I think we finally pinned it hopefully tomorrow will let me sleep... Tapping Problems

[huskermcdoogle]

Personally I would rather use IPR

For tapping only.

 

[huskermcdoogle]

So your Machining center can program the Z axis in Inches / Spindle Rev?

Good 'ol G95 mode on a machining center. I have done it exclusively in the past on mazaks for tapping, some people swear by it, I used to, now I go with that makes people comfortable.

Can you program a X axis to do the same?... say... 0.500 / spindle rev ?

The value per rev is limited to max feed capabilities. But yes you can program that way if you prefer. I think it is parameter enabled.

If so, when did that start?... when Rigid tapping was introduced??

No Idea

I'm guessing that the reduced spindle speed for this tapping operation is helping you more than what ever else you did.

It has definitely helped tread quality, but I don't think it has much to do with the issues we were having with breaking taps, as the 8-32 was ran at more conservative speeds, and in a tapping chuck.

Pressing on.

 

[Dave Cross]

Pressing on.

Dude, no need. Try my procedure in post 11, you will find your problem, 20 minutes is all you need.

 

[jdj]

Dude, no need. Try my procedure in post 11, you will find your problem, 20 minutes is all you need.

Gotta love it when people don't listen to good advice.

Jeff

 

[huskermcdoogle]

Gotta love it when people don't listen to good advice.

Jeff

It't not that simple. Machine has scales. I do not have the procedure for testing backlash with scales, and to my knowledge you can't just turn them off.... I do have the documentation for turning off scales, and it is much more than a ten minute job, and involves re-establishing the axis home positions. Not exactly fun in the Z...

Now, I have put an indicator on the z-axis and done some fed and rapid (programmed) moves into location and checked position. Money on every time. Same if you do the old hand-wheel method. As for using a tension compression holder with the coolant off and using oil, I don't think I would be able to see anything here as the amount of error I believe I have is basically non existent, as I am currently using a solid holder on my M4's and producing great threads. I could try to mount an indicator on the spindle housing and use it on the end of the t/c tap adaptor and watch its motion during reversal. That might tell me something, but I will need a moment to do that, I can probably find a moment to do that tomorrow. In fact maybe I will take a video and post it.

In the mean time, I have other reasons to not be convinced that there isn't a mechanical problem with Z. Before I left today, I was getting a probe open error. The probe was not able to rearm on retract. I do not know if this was an isolated issue, but it would do it with this particular point over and over again. My solution was to increase the back off distance for the second hit. This tells me there is slop in the system. The issue here is the scales shouldn't let this happen from what I understand.

I have and will continue to listen to anything anyone has to say, my issue is that I need to make parts really bad, and currently we are limping at an almost normal running pace. So, I can't start chasing things until I have all of the things or information lined up to go after it. We need parts super bad, and will get to the bottom of things soon. One thing on the list of need to do it now is a Y-Axis screw, from what we can tell it or something in that assembly is boned, and will likely come apart soon.

If anyone recalls this machine had some lubrication issues. The grease pump was turned off in the K parameters by a previous owner, and the lack of grease consumption didn't cross our lube guys mind. I had stared at the grease tube and wondered if it was using grease and continued to ask about, but neither the boss man or our part time maintenance guy seemed to care until we started to generate noise in the Y-axis. At this point it is obviously too late, my concern lays in what damage has been done to other systems on the grease pump, namely the x and z axes, but for now they don't make any noise whatsoever other than the standard ball screw and way cover noises. Now I won't write off the fact that these might have damage, but there are tests that we can have done such as laser displacement, and ball bar tests to check the condition of these parts, and this is becoming a pretty high priority for us to get done. In fact it will probably happen sooner than later, hopefully in a week or so. IMHO, spectacular failure is not a surprise for me at this point because preventative maintenance has not been a strong suit of our company. We don't even have full time maintenance. Any maintenance items that get done are purely reactive, even on simple things like changing compressor filters. It makes me sick sometimes, the things that get pushed to the wayside, and the amount of things that break as a result of not taking care of them when the first sign of trouble occurs. It usually doesn't get fixed until it can't be used. Such is the case with this machine, right now it is barely usable, in fact I am not comfortable making parts with it, for reasons that we don't know for certain what has been causing us our tapping troubles. But the fact that it will make parts and due to all of the other things we have going on, the machine will be making parts until we notice a show stopping quality issue. Something that I personally don't want to run into, though am not particularly worried about, because we would have to have some serious problems (like breaking taps) for us to stop. These parts generally have a pretty open tolerance.

I don't know. At this point i am just more frustrated than anything. I like things to be working right, and have been pushing for us to stay on top of things, but I can only do so much and have so so so much on my plate. Last thing I need is to have to spend my day at the machine doing diagnostics, but if that is what I end up doing, than that is what I end up doing. All this does is push the dates out further for the next parts that need to be done on the machine, and push me deeper and deeper into the hole that is in the middle of the pile of work I have sitting around my desk in my office.

Bed time.

Husker

 

[jdj]

It't not that simple. Machine has scales. I do not have the procedure for testing backlash with scales, and to my knowledge you can't just turn them off.... I do have the documentation for turning off scales, and it is much more than a ten minute job, and involves re-establishing the axis home positions. Not exactly fun in the Z...

I don't know. At this point i am just more frustrated than anything. I like things to be working right, and have been pushing for us to stay on top of things, but I can only do so much and have so so so much on my plate. Last thing I need is to have to spend my day at the machine doing diagnostics, but if that is what I end up doing, than that is what I end up doing. All this does is push the dates out further for the next parts that need to be done on the machine, and push me deeper and deeper into the hole that is in the middle of the pile of work I have sitting around my desk in my office.

