Can't hold tolerance with finish cuts

[BadDog]

Again, play in the lead screw is the norm. The only question is, how much play you have, it's always there; but the truth is that unless it is truly excessive it doesn't much matter in this context, and wouldn't produce your results even if it were excessive, . And if something is pulling it into the cut, then screw/nut wear is really irrelevant to fixing the problem.

To recap, there should be no force working in that direction to move the cross slide if the screw and nut were not even installed. And if you are still not convinced, then consider that an experienced machinist can do quite reasonable (and even great) precision work on a relatively clapped out lathe with tons of wear and "back lash". That's because the impact is generally predictable and consistent, yielding to application of good practices, and that is not this. The real question to ask is, "what force is causing it to move into the axis center line." Figure that out, and how to neutralize it, and you should move directly to decent predictable results, and from there learn how to anticipate and neutralize the flex that is always there to get into more consistent and close precision results. But with it pulling inward, you're not getting anywhere until that is solved.

[spro]

Is it that we are looking at pure manual feed and traverse while the lead screw is turning ? I mean, it has a single lead screw with a slot and "it jumps .002 ..." what else could cause that but the feed worm? I can't say.

[BadDog]

If he's doing standard turning (i.e. just reducing the diameter of a cylinder), the all inputs, manual or otherwise should be along the axis. And yet, something is causing the cross slide to move in toward the axis and reducing the diameter more than intended. Other than negative lead angle or excess positive rake, I don't have any idea what could be the motivating force that moves the cross slide inward. Even if you didn't follow proper practice in taking up slack by setting the cut with an inward movement, the result would be moving out of the cut, not in. That's because all cutting forces generally move you away from the cut, not in.

Hmm. Not sure how it would happen, but a rocking motion could also do that. Imagine if the cross slide were bowed along it's length. Then the downward component of cutting forces could force the end down onto the apron ways, rocking across the fulcrum, and thus moving the cutter into the work. I haven't done the calculations, and don't have measurements to base them on even if I wanted to, but to get 0.004 of inward movement at the cutting edge there would have to be quite a lot of rocking going on. But that could easily be tested with another pass with dial indicator checking the cross slide for vertical movement at the end away from the operator. I suppose you could contrive a similar scenario with the apron twisting (or squatting down on the back) but that one seems even more implausible.

Another hmm. Maybe if the cross slide were excessively tight to move? I could imagine that causing the screw to bind/bow a little as you wind inward to set the cut. Once the cut starts, cutting forces/vibration could be allowing the stored energy to release causing the slide to move inward. But that would still require that either the bind resisting inward movement would have to exceed the normal cutting forces trying to push outward, so the bit profile would still be involved to minimize cutting forces, though that would eliminate the requirement that the bit be "hogging" to produce the movement.

But I can't come up with any scenario where a worn or loose screw does anything but allow the cross slide to move out of the cut, never in.

[John Hasler]

Seems very significant that the cross-slide jumps inward when the motor is started with the cutter not engaged.

[GlennW]

Possibly a worn out or improperly adjusted gib.

That tool post setup extending the tool way foreword isn't helping much either.

In other words, perhaps there is some rotation of the cross slide or carriage when the carriage traverses and the tool engages the stock.

[Glenn Brooks]

Spro, Russ, John, Glenn and Pete. Appreciate your comments and discussion. It's really very helpful to be able to discuss possible causes and how to solve this. I bought this machine specifically to improve accuracy and get away from never ending tolerance issues.

I ran around 200 test cuts over the last few days, the last batch of 100 or so made with a dial indicator set on various (all) parts of the crossslide. I Checked for tool movement into the work also. Generally everything is normal, except for the forward movement of cross slide (not underlying saddle- just the cross slide), and a noteworthy in and out 'wavering' motion of the bit, as the cross slide traverses along the axis of the work. (Note: the compound and tool post are locked down to the cross slide - no unusual movement- it's all in the cross slide...)

I called Miller Machine and Fab - they make and sell replacement cross slide screws for old South Bends. The advised if the dial indicator shows the bit is wavering .001-.003" in and out when making the cut, then certainly something is amiss in the nut.

So After stripping out the screw and nut, my detailed inspection of the cross slide screw and nut show a few obvious wear patterns.

- shank of the screw at the two rear thrust bearings is worn down noticeably- a bit smaller diameter and fingernail rough at the least. So not a tight fit at the thrust bearings.

- threads in the nut are worn stepped shaped. By this I mean the screw has cut two separate steps into the shoulders of the thread pattern. Think of the side view of the Eqyptian pyramids - where the blocks of the pyramid jut out of the slope of the structure. I surmise this may cause random misalignment of the screw threads, if the major diameter rides up, or is pushed around by these little notches. Maybe the screw threads jump or seat into or out of these little ridges when the machine starts up and tooling enters the work. It they jump out of the seat, to the wider diameter at the root, would that not deepen the measured depth of cut??

