Tramming The WM18 mill head

[ChrisBParticipant @chrisb35596]

Some time ago I had tilted the head of my WM18 mill to cut some angles but when I was done and tried to tram the head back to normal I ran into problems.

I could get the head trammed in the left to right direction ( that would be the X axis?) but the forward to aft direction ( Y axis?) was off and I could not find a way to adjust as the column is fixed.

I remember asking of advice on how to adjust this and the general consensus was to leave as is if the out of tram was not excessive. The only thing that could be done was to shim the column.

I left it as it was as I was, but stayed at the back of my mind. This afternoon while in the workshop my thoughts revisited the issue, so I checked the tram again and wondered if I could somehow fix the head nod issue. Measuring with a test indicator over a diameter of 70mm I got a difference of 0.05mm FWD to AFT and 0mm left to right. The head nod is towards me, so if I could find a way to push the column back I should reduce the tilt.

The column is bolted to the base by 4 allen bolts and there's filler all round the base and column – shimming the column is not something I would like to do as it's not straight forward, and it will take a lot of effort and time. So I thought what if I could jack the column?

As a make shift jack I used two long m10 bolt and a coupling nuts, and placed them under the dove tail between the column and base. A couple of turns on the coupling nuts and the test indicator began to move, in a couple of minutes the head not problem was fixed. I read this is a common issue with these type of vertical milling machines and this could be a quick and easy fix to the problem.

[ChrisBParticipant @chrisb35596]

[Thor 🇳🇴Participant @thor]

Thanks for the tip Chris, I must check if it is possible to do the same on my milling machine.

Thor

[mechman48Participant @mechman48]

Did you back off the two front SHCS before you jacked up on the column ? as you say it has cured your 'nod' but are you not putting stress on to the column, not that it would make a lot of difference for the small amount you needed 0.002". As you have managed to move it 0.002" with the location pin in place & SHCS fastened down ( I presume ) then can you not slot a 0.002" shim under the column front & do away with the jack screws? or as intimated you have a 0.002" stress on the casting itself?.

George

[ChrisBParticipant @chrisb35596]

You'll have to take the way covers off. On mine I had previously taken them off when I installed the DROs, and then installed a different cover.

[JasonBModerator @jasonb]

Posted by ChrisB on 09/11/2019 16:52:58:

Some time ago I had tilted the head of my WM18 mill to cut some angles but when I was done and tried to tram the head back to normal I ran into problems.

If you got the problem from loosening the swivel then surely that is where the error is not in the column as you did not alter anything there.

As well as checking the tram you should check the head moves true and that the quill movement is true to the table

[mechman48Participant @mechman48]

JasonB has hit on the more likely cause; I thought about that later on but not having ever loosened mine off ( WM16 ) it didn't cross my mind. I did read some time ago that other owners have converted their machines into a 3 bolt head securing method but as mentioned, once my mill was checked out & trammed I never had the need to move mine, any angle set up needed I have used home made angle plates or digi angle gauge on the job set in the vice / angle plate.

George.

[ChrisBParticipant @chrisb35596]

The head only swivels in one direction – left to right, and that's adjusted and fine, using a flycutter gives a nice crosshatch pattern so I think the head is trammed ok left to right. The problem with head nod is there's no adjustment for it. Don't know if the head nod was ok before I moved the head as I never checked.

As for stressing the casting, I don't know. When I push on the head forwards or backwards with my hands I can see a deflection on the test indicator of about 0.05mm (at a 70mm diameter) so I'd say the mill is quite flexy, the allen bolts holding the column are tight.

Well I'm just saying what cured my problem, it was a simple easy fix for me.

[ChrisBParticipant @chrisb35596]

Posted by JasonB on 09/11/2019 17:17:02:

As well as checking the tram you should check the head moves true and that the quill movement is true to the table

Ah, didn't think of that, will have a look at that tomorrow and take a couple of passes and measurements.

Agreed George, I'm not moving the head again unless really necessary. I got a digital angle gauge which should be enough for my needs.

[I.M. OUTAHEREParticipant @i-m-outahere]

I had a problem like this and found that the mounting surface for the column to base had some burrs and some gunk in there , after a clean my readings improved but were still not perfect .

I layed a thick parallel on its side across the table and re checked the front to rear tram and it was spot on so i think my table has a dip in the middle or the top of one t slot is a little low .

