What's new
What's new

Lathe Chuck Design

Badbascom

Plastic
Joined
Mar 2, 2014
Location
Hondo, TX
I am not super familiar with lathes but I would like to know if a lathe chucks exists that has fine adjustment screws or similar for centering and tilt (edit the screws are on the outside for bolt tilt and centering so they are accessible after the part is chucked in). Preferably where each adjustment screw pushes and pulls so you don't have to loosen the opposite side like a 4 jaw chuck. After we turn a part it is inspected by placing on an air bearing rotary table that has two hand knobs (X & Y) to adjust the centering and another two that adjust the tilt. These are adjusted in unison until software says it cannot achieve any better due to flatness and circularity issues. This machine measures runout to the 1/100,000 so I would like to be able to adjust my lathe to similar standards. The 4-jaw chuck is not cutting it.
 
If I read that correct you are trying to get to 10 millionths? Is the spindle on your machine even capable of that? I've never had to do anything that fussy but imagine if you fart in the wrong direction your part is scrap. Sounds like grinder territory to me
 
If a 4 jaw can't do what you require, how about collets? As suggested, maybe a "Griptru", but still a mess on to get it to run true. What is the job which requires such accuracy from a chuck?
 
yea, I think getting that accuracy from a chuck no matter the number of jaws or the adjustment system, is known as "barking up the wrong tree" !
 
I think Hardinge guaranteed 40 millionths on their lathe spindles... You could just get a Moore Nanotech lathe, I'm sure they can advise on chuck options. At 10 millionths you better have surface finish to go along, the Moores can do it.
 
OP:
I know of nothing that has "tilt" in the chuck.

Likewise, in a lathe, I cannot see tilt as being desirable.
(adds axis, aka pwr 2 less rigid).

6-jaw chucks can easily adjust to sub-micron positioning.
50 millionths = 1 micron, 10 millionths = 0.2 microns.

A 4-point or 6-point adjustable chuck can easily be made able to adjust to 0.2 microns via differential screws and/or offset levers.
I would quote 4000$ for one, 165 mm D, plus 4000$ NRE, as a ballpark.
With customer supplied metrology and gages.
 
The direction you should look will be the least expensive home made shrink chuck.
Very simple mount a disk of steel to your spindle, bore the disk to just under your shaft size, gently heat the disk, and slip you shaft in.
You shaft will now run as accurate as your spindle bearings, not the spindle itself.
You have some control as to how much shrink fit, some materials are sensitive to stress risers, that eliminates jaw chucks.
 
Yes, they make chucks that you can dial in. The magic is in the back plate, not the chuck itself. They go by names like "True Set" or "Run True".

BUT...

All the ones that I have seen only allow you to dial in the concentricity, not the angle. The best way that I know for controlling the angle is to mount a back plate on your spindle, being very careful that there are no foreign substances between the back plate and the spindle, and take a light cut with it in place to bring it true to the spindle's rotation. Then, mount a high quality chuck on that "dialed in" back plate.

And again, all the ones that I have seen have four adjustment screws that operate just like the jaw adjustment screws on a four jaw chuck. They do not provide the mounting pressure so they will have a finer thread which allows more precise adjustment. They still have to be loosened and tightened in pairs but the final tenths are often done with just increasing the pressure on one of them. Then, when all is OK, the bolts that hold the chuck to the back plate are tightened. Perhaps you could make one with two screws and two springs. I've never heard of this being done so, if you do, let us know how it works out.

I guess you could add three or four screws that would allow adjustment of the angle, but there probably is no way to eliminate interaction between the various screws and they would probably be more trouble than the added adjustment is worth. Just my opinion.

I have a shop made back plate that does the above with four set screws. It was not difficult to design or make. I did use some spring washers on the six bolts that hold the back plate to the chuck so they stay in contact when those bolts are loosened for adjustment. Here is a photo where you can see two of those four adjustment screws:

attachment.php


Those adjustment screws simply rest against the OD of the chuck.
 
As mentioned, your lathe spindle is almost certainly not precise enough to allow a 10 millionths adjustment to work, regardless of the ability of the chuck to do so. If you really need precision in that range, you'll likely have to go to an air or hydraulic bearing spindle of similar precision to the inspection table you're using.

Along with that, you'll need lathe ways of straightness and uniformity to make use of the super-spindle, so essentially one of these: Nanotech 450UPL V2 Ultra Precision Lathe | Moore Nanotechnology Systems

In the mode of Hanermo - I can make one for you for the modest price of $150,000 USD, with a NRE of $30,000. Lathe chuck would be an extra accessory... :D
 
I am not super familiar with lathes but I would like to know if a lathe chucks exists that has fine adjustment screws or similar for centering and tilt (edit the screws are on the outside for bolt tilt and centering so they are accessible after the part is chucked in). Preferably where each adjustment screw pushes and pulls so you don't have to loosen the opposite side like a 4 jaw chuck. After we turn a part it is inspected by placing on an air bearing rotary table that has two hand knobs (X & Y) to adjust the centering and another two that adjust the tilt. These are adjusted in unison until software says it cannot achieve any better due to flatness and circularity issues. This machine measures runout to the 1/100,000 so I would like to be able to adjust my lathe to similar standards.

There is a forum member called "ASQL2580" that does this kind of work. The thread to look for is titled "Fanuc 18 instability" His employer uses a air bearing supported chuck mounted on a Moore G18 base to improvise a high precision lathe. The pictures in the thread will give you an idea of how one shop produces high precision parts. The key point is that you need to use a air bearing turntable for the lathe headstock if you want air bearing accuracy for the part. A oil hydrostatic spindle might approach this accuracy.

