins had worked lose, which they do.

Phil

Hi Phil,

It does seem unlikely, yet my measurements seemed to conclude column lean as opposed to head tilt. Specifically, I used a 4"x12" cylindrical square held vertically on the table, and measured deflection in both X (mostly) and Y (slightly) during movement of the Z axis. If the head was tilted, there would be no relative change during Z movement, only during a sweep of a spindle-mounted indicator across the table. Instead, I measured consistent angular deviation results using both the vertical-motion-with-square test and also during the spindle-mounted indicator sweep. After inserting my shims, the tram was nearly perfect, both during sweeping and Z traverse.

I'd be interested to know if you have a different interpretation, given this information. I was pretty shocked to think that a crash could deform the column, yet that's what the measurements seemed to show.

-Bob

[bobeson]

If I could borrow a precision level again and ensure the Y ways don't have any twist, how stable might that adjustment be, especially as a "foundation" for subsequent tramming steps?

I'm not entirely sure I follow your question, so pardon me if I'm answering the wrong thing I would do your best to make sure that the base is level/not twisted, as I did encounter measureable twist in my base when I used excessive base-to-stand shimming, and if your stand pads are not planar then you may indeed be experiencing twist. I can't recommend a procedure offhand for compensating for the twist precisely, but I imagine you'll get there with a little determination. Note that the table-to-column angular relationship is *not* the same as the y-axis-ways-to-column relationship! Only use the actual table for checking tram, not the ways.

As for stability, although they seem pretty cheesy, I don't see why a properly constructed and positioned shim wouldn't have perfectly acceptable long-term stability. It's not like this is part of a highly dynamic motion assembly - it's clamped solidly between two hunks of cast iron that don't move. Once you get the distances correct, I don't see what there is to change, unless you go and bend your column again.

-Bob

[philbur]

Bob, it sounds to me like you've got it pretty much covered.

Phil

I'd be interested to know if you have a different interpretation, given this information. I was pretty shocked to think that a crash could deform the column, yet that's what the measurements seemed to show.

-Bob

[philbur]

Randy, My method for setting the mill on a roughly welded stand was as follows:.

- Put adjusting feet on each leg of the stand.
- Sit the mill on the stand
- Adjust the feet so they are all touching the floor.
- Lightly bolt the mill to the table on three corners.
- Fully shim the fourth corner. Shims should have resistance in the gap but not forced.
- Bolt all four corners of the mill base down onto stand.
- Check for none twist of the base (and the stand) by undoing each hold-down bolt in turn and check for lift of the base from the stand with a thin shim. Re-shim as necessary)
- Adjust the four legs of the stand until the table surface is level.

You might get lucky and a table sweep with an indicator shows everything as OK, don't expect perfection (look at the test certificate for the acceptable tolerance). If the sweep is not good then you can check as Bob did, with a cylinder square or alternatively with an angle-plate to determine if you have column lean or head lean. The only problem with the angle plate method is you have to move it lots of times and the math can get a little confusing. (warning: cheap angle plates may not be square, however you can compensate for this by repeating each test with the angle plate rotated 180 degrees, lots more checks and the maths/geometry will make your brain ache).

I don't think that attempting to twist the base to correct a lean of the column as in the video is a good idea as it will not only impact the column but also potentially the Y-axis ways.

Which way does the fly-cutter indicate the lean, front to back or side to side, or both.

Phil

[rossrods]

Thanks guys for the input. Sorry for the 24 hour delay. Got busy. The only thing that seems to be out a little is the x travels. It might be out .005 over 16" in x travel from the negative side towards positive side. We are doing some critical venting and if the vent lines are over .003 the material will flash. I can touch off on the left most side of my part with the ETS tool setter, and when the program runs the vent lines there is no venting in the right side. We are working off the vice with 3" parallels underneath the ends. Its possible the vice could be out a little also. However the level shows level but thats only so accurate. I havnt tried to take the part down flat on the bed as of yet. Will give that a try next project. I will say this, the tormach is running very good in accuracy in the Y direction. The projects are turnning out nice other than the venting on extreme right. Also these programs are very long in length and are taking 24 hours run time. So the tormach is doing really well. Thanks for all the advice.

[philbur]

Not sure what venting is. However it sounds like the table surface is not in the same plane as the x-axis motion. Correcting it could be a major operation or it could be due to lean, if the table is offset to one side. Leveling the table surface will not correct it. Normal practice would be to shim the part until the surface of the part indicates "flat" corner to corner, or put the part in the middle of the table.

Phil

I can touch off on the left most side of my part with the ETS tool setter, and when the program runs the vent lines there is no venting in the right side. We are working off the vice with 3" parallels underneath the ends. Its possible the vice could be out a little also. However the level shows level but thats only so accurate.

[zephyr9900]

- Put adjusting feet on each leg of the stand.
- Sit the mill on the stand
- Adjust the feet so they are all touching the floor.
- Lightly bolt the mill to the table on three corners.
- Fully shim the fourth corner. Shims should have resistance in the gap but not forced.
- Bolt all four corners of the mill base down onto stand.
- Check for none twist of the base (and the stand) by undoing each hold-down bolt in turn and check for lift of the base from the stand with a thin shim. Re-shim as necessary)
- Adjust the four legs of the stand until the table surface is level.

