Tramming a Mill - Am I Doing it Right?

Not wishing to hijack Edward Crouch's tramming thread so I've I've opened a related question. I've just re-trammed my WM18 after a dig in turned the head slightly. The mill is fitted with a stop screw to facilitate re-alignment but I don't entirely trust it.

As usual I've done the job by guessing the procedure. This isn't clever given that I could easily have asked on here first!

My first attempt to measure unwanted tilt involved using a DTI in a collet and a set-square bolted to the table. It was immediately obvious that this method is badly flawed.

My second attempt used the set-up shown in the photo. A DTI is mounted at right angles to a collet mounted rod in the spindle such that the DTI point touches the table about 160mm away from the spindle axis.

To check right/left tilt I turned the spindle until the DTI is due east (assuming the pillar to be north). In that position the DTI was lowered to touch the table plus one dial revolution to zero the meter and show any ± movement.

Then I rotated the spindle through 180° so that the DTI was in the due west position and noted the measurement. (The point of the DTI was lifted carefully over the slots.)

On my mill due west measured 0.22mm lower than due east, the two points being 320mm apart. After about 40 minutes spanner work, tapping with a rubber mallet, and many, many expletives, I managed to reduce this to a DTI needle thickness zero.

There are a number of ways I feel this method could mislead, for example the joints of DTI assembly shifting as the spindle is turned, spindle run-out, a bad collet, loose gibs, or the DTI sticking. Am I measuring tilt the right way, and what would the team advise?

Thanks,

Dave

Edited By SillyOldDuffer on 29/01/2017 14:20:24

Edited By SillyOldDuffer on 29/01/2017 14:20:47

Dave,

Your general principle seems sound enough, but the support for the DTI does look rather delicate for the job ... something more like a 'girder' might be more appropriate.

That said; if you returned to 'East' and still got a zero reading, then presumably nothing shifted during the process.

MichaelG.

That is pretty much as I do it, but rather than running direct on the table top I set up on a slip block or parallel. When the DTI is zeroed one side, slide out the slip block, rotate 180 degrees & gently slide the block under the clock to take the reading. I generally don't have the DTI bearing on the slip by more than 1/4 of a rotation of the needle, so that the likelyhood of disturbing the setup when the slip block is slid in & out is minimised.

I used to use a "bobbin" at work for this - basically a roller out of a large roller bearing sat "end-on". This had the advantage of having a nice radius on the end, so the DTI plunger didn't have to ride up a sharp step. Must have put is somewhere "safe" as I can't recall seeing it for a while !

What was the reading like front to back ?

Nigel B

Dave

I have no idea whether the illustration below would be of any use to you.

It is the test report for my milling machine. The order shown is probably the sequence used in setting up the machine.

I wish you every success with your machine.

JA

This is the same method I was shown when I did my apprenticeship many years ago. Did you check the north to south alignment? to increase the range, move the cross traverse from one end stop to the other.

Angus

Hi Dave,
​Not a lot wrong with your method of tramming as you've already considered extraneous influences that may alter your reading, & achieved 'zero' albeit the dti 'needle thickness' I trammed mine up using a new piece of plate glass, some have used a new cast iron brake disc for a Corsa iirc purchased very reasonably from auction site/ car parts suppliers. I fitted a couple of locking head blocks & so far never had the need to fit 'extra' bolts to the head tilting axis that some have. Have a look at my post...

http://www.model-engineer.co.uk/forums/postings.asp?th=82914... it's an old post so the pics are not viewable so here are the original pics ...





If you can get hold of a digital DTI that would give you a better reading to see how far out you really are. I have one so maybe now is the time I should do a recheck on mine.

George.

Edited By mechman48 on 29/01/2017 15:12:40

The 'modern' approach is to place a Nova brake disk on the table for the DTI to run on, to save having to lift over the slots

Wow, thanks all. Nice to know the basic method is sound and that there are some easy improvements. I particularly like the brake disk / slip / bobbin suggestions as they minimise risk of disturbing the DTI.

Until now I've been happy with my inexpensive DTI - it did all that was asked of it. A slight tendency to stick hasn't been a problem before but it was an unwelcome time-waster doing this job.

Other lessons learned: tighten the gibs first. The most off-putting aspect of trammming my mill is the way doing up the axle and locking nuts disturbs the setting. I dunno if it's 'best practice', but I succeeded is keeping the head vertical whilst tightening up by tapping it with a rubber mallet to correct any unwanted movement reported by the DTI.

I've measured North/South using the same technique with disappointing results. The head is leaning forward by 0.16 mm over 340mm. I've not noticed this to be a problem in practice but still...

Cheers,

Dave

I have a small granite surface plate (9 x 12 x 2 inches) that I lug over to the mill when I do this. Gives a nice smooth surface for the indicator and doesn't move around.

Edited By Bandersnatch on 29/01/2017 17:04:20

Dave, your method is correct. I would offer a few suggestions:

Instead a jointed arm, make a right angle bend in a piece of steel rod. Use that as the indicator support. You can even thread the end to fit the lug back of the indicator and clamp the lug between two nuts.

Follow Nigel's advise and use a parallel instead of indicating directly on the table.

If you have a back plunger indicator or dial test indicator use it. You can have the dial face always facing up.

In my opinion using a bearing race, brake rotor, or plate on top of the table only complicates the job.

The adjustable stops that Mechman48 showed will help a lot getting the last thou or two.

This is what I use. En 16 bent and tapped at one end for the m6 screw to hold the Indicator. I hold the whole thing in a collet chuck. and as John Reese mentioned dont use anything on the table, just make sure the table is clean and read directly off it. This one has a leg of 120 long so gives a sweep of 240mm.

I built a spindle square a while ago having read this thread on Model Engine Maker.

Seems to work well enough and is a lot less faffing about then using a single clock gauge.

I bought two cheapo metric DTIs off ebay, nothing special and something like £4.50 each.

Here's a couple of photos to give you an idea, but I can't show it on the mill, as the Centec's in Sheffield, and there's something on the Westbury at the moment over here in Buxton.

I made it pretty much according to the aforementioned thread, but on a 2MT arbour as that's what's in the spindle both my machines. I was quite pleased ,when having fitted the arbour to the crossbar, I popped it in the Myford taper and clocked both the bottom two faces. No discernible difference with a 1/10s DTI.

And one to show it's set up correctly before adding to a milling machine spindle

All the best

Bill

Edited By peak4 on 29/01/2017 19:57:22

Inspired by John, Raymond and Bill I knocked this up today. Quick straightforward projects producing a useful tool are very satisfying, even if it's not as posh as Bill's double DTI!

With hindsight there's a design fault. It would be much easier to grip the DTI if I'd made the knurled brass knob 30mm longer. As it is the back of the DTI gets in the way of my fingers. Never mind, I've learned another lesson.

Thanks, Dave