The adjusting screws on the air bearing tables are 40 threads/inch. They will not take a lot of abuse. They would not be suitable for adjusting the position of a heavy chuck on a "Adjust True" type backing plate.

You could make a equivalent design by using screw adjusted low angle wedges to move the chuck about. The wedges would support most of the chuck weight. The wedges would also reduce the effective lead of the fine thread adjusting screws.

There are other designs using differential screws. For example a 20 pitch external threaded bolt with a internal 22 pitch threaded bore provides the same lead as a 220 pitch bolt. The differential screws require some provision for radial bearing support for both the primary and secondary screws. Otherwise there will some wobble in the differential motion.

The Thor Optics company sells the fine adjusting screw assemblies ready made.

One significant difficulty with these adjustment mechanisms is that they also need a non influencing locking device to keep the chuck in place once the adjustment has been made. The locking device needs to support the machining loads rather than the adjustment mechanism.
 
You seem to be saying opposite things here. In the early paragraphs you talk about a 40 TPI thread, diameter not specified, not being capable of making the adjustments. In addition to the diameter of that thread not being specified, no mention of the size of the chuck or of the size/weight of the work. A 4-40 would, indeed be somewhat wimpy but a 1/2-40 would be another thing and a 1-40 would be capable of exerting a lot of force. I would say that weather a 40 TPI thread could be utilized to make the adjustments would depend in large measure on the diameter of that thread and the number of threads engaged.

Then in the final paragraph you correctly state that the two functions, adjustment and locking, should be handled separately. That is the key here. The back plate that I showed in my post above uses 10-32 set screws to adjust the chuck and they are completely adequate for that 6" chuck and almost any work I will ever hold in it. And they can easily make adjustments in the tenths range. That back plate was made in 2005 and, to date, the adjustment screws show no signs of wear. There are SIX much larger screws, 3/8" IIRC, that secure the chuck to the back plate. It is those 3/8" screws that hold the chuck against the weight of the work and the forces involved in cutting.

I used 32 TPI adjusting screws because they were a readily available fine thread and I did not have any delusions about my SB-9 having any accuracy in the millionths range. But 40 TPI would not be that much smaller and with a sufficient diameter and more than 3 or 4 threads in engagement, they would be quite sufficient for chucks up to 10 or even 12 inches. They would NOT be used to lock the chuck down for actual work.



There is a forum member called "ASQL2580" that does this kind of work. The thread to look for is titled "Fanuc 18 instability" His employer uses a air bearing supported chuck mounted on a Moore G18 base to improvise a high precision lathe. The pictures in the thread will give you an idea of how one shop produces high precision parts. The key point is that you need to use a air bearing turntable for the lathe headstock if you want air bearing accuracy for the part. A oil hydrostatic spindle might approach this accuracy.

The adjusting screws on the air bearing tables are 40 threads/inch. They will not take a lot of abuse. They would not be suitable for adjusting the position of a heavy chuck on a "Adjust True" type backing plate.

You could make a equivalent design by using screw adjusted low angle wedges to move the chuck about. The wedges would support most of the chuck weight. The wedges would also reduce the effective lead of the fine thread adjusting screws.

There are other designs using differential screws. For example a 20 pitch external threaded bolt with a internal 22 pitch threaded bore provides the same lead as a 220 pitch bolt. The differential screws require some provision for radial bearing support for both the primary and secondary screws. Otherwise there will some wobble in the differential motion.

The Thor Optics company sells the fine adjusting screw assemblies ready made.

One significant difficulty with these adjustment mechanisms is that they also need a non influencing locking device to keep the chuck in place once the adjustment has been made. The locking device needs to support the machining loads rather than the adjustment mechanism.
 
There are chucks that will do centering and tilt, but I've only seen them on ornamental lathes like Leinhardt or Holzapffel and they see light cuts and maybe 5 RPM.
 
Thanks everyone for some great ideas and discussion. The part I turn is an aluminum bushing for lack of a better term. The bushing comes with the ID already sized, my job is to turn the OD down to match a corresponding bore with the TIR of .001 between the ID and OD. This may sound easy however the ID is not perfectly round which I can see on the runout machine plus the ID has intricate fins.

The first attempt was to use an expanding arbor however the alignment was the problem and the part failed. We finally turned down an aluminum arbor of sorts which the bushing holder pressed over and without disrupting the setup, we pushed the bushing over the holder and did the machining. The bushing marginally passed. If I were to remove and re-install this homemade arbor the alignment would be ruined.

My thinking was that a good hydraulic expandable arbor with a X & Y tilt and center table would be the trick.
 
Thanks everyone for some great ideas and discussion. The part I turn is an aluminum bushing for lack of a better term. The bushing comes with the ID already sized, my job is to turn the OD down to match a corresponding bore with the TIR of .001 between the ID and OD. This may sound easy however the ID is not perfectly round which I can see on the runout machine plus the ID has intricate fins.

The first attempt was to use an expanding arbor however the alignment was the problem and the part failed. We finally turned down an aluminum arbor of sorts which the bushing holder pressed over and without disrupting the setup, we pushed the bushing over the holder and did the machining. The bushing marginally passed. If I were to remove and re-install this homemade arbor the alignment would be ruined.

My thinking was that a good hydraulic expandable arbor with a X & Y tilt and center table would be the trick.

The cheap solution:

Make a long hollow cylindrical holder with a front and back row of four radial adjusting screws to fit the headstock taper. This could also be a thick wall pipe with set screws held in a 3 jaw chuck. Support the arbor by placing the end in the cylindrical holder and center by adjusting the 8 set screws. You will be able to set both the center line and the tilt of the arbor.
 








 
Back
Top