Motion when locking the axis

[Bandit]

I was wondering about the cove in the gib for the set screws. There did not appear to be any,( as per Dabbler also). Have found a few gibs that were bent or twisted, appeared to be caused by locks being overly tightened to lock slide when the gib needed to be adjusted.
Maybe remove gib, check for above and for damage caused by lock(s), oil well, gently put back in place with narrow end adjusting screw backed out. What you are looking for is weather things start to tighten up, back out adjusting screw allowing gib to move, do not get things to tight. Don't want stuck.

 

[Dabbler]

The only way for the gib to lift is for either on of its mating faces to have the wrong angle(or not be perfectly flat) or the the same with either side of the dovetails.

Yes, the math seems to indicate that what you model says is is true, but it isn't. Not close. There has to be clearance on the sliding face of the gibb enough to hold a film of oil. Somewhere between.002 and .003 - if there is much less then the gibb/dovetail interface will wear very quickly.

Since the taper angle is very shallow - usually around 2 degrees, that will allow enough lift to 'pinch out the oil' and move diagonally. The cove in the screw prevents this. Without it, it will lift (diagonally) and change the geometry. With wear, in this unnatural configuration, the gibb interface to the dovetail will be permanently altered/ruined.

That is why the gibb height must not have a clearance greater than about .003 to the dovetail height. With the cove and the tiny clearance the gibb cannot move enough to matter.

 

[RobinHood]

Does this logic make sense? Did I misunderstand your shimming suggestion?

Yes, it does. No, you did not.

It is the screw that provides the uplifting force in your force system. This can cause the gib to “jam” because it is tilting up.

Another question: is there a second screw at the other end of the gib?
My concern with the present arrangement since you don’t have a provision for a coved gib (as @Dabbler is suggesting), the movement of the axis in the direction away from the adjusting screw could pull the gib along with it and really jam things up hard as the gib is wedged between the two dovetails.

Perhaps the gib locks have a protruding part into the gib to stop movement axially once it is adjusted?

Perhaps you can pull the gib out and take pictures of both sides of it?

Edit: see post below for answers to above questions…

 

[RobinHood]

Actually found the answer about the gib style used:



There is no need for a “coved part” on the wide end of the gib as there is a counter screw opposite to the main adjusting screw. See above.

So it comes down to A) preventing the gib from lifting up while tightening and B) fit.

Once both of these issues have been resolved, the machine should work properly.

 

[Eyecon]

@RobinHood I'm so sorry I thought this style of tapered gib setup with the front and back screws was more common so I assumed everyone knew what I was referring to by snugging the gib in place
My procedure for setting up the gibs was to push the gib in using the front(thick end) screw while backing up the narrow end screw in tandem until the hand wheels feel heavy( no reference here, just what I consider not smooth enough). Once I get there, I back off the front screw about 1/8th of a turn and push the gib out using the back screw. I repeat this process alternating between front/thick end and back/narrow end screws, until the wheels feel smooth enough, then ultimately lock the gib in place by tightening the front screw slightly.

On the X axis, this worked perfectly, and the slop when rocking the table between the table and the saddle was in the .0005"-.0008" range depending on where I measured this twisting motion. Then I proceeded to do the same thing on Y: got the wheel action to a good place, then snug the gib using the front screw. At this point I noticed the gib lifting...I ignored it and proceeded to measure the saddle to base casting slop, I was reading about .0016" on the indicator. I also noticed the same movement just by locking the Y axis. So I couldn't understand how the gib can be pushed just far enough to get good wheel action yet the saddle was having so much slop(or was it too much slop? hence my earlier questions). When other members confirmed that .0016" was too much slop/play, I figured that the gib lifting and slop were related.
I guess my next step as planned is to confirm this by bluing the Y dovetail and checking for contact on the gib(of course after checking that the static side of the gib is flat on the surface plate). I'm expecting to see either contact in the centre only, or contact towards the narrow end. If this is case I'll try scraping for a better fit and then add a shim to reduce the clearance and prevent any warping. If I get decent contact all along the gib, then it just the clearance that @Dabbler mentioned and I'll have no other option but to add the shim as you originally suggested.

