How often should I Qualify Styli?

[[Co...]]

I work at a production shop and a lot of our jobs we run every part through a CMM program as they come off the CNC machine.  How often should I be running my qualification program?  Normally I would only do so if a probe breaks or if I set up a new probe.  I'm wondering if I should be qualifying the styli every week now that we are using the CMM all day during production.  

Thank you!

[[Ja...]]

Probably lots of trains of thought on this and while this answer is unhelpful it would be - it depends.

How clean is your CMM environment, is junk prone to accumulated on your probes? What kind of material are the parts you are running? If it is a rough surface and you are using scan paths then you may wear down the probe tips without realizing and not catch it until your next qualification. Are you the only one operating the CMM or have full confidence in other operators not to have a crash? How would you know if something wasn't out of whack without qualification at regular intervals?

That said for us I run a basic qualification program at the beginning of each shift on our CMMs that covers our main couple of stylus systems that cover about 90% of the parts we inspect. The program takes about five minutes to run and gives me some peace of mind that something didn't get screwy if I wasn't around to babysit the CMM or operators running on it. I do a full requalification if a probe is being replaced or setting up a new one, otherwise that daily qualification is sufficient for our needs. We could probably get away with a weekly qualification but it's quick enough to just get it in every morning.

Edited 19 hours ago

[[Ch...]]

So once a week would be recommended. We used to do it every morning, but then production complained that it was blocking the CMM for too long. Since then, we only do it once a week.  Or in case of a major collision.

[[Da...]]

When I wanted an answer to the exact same question, I chose to try a few things out. Instead of calibrating after a certain time, I regularly checked my calibrations for two things: 1. Does the probe still measures the correct diameter of the ref sphere and 2. If I use multiple probes in one program, how far do the X/Y/Z coordinates of the ref sphere spread.

I found that a probe holds its diameter calibration value quite stably easily over at least two weeks, but multiple probes tend to drift away from each other after some time.

So, what I do is, when I use only a single probe in a program, I check daily if it still holds its diameter value. No calibration needed. But if I use multiple probes in a program, I check them daily to each other and if they aren't within my spec (I want them to hold their position to each other within 5 microns), I calibrate them all in one go (but just the ones that I need). I'm pretty happy with that. 

[[Cl...]]

Try adding a daily master part verification. This will immediately alert you if there is any problem with the styli.

Or instead of a master part, create an artifact with multiple geometries. Use it to measure size/form/position/orientation

of those geometries using every styli rotation/angle. This may take 10-15 minutes, but is well worth the time.

[[Ow...]]

As defined above, every work scenario is different.

We have a lot of Duramax CMM's that I set the global stylus limit values to require a qualification of all probes every 24 hours, with a sigma value of 0.002mm max.

The protocol requires that all styli be cleaned before running the qualification program.

I do this because on some of them, several different operators use the CMM (set to run in the autorun interface) on different shifts and there is no way to absolutely ensure every part is clean before every part is run and dirt/debri/sticky-stuff can accumulate on the probe that can not be seen with the naked eye and will affect results.

I have a daily (every 24 hours) qualification program set to qualify as passive (doesn't take long) and a once a week (generally Monday) that qualifies with the geometry qualification, which takes about 25 minutes.

In the limit values, you can set how many hours elapse before the CMM won't run without the styli being qualified. 

[[Ke...]]

I will add another point of view, which is that the time between calibrations/probe checks determines how many parts need to be quarantined and sorted if an issue is found. Calibrate Monday, run all week, Calibrate the next Monday and find an issue? 1 week of production to quarantine and sort. Calibrate Monday, run all day, Calibrate Tuesday and find an issue? 1 day of production to quarantine and sort. So calibration frequency may be less about the CMM losing accuracy over a certain period, and more amount what is an acceptable amount of risk for the company