If i set up a test bar in the spindle (after i tram the head) i can move the head up and down on the column and get a zero reading on both the x and y axis along the z axis column travel so i know the column is perpendicular to the table in the x and y axis – you should also then get a zero on these axis using the quill feed .

This becomes even more tricky with the old tilting column X2 Sieg mills because you have a tilting column and a headstock made from two separate pieces which can also move .

[ChrisBParticipant @chrisb35596]

One thing I noticed is the ammount of flex the column has. Pushing the column by hand forwards or backwards I could see the test indicator give a reading of about +0.05mm and -0.05mm , that is by applying a moderate force, it would deflect more if I pulled harder. That got me thinking….could it be that the column is in fact machined true to the base and table, but the weight of the head makes it nod that tiny ammount?

[SillyOldDufferModerator @sillyoldduffer]

Posted by ChrisB on 10/11/2019 06:11:22:

One thing I noticed is the ammount of flex the column has. Pushing the column by hand forwards or backwards I could see the test indicator give a reading of about +0.05mm and -0.05mm , that is by applying a moderate force, it would deflect more if I pulled harder. That got me thinking….could it be that the column is in fact machined true to the base and table, but the weight of the head makes it nod that tiny ammount?

I've wondered the same myself. If correct, it means the error will tend to reduce whilst cutting, and could well be a deliberate feature. It would be ironic to go to a lot of trouble correcting resting nod only to find actual cutting error is made worse!

All machines deflect under load and gravity. One way of reducing flex is simply to make the machine massive – cast-iron is effective. The other way is scientific, analyse the forces involved and compensate for them in the structure. As the WM18 is a later design, it's not unlikely it's been compensated. However, although my WM18 does all I want, it's a relatively light machine with a simplistic approach to adjustments. In particular, I find nipping up the final tram adjustment is tricky to do without introducing a slight error. It's a basic hobby mill, not a Jig Borer!

Dave

[ChrisBParticipant @chrisb35596]

Well just back from the workshop….as usual, Jason is right

I removed the screw jacks from the column and checked the column and quill. Found that the column was almost spot on straight, while the quill is tilted backwards (that's why when I jacked the column I got the tram right in the fwd to aft direction)

The tilt in the quill is significant, I mounted the test indicator on the quill at a diameter of 8 in, and on the full travel of the quill the indicator would read +0.3mm a the front and -0.3mm at the rear. The column max reading was 0.02mm.

Now how to adjust the quill if it can be done at all is beyond me!

Chris

[Anonymous]

Posted by SillyOldDuffer on 10/11/2019 10:23:20:

All machines deflect under load and gravity. One way of reducing flex is simply to make the machine massive – cast-iron is effective. The other way is scientific, analyse the forces involved and compensate for them in the structure. As the WM18 is a later design, it's not unlikely it's been compensated.

Old school manual mills are massive, but the material is essentially in the right place. Look at an industrial vertical mill and the column is much deeper than wide. And the column depth narrows towards the top. In simple terms the column is a beam in bending and unsupported at one end. The equations for bending and stress have long been available so basic deflections and stresses can be worked out with pen and paper.

The column on the WM18 is much narrower fore and aft than the width, So it is poor at resisting fore and aft bending due to the weight of the head and vertical forces resulting from, say, drilling.

Yer pays yer money and takes yer choice!

Andrew

[Ex contributorParticipant @mgnbuk]

As often seems to be the case, you appear to be trying to correct one error by introducing another – in this case mucking up the column to table squareness to "solve" a spindle alignment issue.

As a former colleague ( a machine tool fitter who did his apprenticeship at Asquiths in the late '50s) used to say – "build 'em right from the base up or you are just making problems for yourself " So …

First check that the table top is parallel to the base in both X (table left-right) and Y (table front-back) directions. Adjust as required.

Then check that the column is perpendicular to the table top in both planes (used to be refered to as "front square" (X-Z) and "side square" (Y-Z) at my last employer) using a known good square of suitable accuracy. Adjust as required.

Then check the quill alignment to the table top in both planes using the square. Adjust as required.

Finally, check the spindle alignment (trammel). Adjust as required, but adjust at the point that doesn't muck up the previuosly set alignments – it is pretty pointless having a spindle that "trams" perfectly if the column or quill is not square to the table top.