Bed time.

Husker

I sure hear ya' on all this! It is rarely simple AND the last part about getting buried in work! What makes it more aggravating is that I seem to be the only one in the company that is concerned! Everyone else acts like it is play-time! Mgmt especially! WTF?!?!?

Jeff

 

[Bruce Griffing]

I am not saying this is the answer, but I will make a comment about the linear scales. What they guarantee depends on how the servo loop is adjusted. Any servo loop must allow some following error or it cannot function. The question is how much. For a given motion, the scales probably guarantee the correct endpoints (though that is probably also an adjustment), but there could be some error during the motion if the servo has drifted out of spec. If there is a bad spot in the ball screw the problem may only occur in that area. I would do a check of following error for the servo in question if it were my problem.

 

[Duak]

I help out with electronic problems at a small shop. One no-name VMC with a Mitsubishi control was having trouble working in certain speed ranges. It would go but wasn't quite right. The control thought it was having encoder problems. The servo eventually said no more and refused to work at any speed. It turned out to be a bad seal leading to a thrust bearing with non-spherical ball bearings. Strange things happen when the mechanicals are not right.

Best Wishes,

Duane

 

[huskermcdoogle]

Strange things happen when the mechanicals are not right.

Ain't that the truth.

 

[otter]

This tells me there is slop in the system. The issue here is the scales shouldn't let this happen from what I understand.

If the above bolded this refers to slop, then it is true and not true. The control system should use the linear scales to eliminate the slop. But the control system takes time to act, so whether you think it eliminates the slop depends on your time frame.

More specifically, cutters cut, even if they are just somewhere for a fraction of a moment. If you are getting slop related bad motion, even for just an instant, you will get strage finishes, divots, etc.

Is there a way to check for slop (in non-gravity axes, at least) by turning the machine off and pushing on the table using a crow bar or whatever works, with an indicator to look for some slop that is easier to cause than backdriving the screw/ motor? I understand that you don't have a ball bar or an alternative backlash checking procedure, but most opinions here lean to backlash. It seems like it would be worth trying to cobble something together to look into backlash.

Erik

 

[huskermcdoogle]

I never thought about trying it with the machine off, granted it wouldn't be as accurate due to potential minute screw back feeding, but it would definitely give us an idea of what we are looking at.

Husker

 

[scudzuki]

I never thought about trying it with the machine off, granted it wouldn't be as accurate due to potential minute screw back feeding, but it would definitely give us an idea of what we are looking at.

Husker

That is the first thing I do.
I don't know why the natural instinct is to complicate everything.
The bigger the machine, the more force it takes to overcome stick/slip and the mass of the axis, but it can be moved. Even the vertical axis can be moved with a bottle jack (carefully!) with a little creativity. Lock the ballscrews from turning and shake it back and forth, any cumulative play (screw + thrust bearings) can be measured with a dial indicator. Of course the contibution of the screw can change per axis position but usually the nut or balls wears more than the screw.

Joe

 

[huskermcdoogle]

Update:

I have pretty much come around full circle back to lubricity problems, granted I have not ran all of the parts I run on this machine that have had issues, but I have now successfully tapped about 350 M5 holes in a tension compression holder and about 50 M4 holes in a rigid ritmar extension between two parts that we had the most trouble with. In fact these parts have the most risk potential from a material and heat treat standpoint, as we are tapping on the border of a friction stir weld.

What we did was add 3% Trim E206 to our 9% Trim Microsol 585XT to give it added lubricity per recommendation from Master Chemical. We are going to be looking into a whole new coolant asap, but needed something to get us by so this seems to have worked so far.

Crossing Fingers.

Ran a 12" circle ballbar test. Machine is still in pretty good shape from a backlash and whatnot standpoint, though has a cyclic harmonic in Y. This takes this machines positional capability from .0003" to .0008" and circularity from .00018" to .00087" (yes five place). Still better than some other machines in the shop. Backlash in the X and Y are .00018" and .00024" respectively, though our calibration guy said the Y value is probably better, but the cyclic nature of the Y axis is probably throwing that out a bit given the the amplitude of that variation is near .00035". We didn't check XZ or YZ, as we ran out of time, but our guy basically said that with the scale, and the fact that we can tap a hole rigid, that the values in there are basically no different than what we had before and will be near perfect like the other axes.

As for the scale in Y. I pulled the scale, and with assistance from our calibration tech (very well trained machine calibration guru), carefully disassembled the reader head from the glass scale, and thoroughly cleaned the glass scale with warm water, dried it with alcohol, and reassembled and aligned the scale. Set it up and tried to re home the scale, but kept getting a serial data error, which was the error we had which causes us to shut it off. So we are going to get a new cable, and test if it is the cable or the hardware, or on the control side. So the scale is still off, but we will fix it (soon), overall the machine is still pretty decent.

Thanks for the help and suggestions so far. It is always good knowing you have some support behind you when nobody else you know can help.


Husker

 

[Ox]

between two parts that we had the most trouble with. In fact these parts have the most risk potential from a material and heat treat standpoint, as we are tapping on the border of a friction stir weld.

Need I say more?


So - are you running on the motor encoder for now then?
Or are you down....


-----------------

Think Snow Eh!
Ox

 

[huskermcdoogle]

Running on the motor encoder again. As we have been since July. Just need to get this thing fixed fixed sooner than later.