Ball Bar Evaluation

[dtmtim]

I am setting up an internal standard for regular verification of our CMM using a ball and 20 positions per ANSI B89.4.1-1997. I have ordered the spec but haven't received it yet. We also have B&S come in and do a full calibration every 2 years.

How often do you evaluate your cmm using a ball bar?

I would also like some advise on writing a PC-DMIS program to report the distance between the spheres in all 20 positions. It seems like an abnormal program since it will require 3 manual hits on each sphere in each position and I'm not really sure how to approach this in PCD.

Thanks!

t.

[cmmguy]

Read this thread first, it may answer a lot of your questions.

http://pcdmisforum.com/showthread.php?t=1636&highlight=Verification

[dtmtim]

Read this thread first, it may answer a lot of your questions.

http://pcdmisforum.com/showthread.ph...t=Verification

Thanks, I need to get into the habit of doing a search before asking a question. Sorry. There doesn't appear to be anyone doing a full 20 position verification per the ANSI spec though. Any advise on writing a program where I can position the ball bar, take 3 hits manually on each of 2 sphere's, DCC Inspect the distance between the spheres, the reposition the ball bar and repeat. I was thinking doing a loop to run 20 times then grabbing the results each time with Excel. I can then store the excel spreadsheet with the date the verification was run for my records.

t.

[cmmguy]

Unless you do bar bar tests on a regular basis, I think that you will frustrated by the results. The barbar test is done at the periodic calibration of your CMM. You would want to do some sort of golden part checks on a regular(weekly/Daily?) basis instead. That is why I directed you to this thread.

[Bill **********]

If you had B&S do your ball bar then search for a B89 directory. Their 32 position ball bar program should be in there. It takes a single hit on each sphere measures three times and gives you a standard deviation between the three. You can modify it to do what you would like. If it isn't there, let me know and I'll send you the one that I've modified for my own use.

Bill

[T Miller]

Mine was in my pcdmis folder under certification

[dtmtim]

If you had B&S do your ball bar then search for a B89 directory. Their 32 position ball bar program should be in there. It takes a single hit on each sphere measures three times and gives you a standard deviation between the three. You can modify it to do what you would like. If it isn't there, let me know and I'll send you the one that I've modified for my own use.

Bill

Bill,

My file search did not return any results for B89. Would you mind sending me a copy of yours?

Thanks!

t.

[dtmtim]

Unless you do bar bar tests on a regular basis, I think that you will frustrated by the results. The barbar test is done at the periodic calibration of your CMM. You would want to do some sort of golden part checks on a regular(weekly/Daily?) basis instead. That is why I directed you to this thread.

My intention was to do a monthly check. Why do you think the results will be frustrating? I'm just curious - I haven't done this before and I know you have a lot of experience with this.

t.

[Jan d.]

I am setting up an internal standard for regular verification of our CMM using a ball and 20 positions per ANSI B89.4.1-1997. I have ordered the spec but haven't received it yet. We also have B&S come in and do a full calibration every 2 years.

How often do you evaluate your cmm using a ball bar?

I would also like some advise on writing a PC-DMIS program to report the distance between the spheres in all 20 positions. It seems like an abnormal program since it will require 3 manual hits on each sphere in each position and I'm not really sure how to approach this in PCD.

Thanks!

t.

Ever thought about running the Renishaw machine checking gauge? It is something that you can run daily and I have found that it will flag problems before they get out of hand.

http://www.renishaw.com/client/product/UKEnglish/PGP-66.shtml

By the way, we do our calibration every 6 months. We start with once every 3 months and see how much the machine "degradates". We found that on our machines, it is required to fine-tune them once every 6 months.

Jan.

[dtmtim]

Ever thought about running the Renishaw machine checking gauge? It is something that you can run daily and I have found that it will flag problems before they get out of hand.

http://www.renishaw.com/client/product/UKEnglish/PGP-66.shtml

By the way, we do our calibration every 6 months. We start with once every 3 months and see how much the machine "degradates". We found that on our machines, it is required to fine-tune them once every 6 months.

Jan.

You are doing your own calibrations then instead of having B&S or another outside source do your calibrations?

[dtmtim]

btw - my machine probably runs about 20 hours per week. I am the only user.

[cmmguy]

..... We found that on our machines, it is required to fine-tune them once every 6 months.

Jan.

Are they out of spec?

[Jan d.]

You are doing your own calibrations then instead of having B&S or another outside source do your calibrations?

For the real work we do, we only have top of the line Zeiss Carat Special Accuracy machines. We want them to be accurate to within 0.000040", which is the spec. We do not do that ourselves because we do not have the proper artifacts. We use an outside service (not even Zeiss). I do NOT think it is impossible to do it yourself. However, it may be more cost effective to have someone else do it. How many CMM's do you have? We only have 3 so it was not cost effective to get all hardware, train several people to just do this 3 times every 6 months.

We have some other low accuracy LK's that are done by LK (same reason as Zeiss's, we only have 3 LK's).

We calibrate our machine tools ourselves. We run laser and ball bar tests. Then create comp tables in the CNC controls. We decided to get into this because we have about 30 highly accurate 5 axis machining centers that need to remain very accurate. We have a lot of experience with in-house machine tool calibration.



Jan.

[Jan d.]

Are they out of spec?

Yes, they are. We seem to be loosing about 0.0001" per half year. That is unacceptable in our case and the machines must be brought back in. It rarely happens that after 6 months our CMM's still meet spec.

If they were to meet spec at every calibration interval, we can do 2 things:
1: either relax the calibration requirements (go to once a year or longer).
2: change the accuracy spec for the machine.

We tend to do number 2. If the machine is so stable that it always meets spec, we clamp down on the spec.


Jan.

[dtmtim]

[QUOTE=Jan d.;49870]For the real work we do, we only have top of the line Zeiss Carat Special Accuracy machines. We want them to be accurate to within 0.000040", which is the spec. We do not do that ourselves because we do not have the proper artifacts. We use an outside service (not even Zeiss). I do NOT think it is impossible to do it yourself. However, it may be more cost effective to have someone else do it. How many CMM's do you have? We only have 3 so it was not cost effective to get all hardware, train several people to just do this 3 times every 6 months.

QUOTE]

I really have no intention of doing a calibration myself. The intent was to use the ball bar program to establish a base line to periodically check the machine in case something goes out of spec. The renishaw solution looks good but is probably beyond my budget - as a small shop owner my budget consists of what I have in my wallet :-).

Thanks!

t.