Column Alignment with Tilting Head Type Mills

I have a question and a theory regarding the column alignment on tilting head milling machines - I'm referring to a WMD20V but I suppose any mill with a rotating head and fixed/adjustable column.

Preface; The chap I picked the mill up from mentioned that he had issues with losing x,y coordination as he raised the head of the machine (it would drift off in one direction). This was caused because he had (mis)aligned the spindle to the saddle using the printed scale (the source of all the problems) and then using a DTI gauge in the spindle aligned the column & head to the table.

So the spindle remained perpendicular to the table, but the column was tilted so it would move left and right as he raised/lowered it.

I seek a better solution and have read a few methods that rely on the machining accuracy of the parts, digital angle gauges, feeler gauges between the column and base etc. - which may work, but I thought a better method must exist - so I wondered how other people may approach this and could you review my theory below?

My theory is that I could mount a suitable test bar (£35 delivered) in the spindle and using a simple jig I can slide my DTI gauge parallel to the dovetail. I could measure the stationary spindle to column alignment and set that (maybe rotate the spindle 180° to 1/2 some of bearing/test bar alignment error if any.... ?), and then using the DTI in the spindle as per the norm - align the column to the table and Bobs your uncle ?

What do you think?

Edited By Russ B on 20/03/2014 12:04:52

Russ,

What you describe seems to be a viable alternative to the usual "tramming" with a DTI ... but I'm not sure if it offers any improvement.

The important thing that you have highlighted is that "two wrongs CAN make a right" ... within a limited set of circumstances.

The previous owner had evidently made two angular errors, which cancelled-out: The interesting bit, I think, is that in that condition movement of the Quill remains correct because [being subject to the sum of the two angles] it is running vertically, albeit laterally displaced from its "design position" ... but, movement of the Head on the not-quite-vertical column will disclose the angular error [because it is only subject to the effect of one angle] by producing a lateral shift.

The process you describe should work, but you could probably achieve the same result by tramming; provided that you check at more than one position of the [Mill] Head.

MichaelG.

Thanks for the reply, I'll need a while to digest and ponder on that, I'm still not sure how tramming with more than one position of the head will do it - could you expand a little more

(unless your talking about accurate x,y coordination and comparison of a projected point from the spindle axis to the table - opening a whole new can of "how to" worms - especially without a dro?)

My problem is, I'm dreaming up these methods of metrology with experience from a different line of engineering which I'm more familiar (but less natural) with (Civil Eng and surveying), I'm not sure how applicable my theories are here but it sure is fun.

Hi Russ,

I'm not sure what the configuration of the head to the column is on the WMDV20. I have a Warco WM16 which has a rotating head as opposed to a fixed head on a column that rotates at its base like some other makes/models.

What I did when I set my machine up was to clamp an engineers square vertically to the milling table and hold a DTI in the spindle. The DTI runs up/down the vertical edge of the square when raising/lowering the head. This tells me if the column is perpendicular to the table or not (assuming I have a perfectly square engineer's square).

After adjusting out any tilt in the column I then concentrated on tramming the head and spindle assembly to ensure that the spindle is also perpendicular to the table.

Hope this helps, I must admit I'm fairly new to machining.

Regards,
Ed.

Russ,

Our electricity has just come back on, after a power cut ... so this corrigendum is a little late.

Although I stand by the "discussion" in the middle of my previous post; I realise that my conclusion, about tramming being equally effective, is probably wrong.

This is a very interesting "puzzle" and I will give it some more thought.

... Going back to sleep for a while now !!

MichaelG.

Ed - I see thanks, yes that should achieve a fairly good level of column to table alignment without the need for special tools - albeit relying on the accuracy of the square and table -

this is basically the same method different exicution (relying on no special tools/test bars), so (to aid people are maybe struggling to visulise this)

> rather than running a dti up and down the column using a custom jig, while reading alignment on a test bar in the spindle - thus aligning the spindle to column ways - then tramming the spindle to table which should result in a square column less any misalignment of spindle to column.

> your running a dti up and down the column held by the vertical saddle (either in the spindle, or magnetic base to the saddle perhaps?) and running up and down an engineers square clamped to the table - thus aligning the column to table, then aligning the spindle to table which should result in the column and spindle now being aligned less any misalignment in the column to table.

Is everyone on board

MichealG - I think maybe this is what you where aiming at possibly - sorry for confusing things.

