Thermwood Alignment Accuracy

[xform precision creations]

I realise that this is not a Thermwood forum however I know there are a few Thermwood & Mastercam users on this forum who may be able to help me out. I have been having some issues with head alignment on my 67 5 axis machine and have just finished traming in the machine (head allignment for the B & C axis). All my tolerances are less than 10% of the allowable as specified by Themwood by yet I still seem to have a significant error somewhere. Note that all these tests have been double and sometimes tripple checked and I am confident that it is very well aligned. Hopefully the following explains the problems I am having.

 

The error showing up has something to do with the B axis and shows up when the C axis does a reversal (ie when it reaches its 360degree limit, retracts, flips B and C 180deg and enters in to the part again). The error showing up is about a 0.4mm relative to the z axis height. I can simulate this in testing using a test pin in the collet and manually driving Axis B90-, bring z down to a DTI gauge and take reading, lift Z to clear DTI, rotate axis C 180+ and axis B 180+, reposition X to position the pin over the DTI and lower Z to the original Z setting. The difference in readings is about 0.4mm, a considerable error. But yet when checking axis B home position it is within 0.01mm and axis B width centred on axis B is within 0.01mm and any backlash in the B drive system is so minimal that it is unmeasurable. I am wondering if it has something to do with software, ie I am telling it to go B90 but it is not actually going 90? All my code comes from Mastercam but the errors are completly to do with the machine, not Mastercam.

 

I am wondering firstly if anyone can give me some hints as to what I the problem might be and secondly what sort of accuracy are other people getting on the Thermwood 5 axis machines, particually when if does a reversal?

 

I will discuss this with Thermwood tommorrow by if anyone can help me out prior it would be most appreachiated. Thank you.

[crazy^millman]

I would wonder how much wear you have on your worm gears on that axis? I would also wonder what type of run in program you use everyday to make sure the worm gear keep even and normal wear. These machines have a high wear on the worm is used only in one area alot like 90 so it is important to use a program to run it through it motion and check your back lash as well as the amperage spec for those drives all the way through the axis movement ot make sure you are not tight in some places and losse in others. I would also double check your poivot point and make sure it did not change by a drastic amount. I would also check the sqaureness of the machine I have seen this move and cause all types of headaches getting it to repeat.

 

HTH

[jmjftw]

We had two Thermwood 5 axis routers that we used to machine the superlite ablator off of the component parts used on the Space Shuttles external tank. We had accuracy/repeatablity problems with our machines as well. The machines were delivered with resolvers and we had the retrofited with encoders, it still didn't help.

They moved them to the SOFI area (spray on foam insulation) because the tolerance(s) on the foam was plus or minus .250 in most cases.

Now if you want to make a nice wooden door...

[Guest CNC Apps Guy 1]

The Thermwwods I've seen in "action" have had eccentric spindles (read when the spindle in parallel to Z and you rotate C there is eccentricity). About .030" has been the average I've seen. I'd look at that FIRST because that will give you BIG headaches.

 

JM2C

[Slick]

Excuse me if this does not relate, but, is this showing the air in posted NC code only? The reason I ask, if you are doing all the tram checks successfully (manually) and then coming up with an error in your programmed files, it could be that you don't have the correct pivot distance. Maybe?

[xform precision creations]

Thankyou everyone for your replys. Further investigation has revealed that the B axis gears may have some wear, and while I thought I had no backlash, there does seem to be some in certain spots of rotation. At 90 degrees there is virtually no lash but at 0 degrees there is approx 0.15mm at a point approx 230 from the pivot. This could be amplified during a reversal to the 0.3 - 0.4 reading that I am getting, combined with the fact that setting axis 5 home could have been done with the lash sitting in the opposite direction. Therfore I will look at getting some new gears or maybe temporaraly rotating them to a less worn spot if this is possible. Hopfully it isn't too bigger task to install new ones!

 

James, regarding concentricity I had this adjusted to within 0.02mm but I am sure now that if I tested it again with the backlash in a different position it may be much larger.

Pivot distance is correct as I have measured it and run a test piece to test the various axis accuracy and the only error showing up is B axis roation being off. Also verified that the Gcode is being posted correctly and it is.

 

Ron, Do you run your machine in a warm up mode at the start of every day to excercise the B and C axis, regardless of whether you are using it for 3 or 5 axis work. What sort of time should I be warming them up for, or are they still going to wear due to it deing vertical more often thatn any other position? I presume you would also use this time to warm up the spindle at a low rpm before bringing up to 18000rpm or is this not nessasary.

 

Regards,

Howard

[Mike Gates from Barefoot CNC]

Just curious, did the Thermwood Tech. tell you about the worn gears or did he actually show someone at the shop what he found. The 5 axis Thermwoods I dealt with were all belt drive on the B axis. We had a lot of trouble with the belts stretching over time causing it to be very inaccurate, especially when doing any heavy machining, it finally slipped a notch on the belt. We replaced the belt after we realized what was happening and things did drastically improve.

[MATT Q]

The newer machines have a better belt setup they have bands of metal (radials )in them they are alot stouter than the old ones. Howard that gear change out is not fun we will be doing that this week also on one of the machines we cut fiberglass on. Isn't there a tram adjustment for the error you are seeing? You are running a toolchange machine right?

[xform precision creations]

Mike, I was able to pick up the gears were worn myself by feeling the difference in backlash at different positions of B axis. I then discussed it with Thermwood who verified that this probably was the problem. I have since priced up some new gears. I think that I will get some of the new drive belts. As Matt said they apparently have steel bands in them making a more positive drive and takes the spongyness out of the drive. Matt, it is not a tool change machine but a dual end spindle. I will be able to tram it correctly in an area of no lash now that I know what the problem is. I have also learnt that I perhaps need to warm the gears up for longer each morning to prevent this happening again.

[crazy^millman]

Sorry to be getting back to this but crazy around these parts. I would run it for 1 hour every night and 4 hours on the weekend. I would run it a slow speed as well the help it wear even. The gear is a PIA but you can do it yourself with their tool they need ot supply to you.

 

HTH

[Jimmy Wakeford from Barefoot CNC]

I usually put a G92 B0 in the top of prog for adjusting the B per day if cutting thicker materials on therm.

To check what the daily B should be z down to bed and turn c 180. Then tilt b till it is straight , and that is your b value for the g92

 

I think it is 1 degree for every .125"