We did, on occasions, tweak different alignments to one end or the other of the acceptable range to get machines "in spec" overall, but as the acceptable limits were (at their most lenient) around a thou per foot (and frequently half that – or tighter still if the ward "borer" appeared in a machine description) there was not much leeway. Better to fix the actual problem at it's source as you work upwards than try to introduce errors elsewhere to compensate.

There are no shortcuts to doing it right – if there were, the Yorkshire-based machine tool builders would doubtless have found them by now !

Nigel B

(who, in a former life, had to get rebuilt CNC machines accepted by some very picky customers)

[ChrisBParticipant @chrisb35596]

Posted by mgnbuk on 10/11/2019 12:22:32:

As often seems to be the case, you appear to be trying to correct one error by introducing another – in this case mucking up the column to table squareness to "solve" a spindle alignment issue.

 

Below is an extract from the PM-30V manual which is a similar mill to the WM18. Basically it's telling you to move the column to tram the Y-axis – now instead of shimming the column I jacked it – same result as shimming I think. But still the quill out of alignment will not be resolved by the procedure in the manual as you clearly explained Nigel.

"A procedure similar to the above may be used to check tram
in the Y-axis, front to back. The difference here is that there
is that Y-axis tram is established in manufacture, and can be
adjusted only by these shop methods:

1. Shimming between the dovetailed Z-axis saddle casting
and the headstock itself. It is more likely that the headstock
is nodding forward rather than leaning backward,
so start with (say) a 2 mil shim in line with the underside
(central) headstock nut. This is a temporary fix that will
need to be checked if the headstock is tilted again.

2. Shimming between the underside of the column and the
main base casting. This is a long-term fix. It is a two-person
procedure, requiring an engine hoist or some other
means of un-weighting the headstock (see Section 1, Installation).
As an alternative to shims, which do not provide
a uniform bearing surface, consider injecting a metal-
filled two-part polymer. "

 

So the question is, is there a way to adjust the quill?

I have checked the table movement and the difference I could read in the FWD to AFT direction was 0.02mm over the whole travel – which I should regard as negligible. In the Left to Right direction did not get any deflection.

The column is square to the table in the Left to Right direction. In the Fwd to AFT direction it is tilted 0.02mm over a height of 80mm towards the front. (negligible?)

The quill is out by 0.30mm towards me over the same height of 80mm, which compared to the column out of squareness is enormous. Have no idea how to adjust this.

Edited By ChrisB on 10/11/2019 16:04:09

[ChrisBParticipant @chrisb35596]

Can anyone with a WM18 or similar mill measure their quill out of squareness for me please? I have a feeling that this movement is coming from the rack and pinion pushing the quill towards me.

I'm tramming the head with the quill retracted and locked, this gives me 0.02mm left to right over an 8 inch diameter, front to back I'm getting 0.05mm…I would be happy with these readings, however when extending the quill to its max extension and locking it the readings change drastically, the left to right becomes 0.05mm, but the front to back goes over 0.30mm.

If I use it with the quill retracted I think I should be fine but I would like to know if this is normal for this type of mill or I'm getting this totally wrong.

[not done it yetParticipant @notdoneityet]

As I see it, the proof of the pudding, as they say, means trying it to see what you get.

I don’t extend the quill any more than absolutely necessary when milling. So no error at full stretch.

Drilling, to see if it drills straight through a deep piece is the test of ‘nod’. Small drills will wander, but finishing a hole with a milling cutter should show if the machine is sufficiently accurate. But heaving on the quill to make a dull drill cut might well alter the geometry rather more than a light touch with a keen cutter. I drill by raising the knee if I need to rely on a straight through cut with a drill. The quill is there for situations when it is needed or when the job will make no difference.

When thinking about a mod such as altering the column nod, I always consider how it would be adjusted if out in the other way. The mod that could add or take away ‘nod’ would be the preferable method. Shimming fits the bill for either situation, while tensioning the column to pull it forward would be more difficult. So you can guess how I would fix it, if indeed it needs fixing.

[Ex contributorParticipant @mgnbuk]

So the question is, is there a way to adjust the quill?