- Threads of the screw itself exhibit up to .006" wear in the center to end of the screw -where most of the work on the lathe occurs. The wear is not uniformly distributed along the length of thread.

- By hand, the nut wobbles up and down and side to side by hand...

- rear most threads in the nut are worn nearly to a point. Front side of nut more ACME profile normal.

Anyway, I ordered a new screw and nut from Miller Machine and will see what happens. $275 seemed a reasonable price to pay.

Shouldn't be rocket science, but apparently with this machine, it is... either that, or I am just dealing with some form of alternate universe.

Hopefully, Monday or Tuesday will tell the tale...

Glenn

[spro]

YES. Possibly trapezoidal thread was generated by the cross slide screw being bent. As the bent thread works its way into the nut etc.

[BadDog]

What you describe is indeed severe wear, but I'm still not seeing where the force moving the cross slide deeper into the cut is coming from. If you approach the cut properly (feeding inward to take up the slack in this case), then the thread should be seated against the far side of the nut, presumably in the wear surface that it has created. Even if there is a ramp ramp that would be traversed thus pushing the nut and attached cross slide toward the axis should the screw be forced sideways in the nut, what force is pushing it sideways hard enough to overcome friction AND the step in order to shift it outward and up onto the step?

It still feels like there must be something else going on here because nobody has explained what force is acting to overcome both the cutting forces that generally push the tool away AND friction in the slide. And apparently doing so simply by starting the spindle before the cut is entered. The only guess I have at this point is that the maybe the excessive wear is somehow causing the screw to bind and "wind up" somewhat, then unwinding when vibration causes the friction to drop? Or maybe the screw has a bind that rides in the wear groove as you wind in the cut, then likewise the vibration lowers the friction allowing the screw to drop down (or up, or sideways) onto that step you describe in the nut?

[spro]

Russ you explained it. A vibration across a tensioned, bent screw. We are to know the threads have been reformed to ramp shape. That takes some continual pressure to do.
Glenn B. You pulled the trigger and may have solved the problem.

[stephenc]

Like a lot of you here I don't have a lot of experience with machine tools , so I follow these threads to glean all I can in case I ever have a similar problem of my own .

In this case I became interested enough to search and find an exploded parts diAgram of this machines cross slide assembly so I could sort of follow along .
And that leads to a question I have regarding the taper attachment.

Maybe someone can point me towards the possible error in my thinking here .
But I noticed that the lead screw assembly looks like it has the thrust bearings used to take up end play and backlash on the taper attachment side and not the dial side .

And please someone correct me if I am wrong here , but to my thinking in order to take up any backlash in the lead screw you would need to go in the opposite direction that you would with a normal cross slide ... i.e. .. turn it in towards the work and then back out to take up backlash .

I am plenty sure there is something I am missing here so please correct my thinking

[BadDog]

Some have the thrust restraint on the dial side, some have it on the back side (including most all that have taper attachments, particularly with telescopic screws). Rear restraint is generally preferred (and found on better machines) not only so that it can potentially be controlled from the taper attach (while still being adjusted from the fixed crank), but also because tensile loading of the screw is generally going to be superior for stability over columnar loading that would occur if the restraint were on the dial/crank side. In fact, columnar loading would be more likely to let it flex and act like a spring to potentially cause the observed behavior. Also, if it started to bend, it would be more prone to bend more as cutting loads increase, which is why most put the thrust restraint at the rear in the saddle with nut on the cross slide, though compounds tend to be the opposite.

Regardless of which end is restrained, you take up the slack to hold the tool into the work. And you do that by approaching into the cut so that the friction of the cross slide takes up slack in both screw threads and thrust restraint.

[10KPete]

Like a lot of you here I don't have a lot of experience with machine tools , so I follow these threads to glean all I can in case I ever have a similar problem of my own .

In this case I became interested enough to search and find an exploded parts diAgram of this machines cross slide assembly so I could sort of follow along .
And that leads to a question I have regarding the taper attachment.

Maybe someone can point me towards the possible error in my thinking here .
But I noticed that the lead screw assembly looks like it has the thrust bearings used to take up end play and backlash on the taper attachment side and not the dial side .

And please someone correct me if I am wrong here , but to my thinking in order to take up any backlash in the lead screw you would need to go in the opposite direction that you would with a normal cross slide ... i.e. .. turn it in towards the work and then back out to take up backlash .

I am plenty sure there is something I am missing here so please correct my thinking

I apologize for the short post and link but here is my explanation on another board (forum?). I hope it will help.

http://www.practicalmachinist.com/vb/so ... ts-331753/

Pete