Getting back to this, thank you for your input, Phil. I wish it were that simple. The Tormach stand has an appreciable stiffness, but not as stiff as the machine base itself, and will flex when the machine weight (1100+/- lb or 500kg) is placed on it. Furthermore, the pads on which the machine base are not machined to be coplanar, so Tormach supplied rubber pads to go under the machine base feet to absorb the misalignment.

My scenario is more like this, I'm afraid:

-Set stand on floor and adjust feet so the base doesn't rock. Mounting pads are in an unknown condition.
-Set machine down on stand. Stand flexes under machine's weight, and both stand and machine base take a new unknown alignment. All four corners of machine base are probably resting on the stand pads but what does that mean? Shimming and/or stand twisting are meaningless since there is no absolute reference.

I can't help but think that what I really need to do is get some heavy nuts to thread onto the studs of the stand, so I can place the mill's weight on the stand (sitting on the nuts but suspended above the pads themselves) and then measure the pad's alignment while under load. I could tweak the base leveling feet and/or design shims to make the mounting pads level with each other while ignoring the mill's alignment. Then turn the nuts down (or remove them) to place the machine base directly on the mounting pads and go from there knowing that the stand is not imposing any twist to the mill base.

Does that sound over-the-top to you? The only other thing I could think of is to make an intermediate structure so the mill itself is sitting on a plate that is 3-point mounted on the stand (at work I have designed kinematic mountings for optics and they are always 3-point) so there is no imposed twist to the mill base.

This is all prior to any actual tramming, obviously...

Randy

[flick]

Randy,

After adjusting the feet of your stand so it doesn't rock and then placing the machine on it (as you described above), adjust the feet while checking the alignment of your machine. You should only have to adjust one foot. Keep track of how far you are adjusting the foot (might be helpful to know the pitch of the screw...). Make a shim as large as the adjustment. Place it between the base and stand, and adjust the foot back down until your machines alignment is acceptable again. It should be back about where it started. Now your machine is aligned, and (the goal of this whole exercise) it's weight is distributed somewhat evenly at both the machine/stand interface and the stand/floor interface.

If you're using rubber feet this won't work, because the adjustment is taken up more by the flexible feet than the rigid base. Get rid of them. They're just there to hide uneven weight distribution, which this method addresses. Use steel feet instead.

[philbur]

Randy, in what way does this impact/negate the procedure I outlined.

The Tormach stand has an appreciable stiffness, but not as stiff as the machine base itself, and will flex when the machine weight (1100+/- lb or 500kg) is placed on it.

So what, you shim between them to compensate. If the stand flexes enough that all four corners of the mill are on the stand then great.

Furthermore, the pads on which the machine base are not machined to be coplanar, so Tormach supplied rubber pads to go under the machine base feet to absorb the misalignment.

So what, you shim between them to compensate. Absolutely nothing needs to be coplanar. The only condition that is required is that the four mounting points on the mill need to be without applied hold down (bolt) stresses. That is the all four contact points on the stand need to be in contact with the corresponding 4 points on the mill base (by shimming if necessary) before you bolt the mill to the table.

Phil

[zephyr9900]

If the stand flexes enough that all four corners of the mill are on the stand then great.

My point is that no, it is not great, Phil. The machine base does not have infinite stiffness, and if it is placed on an uneven surface it itself will warp under its own weight. Both machine base and stand will warp when the mill is placed on the stand, and the proportioning of the warp between the two will be dependent on their relative stiffness. But both will warp. That is why I made the point that the stand has an appreciable stiffness itself but is not made to the precision of the mill base, so it is a problem.

Randy

[zephyr9900]

If you're using rubber feet this won't work, because the adjustment is taken up more by the flexible feet than the rigid base. Get rid of them. They're just there to hide uneven weight distribution, which this method addresses. Use steel feet instead.

Thank you, flick. Yes, there are rubber pads in the stand feet, and also rubber pads between the stand pads and machine base. The latter are also there because in the early Tormach bases, the mounting pads weren't skimmed flat after the stand was welded up. That bothers me somewhat because the contact surface between the mill and stand is a fairly large area.

Randy

[philbur]

I think the basic assumption for a mill is that it retains its correct geometry when subjected to its own weight, even when one of the four contact points is not supported. Otherwise the only stand that would work is one that is infinitely rigid and has four mounting points whose relative locations are a perfect match for the mills four mounting points when the mill is not subject to any stresses that would put the mill out of "tram".

Good hunting.

Phil

PS: Quote: "....but is not made to the precision of the mill base, so it is a problem."

I've been running my mill for 5 years on a fabicated stand and it has neve been a problem.

Unlike a lathe, to align a mill you change the head alignment to the column or the column alignment to the table, you don't attempt to twist the base to align anything.

My point is that no, it is not great, Phil. The machine base does not have infinite stiffness, and if it is placed on an uneven surface it itself will warp under its own weight. Both machine base and stand will warp when the mill is placed on the stand, and the proportioning of the warp between the two will be dependent on their relative stiffness. But both will warp. That is why I made the point that the stand has an appreciable stiffness itself but is not made to the precision of the mill base, so it is a problem.

Randy