I think we've highlighted that 2 wrongs DO appear to make a right if your just tramming the spindle to table, so the quick set square bolted to the table method should really be used for the 1st assembly and at suitable intervals thereafter - I'm not sure if my wacky method would be any more accurate - I might have to put it to the test!

Thanks for the input - this problem is certainly more than halved!

Russ,

I've not found any useful views of the WMD20V, but am I right in assuming that the column assembly is similar to the "old X2" ? ... as illustrated on page 3 of this.

MichaelG.

.

Russ,

I think your surveying background may stand you in very good stead ... Optical methods are generally the "best" way [in terms of results vs convenience] of checking angles.

MichaelG.

Eughh! no no no - those tilting columns are a sin IMO, and I'd go as far as saying a discredit to Sieg's engineering ability - 1st hand experience sadly. The number of large components and castings required to achieve that is more than it would have taken to incorporate a swivel mechanism between items 49 and 63 on the link you send (spindle box, and spindle box base - or vertical saddle as I would call it)

The head rotates about an axis on a vertical saddle see page 45 here **LINK**

It's a much more substantial machine than the X2, its more of a match for the X3 - its a little lighter and little cheaper, but equally capable vs the X3 and many say more rigid and reliable - and better finished (hand scraped on all mating faces) which I can't comment on having never seen an X3 in the flesh.

It seems to me that your problem is illustrated in diagram A below. Your DTI method will align the spindle with the column, as in B. Then, by normal the tramming method, you will tilt the column until the spindle is vertical, as in C.

Not knowing the mill, I don't know how you will tilt the column, but the method is sound.

Gary

.

No offence intended, Russ ... and thanks for the link to the beast in question.

However; if I understood your original post correctly; the column on your machine is not accurately perpendicular to the table ... It is evidently not a "tilting column" so it's an assembly error.

The good news is that once you've got it set right you won't be moving it.

MichaelG.

Hi, here is a **LINK** to the WMD20V, which I believe is similar to the Chester 20V seen here **LINK**

Regards Nick.

delete

Edited By Nicholas Farr on 21/03/2014 09:44:18

no offence taken at all - its just my oddball sense of humour!

The column does not tilt to that degree, sorry for the misleading description, it is bolted straight to the back of the base without any locating pins etc, so the user can adjust the alignment - I will use Ed's method to get me going, and if I see a 2 Morse test bar come up cheap I'll grab one and test my theory while hopefully checking how well Ed's method aligned the column and table - I will use Loctite between the column and base to give me less chance of it slipping.

Nicholas, yes that's the one, although the photo on Chester's website I believe is the old design control box layout (they called it the 20VS I think) the 20V they sell will probably have a larger spindle speed readout as shown on Weiss' website - it is purely cosmetic though, the machine is the same and the controller should be the same KBIC unit + it will have the long 700mm table, not the 500 shown on the Weiss site

.

Russ,

There is a wise old saying ['though I can't remember whom to credit]

... If you can't make it right, make it adjustable ...

MichaelG.

[proud owner of a BCA Mk3 ... small capacity but perfectly formed]

Edited By Michael Gilligan on 21/03/2014 10:27:13

Russ

Your method of using a square is fine but if you can use a large angle plate (of known accuracy) instead it will be even better. Whilst at it check the other plane, as if it is out in one direction chances are it may be out in the other direction. I would forget the loctite for two reasons the first being if you should need to part it in future it could be a bit of a so and so. The second reason is that it is likely that the casting has been contaminated by oil and you will not be able to secure a bond, no matter how well you clean it. My method of corecting the fault would be by shimming, that is unless you are very conffident about your ability with a file.

I'm pretty handy with a file so that would be my chosen route I'm not a fan of shimming for anything other than a temporary fix unless the shims are pretty substantial, regarding an angle plate or square I'll probably go with that but I'd prefer a measurable method (which is why I thought up the test bar in the spindle method)

I will go with Ed's engineers square bolted to the table method, just to get me going, and it will be interesting to compare result with a lathe test bar and jig at a later date

Russ

It may seem stupid, but have you taken the machine apart cleaned and examined the mating faces, even run a smooth file over them just to make sure there is nothing emmbeded in the surfaces. Assemble the machine and check again. If you have please forgive me, but it is often silly things that cause some of these problems.

Hi Dusty,

I'm in the process of a complete strip and rebuild including 10k rpm spindle bearings and a quick blue and rescrape of the ways if necessary (as the current hand scraping looks a bit coarse for my liking), my concern was with the accurate reassembly and alignment of the machine with the swivel head (as I said I prefer measurable alignment as opposed to relying on something else to be square wherever possible)