Am I thinking about this the right way?

 

[Susquatch]

I’m totally on board with this approach, I’m just not sure what else I need to do to figure out why the gib is skewing in place as shown in the video other than blueing it. Also based on what PM support told me, what they shipped is a blank: a much longer piece of grey cast stock with the appropriate tapper ground. It needs to be scraped and cut to length before it can be used as a gib. This is why I was thinking to at least start with the bluing process to try and figure out why the gib is moving up and down, try the scrapping process based on what the bluing shows and see if the problem is fixed. This way I don’t have to risk ruining the gib blank.

I am ok with bluing and modelling and any analysis. But I am really against scraping your current Gib until you know much more. To be frank, I'd much rather try scraping the new Gibb than the old one. You can get another new one easily. But once you modify the old one, you cannot get another old one. So ya, by all means Blu and test and measure and model and evaluate and discuss and trouble shoot. But please put the scraper away for now.

Based on the rough kinematic setup I did in Fusion360, the only way the gib can move up and down as shown in the video is a) there is a high spot region in the center of the gib and it’s only making contact in the center leaving the wide and narrow ends free to move or b) the overall taper angle is slightly shallower than the mating surface effectively causing it to only make contact at the narrow end, leaving the wide end I was trying to snug free to move up and down as it did. The static face of the gib will be first blued against a surface plate to make sure that it is at least flat. All this assumes both castings dovetails are flat. I believe that @PeterT already said that is it very unlikely that the casting have any issues and that the gib fitting is the most likely culprit. This is what I’m hoping the bluing process will show.

From my simple minded assessment, the Gibb isn't tall enough to stop vertical movement. The wedge shape causes it to lift when its tightened. No amount of material removal will fix this.

But I'm not at all sure that fix is the right word to use here because that assumes that it's a problem. Gravity will pull the Gibb down when it's loose, and screw force will push it up against the mating ramp. That's only a bad thing if it changes a lot as the table is moved along its axis. Many many many machines don't even have tapered Gibbs.

What happens to the Gibb as the table is moved back and forth? Will the screw provide differing lift at various points along the table travel? Same question for screw torque. Is its tightness variable?

Have you taken the table off yet to examine the Gibb fit and installation and dovetail fit along its length visually? I get the advantages of using Fusion to model everything, but sometimes laying an eyeball on things is the best modelling process ever invented.

Edit - I didn't see @Dabbler or @RobinHood posts before I posted this. They make very good points.

A good question for PM is "What is the vertical clearance in the Gibb channel supposed to be"? You are getting WAAAY more vertical movement there than what Dabbler suggests it should be. So it would be good to know what PM thinks it should be.

I like the idea of shimming to test it too.

 

[Eyecon]

A good question for PM is "What is the vertical clearance in the Gibb channel supposed to be"? You are getting WAAAY more vertical movement there than what Dabbler suggests it should be. So it would be good to know what PM thinks it should be.

I like the idea of shimming to test it too.

Unfortunately PM support initially said that this lift was normal and only agreed to send the replacement gib blank after I stated asking more questions and challenging their logic. I asked about clearances and taper angles but they didn’t have any of this information which is what brought me to this forum to ask the community experts.

In terms of procedure I think we are all on the same page, provided there is nothing funky going on with the gibs mating surfaces(sliding and static), the likely issue here is that the gib is simply too short and is allowed to lift, so shimming is the way to go until I get a new gib blank. I’ll also compare this gib to the one that’s behaving properly on the X axis to be sure.

 

[DPittman]

Forgive me for about what I'm gojng to do, I'm going to ask a question when I have not read and comprehended each and every previous post....

Is it possible the faulty gib is broke in two?

 

[Eyecon]

Forgive me for about what I'm gojng to do, I'm going to ask a question when I have not read and comprehended each and every previous post....

Is it possible the faulty gib is broke in two?

No the gib is in one piece, I already took it out for visual inspection and cleanup when I received the machine.

 

[Bandit]

Found 2 gibs on a lathe both broke, cross slide and compound. Same person adjusted??

 

[DPittman]

Found 2 gibs on a lathe both broke, cross slide and compound. Same person adjusted??

