Not milling flat

[38super]

3in1 machines are not known to be ridgid. My old shoptask mill head would lift away from the cut. Only solution was to use sharp tooling, thin cuts and no more than 30% overlap.

 

[eugene13]

Maybe annealing the material would help.

 

[rzw0wr]

38super, It is not a 2 in 1 type.
Mill/drill press.

I think I may have found some of my problem.

I put an indicator in the quill and trammed the table on the X axis at the middle of the table. Within .001" or so.
I trammed the Y axis, it is within .002" or less.

I trammed the table in 8" sections and the numbers were all over the place.

I had this table machined when I got this mill about 3-4 months ago.

Can anyone give me a good reason why I should not fly cut the table.

Thanks for the replies,
Dale

 

[38super]

Fly cut an aluminum scrap plate then measure it.

 

[Janderso]

I didn’t know they made feeler gauges Half the size of a human hair.
What’s aluminum foil?
Wait, I’ll check.
My heavy duty Reynolds is .001”
I don’t imagine you fold these up like a jack knife?

 

[RJSakowski]

I didn’t know they made feeler gauges Half the size of a human hair.
What’s aluminum foil?
Wait, I’ll check.
My heavy duty Reynolds is .001”
I don’t imagine you fold these up like a jack knife?

I do but the feeler right next to it is .025" I pair them up and fold them as a pair.

 

[RJSakowski]

38super, It is not a 2 in 1 type.
Mill/drill press.

I think I may have found some of my problem.

I put an indicator in the quill and trammed the table on the X axis at the middle of the table. Within .001" or so.
I trammed the Y axis, it is within .002" or less.

I trammed the table in 8" sections and the numbers were all over the place.

I had this table machined when I got this mill about 3-4 months ago.

Can anyone give me a good reason why I should not fly cut the table.

Thanks for the replies,
Dale

I wouldn't mill the table. At least not before nailing down the root cause. A loose gib can cause the table to tilt as you move the table to an extreme position making it look like it is out of tram. If you fly cut the table, you could be introducing a curvature.

Actually, tramming should test only one thing; how close to perpendicular is the spindle axis to the ways of the mill. To check if a table is true to the ways, I mount an indicator to the head or other external surface and sweep the table buy moving x and y. If the table surface is parallel to the ways, the indicator reading will be constant. If it is not, there can be two causes, the table surface is parallel to the ways or the table is lifting due to the clearance. Tightening the gib or locking the table should determine which ( it could be both).

When you say the table was machined a few months ago, how was that done?

 

[RobertB]

I had this table machined when I got this mill about 3-4 months ago.

Can anyone give me a good reason why I should not fly cut the table.

Internal stresses can cause it to warp and twist if you remove material from the surface.

 

[Bob Korves]

Actually, tramming should test only one thing; how close to perpendicular is the spindle axis to the ways of the mill. To check if a table is true to the ways, I mount an indicator to the head or other external surface and sweep the table buy moving x and y. If the table surface is parallel to the ways, the indicator reading will be constant. If it is not, there can be two causes, the table surface is parallel to the ways or the table is lifting due to the clearance. Tightening the gib or locking the table should determine which ( it could be both).

Milling machines (and surface grinders) can give inaccurate readings when a indicator mounted to the spindle area is testing the table (or mag chuck on the surface grinder.) Because the tables of these machines often overhang the supporting saddles when at the ends of their travel, the table and the saddle get worn (or loose) in such a way that the table droops while moving to the ends of the travel. Looked at from the operator's side while standing well back from the machines, a table with this common type of wear is traversing a large radius, close to circular curve as it moves from end to end. The indicator, reading directly below the spindle, will not properly see that the axis is not flat, because it is always reading on the high point of the circle. Another method must be used to test the table or mag chuck for flatness, and to create work that has a flat surface in the table travel direction. Or, the machine must be repaired so the ways are no longer curved or loose. This issue is quite common, but little known or compensated for.

 

[RJSakowski]

Milling machines (and surface grinders) can give inaccurate readings when a indicator mounted to the spindle area is testing the table (or mag chuck on the surface grinder.) Because the tables of these machines often overhang the supporting saddles when at the ends of their travel, the table and the saddle get worn (or loose) in such a way that the table droops while moving to the ends of the travel. Looked at from the operator's side while standing well back from the machines, a table with this common type of wear is traversing a large radius, close to circular curve as it moves from end to end. The indicator, reading directly below the spindle, will not properly see that the axis is not flat, because it is always reading on the high point of the circle. Another method must be used to test the table or mag chuck for flatness, and to create work that has a flat surface in the table travel direction. Or, the machine must be repaired so the ways are no longer curved or loose. This issue is quite common, but little known or compensated for.

That is why I suggested not milling the table.

This is exactly what was happening on my Tormach. The x axis ways were loose and when the table moved to the left, the weight distribution was such that center of gravity of the table os no longer supported and the right side of the table lifted by 4.5 thousandths over the 14" of travel. The lifting started at 10" from the homing position with the table to the far right. Left uncorrected, a facing operation would actually cut a curved face.

Adjusting the gib brought the lift down to a little over a thousandth. This lifting is also prevalent on other machines and warnings are given about loss of accuracy when using the table in extreme positions. Separating lift of table from surface irregularities is problematic. The best way to check the surface that I can think of would be with a straight edge and bluing. To check if the table surface is parallel to the ways, I would probably have to remove the table and set it up on a surface plate and check the height at various positions.

In the case of my Tormach, I consider it a design flaw. The saddle is too narrow which aggravates the problem and the stepper motor and housing hang off the left side of the table lending to the imbalance. The problem is further exacerbated when the 4th axis RT is added as it mounts on the left side of the table.