Problems milling a "perfect" straight line on a VTL with live tooling 1

[DrMetal]

We have a 1991 Berthiez (french) Vertical Turning Lathe (model TVM-100). The vertical ram (Z axis) has a live tooling attachment for milling & drilling. In turning / boring mode, the table can reach 600 RPM. In milling mode, a precision, low speed gearbox servo engages the table for a max of 5 RPM. We are trying to mill a rectangle shape into a surface. There is no Y axis ... only X, Z, and B (table rotation axis). Therefore, to mill a straight line, the B and X axes must be synchronized within the FANUC series 15 parameters to yield a perfect point-to-point straight line. However, no matter what line length or feed rate we program, we are consistently getting what appears to be a slight angle at the center distance of each straight line cut. However, it is possible the "out-of-straight" line is really a very shallow curve. For example, in a 16" travel cut, we are getting about 0.005" "out-of straightness" on what should be a perfectly straight edge. As the low speed table rotates clockwise, the part initially rotates "away" from the tool, and then rotates "into" the tool after passing the midpoint of each straight mill cut. It appears the "bends" in the straight milled edges are at the center distance where the X axis starts to reverse direction to maintain the straight cut. In other words, it appears there is something not working correctly with the synchronization of the table rotation (B axis) in relation to the X axis movement. Hope I explained this correctly. Anyone have any ideas where to start, or what to look for? It looks as if this problem may be a common one for any machine that uses this type of syncro between axes to mill straight cuts ... or any other specified shapes as well. One person has raised the question about whether backlash in either axis could be the culprit. If so, how does the Fanuc program compensate for the backlash (assuming it exists)? We have measured near zero backlash in the table, but we have not measured the X axis. However, we have recently completed many precision acme threaded, mating parts using the high speed turning / boring mode of the table, where it must syncronize with the X andf Z axis ... all with no problems. So we assume there is no backlash problem. Thanks in advance for any help on this issue. BTW, we have tried to get technical help from the factory on past issues, but with very little success. Our 1st attempt using the GE Fanuc technician has not worked. GE has another tech coming soon, but thought we may get some help from this forum too. Thanks again.

 

[jbel]

First things first, there is no such thing as

a perfectly straight edge.

You may be able to approach straight but you will never get there.

It seems from your posting that the axis syncronization is working correctly.

Check your setup and programming parameters, live tooling attachments at notorious for not being as rigid as main spindles on CNC's. Are you removing all the material in one pass? Is this too much to ask from the live tooling head in terms of rididity?

You say that

It appears the "bends" in the straight milled edges are at the center distance where the X axis starts to reverse direction to maintain the straight cut.

It appears that there may be backlash in one of your axis drives. When the direction reversal occurs, this is where you are having your problem.

If this is the case, try taking multiple lighter cuts and see if the problem diminishes. You may need to either baby the machine with light cuts in the future or look into having the axis rebuilt. We are not exactly talking about a new machine here, so this may be the correct way to fix the problem.

Good Luck!!

 

[DrMetal]

Jbel,
Thanks for your insights. I realize there is no such thing as perfection (unless you talk to my girlfriend). We are getting a much greater "out-of-straight" condition than what the machine is capable of achieving. I don't believe it is a tool rigidity problem due to the way the live tooling head is designed & attached to the ram. This is the first attempt at using the live milling function since the machine was installed in 2003 (it was purchased used at an auction). The original factory parameters are reported to be still in the software. One thing we did during installation was to rebuild the low speed gear box. In doing so, we reduced the backlash in the main table spindle from about 0.010 to 0.015" to less than 0.005". This was measured as indicator movement at the OD of the 1 meter diameter table while prying back & forth on the table with a 16" long prybar. So what we are wondering is whether there is a "compensation" parameter in the FANUC software that is set for the original 0.010-0.015" backlash. Maybe we are now over-compensating with the reduced backlash. In this particular operation, the cutting pressure on the X-axis (i.e. the cutting tool) is always "pushing" in the same direction, even when the X-axis reverses direction, so we don't think there is an issue with backlash in the X axis. However, the cutting pressure on the B axis (i.e. the table spindle) is in the clockwise direction for the 1st half of each striaght line cut, then reverses to counter clockwise pressure past the mid-point. So we may have an issue with backlash compensation on the turning table. We will investigate that possibility next, as soon as Fanuc tech rep arrives later this week. This run of parts is made from 3"T 316ss plate cut outs. We are taking lots of relatively light cuts. I don't believe we are over stressing the ram or live tool holder. The live spindle drive motor is only about 9-10hp, with only CAT-40 tooling. The main spindle motor is 60-70hp, so the X & Z axes are very rigid for the heavy duty turning & boring functions. In any event, I will pass your suggestions on to the operator to see if that makes any difference regarding taking lighter cuts. Thanks again.

 

[Colega]

I have never seen this exact same problem before but have you tried to compensate the "out of strighness" by modifying the program. Since you dont have a Y axis youll have to make the table spindle go back and forth a little in the are where you have the problem. The controller should have a backlash parameter but I dont think the problem you have is there.

Good luck.

Colega

http://www.corporacioneg.com

 

[BillPSU]

I've never programmed this type of cut. Is the cut broken up into (2) moves? First cut from the outside corner to the centerline decreasing on the X-axis and the second cut from the centerline to the other corner increasing on the X-axis. If this cut is done in one move it would generate a tapered slot unless the controller has some special subroutines programmed in.

 

[DrMetal]

Gentlemen - Thanks again for your input on this problem. I am not the programmer and have yet to speak to him about this problem due to some other more important issues. Once we get together I will post our findings. I do not believe "programming" the problem away is the answer in this case, since each different "straight" cut (or any other cut that requires going past the "fulcrum" or centerline of the B spindle axis) might possibly require a different program adjustment, but I am unsure of this. I do not believe the cut is programmed in 2 moves with this machine. And because there is no Y axis, all extensive milling cuts require an accurate sync of the B & X axes. It appears the milling can get quite complex as far as the internal logic & programming goes. We'll see what we get once we get together & attempt some of your collective suggestions. Thanks again.

 

[CSLufkin]

Here is my guess.
Somewhere in the background of your control there is a type of logarythm that calculates how to "synchronize" the X and rotational axis. This Log uses a constant number in the form of a parameter setting somewhere (probably well hidden, and for good reason). If you try this same type of cut at different distances from center you will most likely find a pattern to how far "out of straight" the line is milled. If you have the machine lazer calibrated by a reputable company, they can probably fix this. Unless it is a mechanical problem (backlash as suggested earlier, or a guideway or encoder problem). This is the gamble, you have to pay them even if they find a mechanical problem.

 

[Colega]

I Agree with CSLufkin in that there should be a parameter on how to sync the x and the b axis. Because if we go to the geometry of the machine, there is no way you can do a complete straight line by using this machine. Your straight line should really be the result of a mathematical calculation, much like the way a machining center can mill a circular hole.

Maybe you just can't do that piece on that machine, it may be out of capability. Sad, but may be true.


Colega

http://www.corporacioneg.com

 

[rupp]

Could this problem, be due to the direction in which the cutter travels, along the part? If I am understanding you correctly your a milling some sort of rectangle pocket out. The .005 run out could be due to the fact that for half of the "straight line cut" your are conventional milling and the other half you are climb milling. Which inturn means for half the cut, the tool itself wants to push away from the part and the other half the opposite is true. This would probably account for some mismatch on a lengthy cut. I would solve this by leaving a light cut for a finish pass along the walls of the pocket such as .01".