Ya likewise!

 

[RobinHood]

the likely issue here is that the gib is simply too short

No, not necessarily.

If you have the rear screw making contact at the same time as the front screw is pushing the gib towards the rear (tighter), then the gib is not too short.

The lifting is happening because the gib in not tall enough. It has nothing to do with its length.



The adjusting screw is providing the upwards force - always - when you tighten on your Y axis. Since the gib is not tall enough, the force lifts the gib up. If you shim the top, to take up the head room, it will no longer lift and be pushed in axially and not cock / jam up.

Your Y-axis gib can never behave like the X-axis gib on that mill. In the x-axis, the screw torque will always push the gib down onto the way. They would have to move the Y-axis gib to the left side of the saddle so that the adjusting screw could be on the left side of the gib (like it is on the x-axis), for it to also exert a down force while tightening. Another option is put left handed threads on the Y-axis adjusting screw.

 

[PeterT]

@Eyecon the @RobinHood sketch is exactly what I was driving at with my questions. I don't quite understand your CAD diagram & it might be leading you to a false conclusion, especially with the 90-deg simplification vs. what should be a parallelogram. An unresolved issue is that you have a large gap between the top of gib & the underside of slide. ie. they are different heights & therefore allowing degrees of freedom that should be very minimal IMO. We also still don't know if the gib height is the same height on both ends, you didn't answer my question so it makes it hard to fully understand the situation.

using this sketch green is the big end & blue is the small end. Typically they would be the same height but should extend to where he has the shim pack. If the gib is adjusted axially in & out of the paper, that translates into displacement in the X-direction (blue arrow) tightening or loosening, exactly what we want


what looks to be going on is the gib is displacing along the dovetail axis (red line) as a function of the adjusting screw rotation. Now it gets interesting. I think theoretically if both big & small gib ends (green & blue) rose exactly together there should be no net displacement in X. But if one end acts as kind of a pivot point staying low & the other end is being displaced up via the rotation direction of adjuster screws (further complicated by the adjuster actually pushing on the gib), then I suspect you will have some have some kind of compound displacement where its sticking in one area & not in another. Detecting this with blue rubbing off is meaningless because this displacement mode is not accomplishing the goal of adjusting the gib in the intended way as above. You have just landed at some in-between point where it feels about right, but its a false fit. Maybe I'm off base but this is how I perceive it with facts so far. Adding shims as a test to see if this eliminates the issue is exactly what I would recommend. Who knows, maybe your new gib will be the same height & the issue is how much relief they milled in the dovetails? Start with the easy stuff first.

 

[Eyecon]

The lifting is happening because the gib in not tall enough. It has nothing to do with its length.

Didn’t mean to imply short in the taper direction, I know it wasn’t too short since the gin was making contact with the 2 screws. I meant short along the dovetail as you’ve shown the drawing.

Hopefully the blueing will confirm that the gib is just not tall enough as you suggested and shimming will resolve the issue.

 

[DPittman]

One other problem I've noticed on my mill with the same gib arrangement is where the axis lock bolt pushes in on the gib. My setup doesn't have a brass pad in between the bolt and the gib and I think it should. The locking bolt makes an indent on the back side of the gib that can prevent the end adjusting screws to properly locate the gib.

 

[PeterT]

@DPittman that is yet another issue. I would assume when one sets up the gib that the lock screws are completely backed out so they don't interfere with the operation. Otherwise, ya, it could behave as a potential mid-span pivot point or be another mode by which the gib is displaced. His excess top gap serves no useful purpose but may well be complicating the gib fit, & would be the first area I would address.