To be able to swivel the head, you have two parts – the base casting that moves on the dovetailed column & the quill carrier / spindle motor mount casting. One of these will probably have a spigot to locate in a hole in the other to act as the swivel bearing & a couple of tee bolts to hold the two parts together. From your description, it would seem that one or other (or both) of the faces on these parts that bolt together is out of alignment. You will have to dig deeper to find out where the error lies in order to determine how to improve the situation

I would proceed like this :

Firstly, completely separate the two castings, thoroughly clean both faces of paint, swarf etc. and deburr. Might be worthwhile to blue up one part & offer it up to the other to see how well the two faces fit together. Re-assemble and recheck the head / quill / spindle alignments to see if the test results have changed. Bear in mind that to separate the two parts to work on them you may need to disconnect wiring & that the quill carrier casting is probably heavy. If your alignments were OK before you set the head at an angle, it might be just contamination of the joint that has caused the current issue – you might get lucky !

If the results are still not acceptable, split the castings again and carry out a check on the base casting joint face to make sure that it is parallel to the dovetail ways / column. Put a dial gauge magnetic base on the table, adjust to touch the joint face at , say, the 3 o'clock position. Zero the dial gauge after locking the Y axis, then manipulate the X and Z axes to move the dial gauge around the joint face to the 6, 9 & 12 o'clock positions. If the 12 & 6 o'clock readings are different from the readings off your square, it would suggest that the joint face is not parallel to the dovetails. To keep track of the numbers, use a marker pen to write directly on to the casting face – saves messing about with scraps of paper.

If you have a surface plate, place the quill carrier casting joint face down on the plate – if there is a spigot to centre the casting preventing the joint face sitting on the plate, use parallels. Fully extend the quill & lock it. With the dial gauge mounted on a surface gauge (or use a magnetic base on wide parallel) slide the dial gauge back & forth across the top of the quill nearest the casting & set zero. Repeat the check at the outermost end of the quill – the quill should be parallel to the joint face. If you don't have a surface plate, there may be room to do this on the table top, which should be flat.

It would probably be worth while to check that the outside of the quill is parallel & check the fit of the quill in the casting bore with the return spring & pinion removed. Bit difficult to sort a loose quill, though, as you need a "putting on" tool ! At the last employment we sent them (and tailstock barrels) out to a specialist firm who honed the bores, hard chrome plated the quill & ground back to get the correct fit. As you might expect, not a cheap option.

I would probably scrape out any errors in the joint faces, as the amounts to be removed are probably not that great, though it may be possible to machine them – depends on your equipment etc. It is worthwhile double checking your readings before putting scraper to metal – it is easy to knock the dial gauge, so I always go back to the initial zero setting position to check it is still zero. If you don't get reliably repeatable results, re-assess your methods.

HTH

Nigel B

[ChrisBParticipant @chrisb35596]

Thanks Nigel for the detailed explanation. Will try to get the head off and check if there are any nicks or anything under it. Below I took a short video clip of what's happening, the first part is the travel of the head up and down movement, the second is the Quill. It looks a bit strange.

I'm using a precision vice on its end as a square, might not be perfect but it's close enough, it's about 200mm tall.

Edited By ChrisB on 11/11/2019 21:33:24

[ChrisBParticipant @chrisb35596]

This morning I had another attempt at sorting the mill tram, following the advice on this thread and other videos of similar mills on youtube.

Started off with levelling the bench and table which were out so I had to adjust a bit. Then checked the table and column, both were acceptable, the column has a nod of 0.05mm over a height of 200mm – I shall call that acceptable. Then I proceeded to tramming the head, left to right was easy enough, but front to back was a bit of a work out. As the column alignment was fine I was not going to touch it to fix the front to back tram, so instead I shimmed the head. I had to place a 0.2mm shim under the lower part of the head where the head lock bolt is located. Not an easy task as I had to do it about four times, and each time I had to support and slacken the head enough to slip the shims in.

The adjustment was done with all locks on the column and quill tightened and the quill at the retracted position. I could not find a way to fix the quill misalignment (0.3mm over it's full travel) but I reasoned I will use the column rather than the quill to take milling cuts.

I did some test cuts, first with a 50mm face mill and then with a flycutter. Judging from the machining marks it looks like the tram is fine in all directions ( that's no expert judgement btw!)

This is a 50mm face mill, probably I could do with some more speed. No noticeable ridge at the cut intersections.

Flycutter at approx 6" diameter left a nice smooth and flat surface.

[Author]

Posts