“Tooling up” - Tramming the Mill

I explained in an earlier post how I planned on fixing one of the shortcoming of my lathe, the tooth skipping issue, which has been causing all sorts of troubles when making airplane parts.

To recap, I’ll have to mill the top of the automatic longitudinal feed casing, in order to raise the pinion assembly into a closer contact with the rack.

Precision in this endeavor is important, and since I am new, as well as self taught, at machining, I will need to give myself every advantage I can.

One of the interesting features of this mill is the ability to lean the mill tower left and right in order to make cuts at any angle between -50˚ to +50˚. This is accomplished by loosening the one big bolt in the back of the base of the tower, and leaning away. 

The base needs to be anchored to the table for this operation, or risk toppling the unit. 

This solution is pretty handy as long as you are not looking for much precision in your angles. Another drawback is that returning the tower back to vertical using the graduated indicator is unreliable at best, if any accuracy is expected, and it's only really handy for initial positioning.

Tramming is the process of squaring the spindle with the table, this is necessary because the angle of the head is adjustable from side to side. Tramming the mill requires the use of a dial indicator mounted so that it rotates with the spindle and reads against the table at the farthest distance possible from the spindle.

Bolt holding the column up (on the right).

So it is, that I finally used the lathe to make something useful: a holder for my dial indicator, to be used in setting the mill tower to precise vertical (as it relates to its base).

Digging into my pile of scrap metal I found two aluminum rods. I turned the round one with the lathe at both ends, and threaded one end. 

Thread on aluminum

I then milled a small flat on the hexagonal rod with the drill press (very slowly), drilled it, tapped it, and joined it to the former. 

Yes, I am aware that hexagons already have 6 flats on them. But had I used any of them, I could not have later held it in my vise for drilling and tapping, since it would have presented its edges to the vise jaws. 

Threaded joint

With my dial indicator mounted at the end, and the rig set in the chuck, I "swung" into action measuring both ends of the table, and moving the column back and forth endlessly, for the next 30 minutes.

Setting the zero vertical reference point on the right side of the X-Y table.

Measuring the left side of the table (1/1000" low)

The best I was able to attain is a 1/1000" difference over a 10" swing, so 1/10'000" vertical per 1 inch of horizontal. In other words the amount the tower is off from true perpendicularity with its table is 0.0057˚, or 89.9943˚ measured in the other direction.

This result was very hard to achieve, so much so in fact that just touching the table with any pressure would skew the measurements by 3 to 5/1000". Tightening the big bolt was very difficult to do as well without affecting accuracy, and was done in repeated incremental steps of tilting, measuring, and torquing.

All of this back and forth motion was very easily seen on the dial indicator, but was virtually undetectable by looking at the mill scale which never even moved from its zero position.

Nice but useless angle indicator at the base of the tower.