 

[Eyecon]

@Eyecon the @RobinHood sketch is exactly what I was driving at with my questions. I don't quite understand your CAD diagram & it might be leading you to a false conclusion, especially with the 90-deg simplification vs. what should be a parallelogram. An unresolved issue is that you have a large gap between the top of gib & the underside of slide. ie. they are different heights & therefore allowing degrees of freedom that should be very minimal IMO. We also still don't know if the gib height is the same height on both ends, you didn't answer my question so it makes it hard to fully understand the situation.

using this sketch green is the big end & blue is the small end. Typically they would be the same height but should extend to where he has the shim pack. If the gib is adjusted axially in & out of the paper, that translates into displacement in the X-direction (blue arrow) tightening or loosening, exactly what we want
View attachment 33655

what looks to be going on is the gib is displacing along the dovetail axis (red line) as a function of the adjusting screw rotation. Now it gets interesting. I think theoretically if both big & small gib ends (green & blue) rose exactly together there should be no net displacement in X. But if one end acts as kind of a pivot point staying low & the other end is being displaced up via the rotation direction of adjuster screws (further complicated by the adjuster actually pushing on the gib), then I suspect you will have some have some kind of compound displacement where its sticking in one area & not in another. Detecting this with blue rubbing off is meaningless because this displacement mode is not accomplishing the goal of adjusting the gib in the intended way as above. You have just landed at some in-between point where it feels about right, but its a false fit. Maybe I'm off base but this is how I perceive it with facts so far. Adding shims as a test to see if this eliminates the issue is exactly what I would recommend. Who knows, maybe your new gib will be the same height & the issue is how much relief they milled in the dovetails? Start with the easy stuff first.
View attachment 33654

Totally understood and I appreciate the additional explanation. When the gib lifts in the green/front/wide end, it’s hard to tell what’s happening around the blue/back/narrow end. It seems to move but not by much. So as you said: it could be pivoting somewhere near the back or center, the gib certainly feels more wedged towards the back/narrow end than the front. Shimming will certainly stop that but will also mean that the gib is not making full even contact along it’s length which explains why the gib felt tight and l the wheels felt heavy while the sadle/table to base was still very sloppy.

Your suggestion makes perfect sense though, I can shim the gib as suggested, if the shim doesn’t correct the compound displacement that could be causing my problem(stiff motion and sloppy table/saddle), it would mean the gib’s fit is off.

 

[Eyecon]

We also still don't know if the gib height is the same height on both ends, you didn't answer my question so it makes it hard to fully understand the situation.

Yes I realized that I didn’t answer many questions. I wanted to take time to compare both the x and y gibs and take measurements and at least blur the gibs against a surface plate to make sure they are flat and report back to the group.

 

[PeterT]

Shimming will certainly stop that but will also mean that the gib is not making full even contact along it’s length

As long as you have some minimal clearance between the gib height & gap height, the gib should be able to bottom out the slide as its designed - by moving in & out axially the gib wedges in the horizontal axis to the extent the table becomes very difficult to move. I don't mean try & lock it up to test the full extent, but in normal setup this is how we know we went too far, the table becomes sticky.

As soon as you add (unwanted) rotational/pivot component to the gib strib, your gib/dovetail surfaces are no longer co-planar & now will interfere in a non-uniform likely somewhat random patch depending on the pivot point. I bet if you did this in a cad model with greatly exaggerated (top view) offset taper angle it would highlight this irregular interference. So if you blued that patch & removed it by scraping, the gib would only (kinda) fit if it was returned to the prior cocked up position. So on that basis is why I don't really see the point?

 

[Eyecon]

I don't mean try & lock it up to test the full extent, but in normal setup this is how we know we went too far, the table becomes sticky.

But that's exactly the problem: the table on Y is almost sticky yet it's loose side to side i.e. not rigid enough. What I'm trying to find out is why this is the case: is it because the gib height is too short and is not seating correctly or seating at a non-coplanar angle or wether the gib geometry in general is off (i.e. taper angle is off). . I'm going to try shimming first to see if I can find a gib position where the table is not sticky on Y but at least as rigid as it is on X, if I can't then this would mean the gib geometry has to be off. Does this make sense?

if you blued that patch & removed it by scraping, the gib would only (kinda) fit if it was returned to the prior cocked up position. So on that basis is why I don't really see the point?

So in this case the only way to get a proper read from blueing would be to make sure that the height clearance is reasonable first or at the very least always consistent(ii.e. shim the gib so it can't lift too much or even at all) otherwise the blueing results will be meaningless, is this what you mean? And again the only reason I was asking about checking for contact via blueing is if the shimming idea doesn't work because the gibs actual geometry is off.