When do you replace probes?

[[St...]]

How do you know when it is time to replace a worn out probe? Have an operator that says he can see the flat spots with his naked eye and wants to change probes. So how long do Ruby and Silicon Nitride probes last? The sigma on the qualification is less than .0005 mm.

Probe Qualification.JPG

[[Da...]]

Ruby is an extremely hard material, only slightly less hard than diamond. In my opinion, you will never see a flat spot on a ruby. I've checked a lot of probe rubies under a 2000x Microscope and never have I seen a flat spot. In fact, I had been searching for a perfect way to clean aluminium build-up on ruby, and I've saved and improved a lot of ruby probes by using sandpaper. Yes, you've read right, sandpaper. The ruby doesn't even shed a smile when I torture it with sandpaper, and it will get the range (not the sigma, but the range, the sigma will even be better) back to values under a micron.

One thing that I did see on rubies is pitting. When I have a bad ruby, it's always pitting. I've had a probe ruby used on a Heidenhain probing system on a CNC mill that was totally unusable because it looked like huge chunks were broken out. My suspicion was the operator had dropped it on a hard surface, no other way it could have happened.

Virtually all rubies I've checked have small pits that are small enough to not affect everyday measurement. Even brand new rubies have those. Every small pit is a potential nucleus for a material break-out, but if you treat your probes right, they can last for years.

My worst calibration results happened on new probes and could be explained by faulty glueing, thus resulting in almost unnoticable movement at the connection to the stem. My small vendor replaces those without lifting their eyebrow.

If you do heavy work on abrasive materials all around the year, even then will you most likely only enlarge the pits that had been already there but not create flat spots. Try it, take an old probe and try to destroy it with sandpaper, and check before and after. Yes, there may be tiny scratches visible under a microscope, but their width is negligible.

If you are unsure, just measure your reference sphere with a scan around the equator and some scan lines on the rest of the surface. The most stressed part of a ruby probe usually is the equator. Any defect will reflect right away in the scan.

As for silicon nitride, I use some probes, but I have too little first-hand information about wear. I will report - in a few years? 😃

[[Ri...]]

Please sign in to view this quote.

I have been using electric erasers for about 20 years, I use Mylar eraser inserts, they do no harm to ruby or Silicon Nitride.
Typically, I can clean rubies from 2mm on up and disc styli using a 10x scope or eye loupe.
This can clean off aluminum, magnesium, titanium, etc. build up within about 10 minutes per stylus.

[[Ke...]]

It depends on what material and what the condition of the surface you are measuring as to how long probes last. It is definitely possible to flat a Ruby, Silicon nitride, or even Diamond-coated probe (sort of. Usually the diamond coating breaks away and the underlying material gets flatted in that case). It is also possible to have buildup occur on the probe in such a way that it appears to be a flat. I personally have customers that have run the same ruby probes for years, usually not having to replace them until an operator incident breaks them, and I have other customers who have to replace the probe every 1.5-2 months due to flatting because all of their measurement is done with the end of the probe, on surfaces with machining marks, at very high measurement volume. One way to catch these is to create a program which scans the reference sphere as a circle, or semi-circle depending on access, in multiple axes, then display and plot roundness with some level of magnification (I believe I used 500x for the customer in question). This will allow you to visually identify using the plot if there is either build-up or flatting.

[[Ri...]]

Did I just read that you will never see a flat spot on a ruby??? Lol.

Someone hasn't scanned a lot of planes on semi-rough surfaces before then.

[[Da...]]

Please sign in to view this quote.

Pictures or it didn’t happen! 😉

[[Ri...]]

Please sign in to view this quote.

Lol.

I don't have them anymore, but it absolutely happens. Scanning planes is the worst wear on probes due to the localized surface contact.

[[Je...]]

Please sign in to view this quote.

Probes are relatively cheap. When I encounter any plausible existence of wear or failure, I replace the stylus. The time and risk in reworking a worn stylus is greater than the time and cost it takes to replace one. If an employee approaches me with doubt over a stylus tip's quality, I immediately replace the tip with no questions asked. I thank the operator for being proactive and communicating about the problem.

Stat. models and PM could be used to detect stylus failures, but these would only add value in large volume situations.

[[Da...]]

Flat spot on a 5mm ruby caused by scanning ceramic. I have an image from a stereo microscope somewhere also.. it can happen. 😃

Yellow line is ‘perfect’ geometry and you can see the gap to the tip.

IMG_3328.jpeg

[[Ow...]]

Been working with CMM's for 18 years and yes, you most certainly can use a probe enough to wear a flat spot on them (especially on cast iron parts working three shifts 5 days a week) and a standard tensor or just points probe calibration will not detect them, especially on the very bottom.
Silicon Nitride probes are best for aluminum, not so much for wear in my opinion but, because you can see the aluminum build up on them.

[[Ro...]]

If everything has been calibrated together, take a Z point on a very small surface of something, then change to a different probe and take a point in the same place. if both points are within a couple tenths of each other, then tell your guy to get back to work.

(do things like this while your operator is at lunch. if the measurement is off by a substantial amount, or off in the wrong direction, it'll just lead to more questions and stress in your life)

[[St...]]

I would like to thank everyone for their input on this subject. It is very much appreciated!! We do quite a bit of scanning on steel as well as aluminum parts. So I can understand that a flat spots can happen, and that the probes do wear out. But as I say we only check the "shiny" surfaces. So the surface is not too rough to really wear the probes out real fast.

[[Ch...]]

Use this program to find flat spots on rubies. This was originally from Derek Stedman.

Styli Roundness Scan.zip

[[Ro...]]

This is after scanning a part made of silicon carbide. It ate up the Ruby, so then we switched to Silicon Nitride which wasn't as bad as the Ruby but still wore it down. We are going to try using Zirconia but if that fails then Diamond it is.

[[Gr...]]

I also work with silicon carbide and see a similar thing on our ruby tipped probes. We usually swap out a probe when the stand deviation is over 1 micron, and it's almost always a flat spot on the bottom. Our 0.5mm probes last about a week before there's a flat spot on them. but the 3mm probes can go on for months or more.

And yeah, we could switch to diamond coated probes, but they're literally 10x the cost of a ruby tipped probe.

[[Da...]]

I agree with Owen, I've been programming CMM's for 46 years and yes i started before Computer Monitors Mice Hard Drive or RAM, we had a Teletype Machine with Paper Tape Punch, you can use a probe enough to wear a flat spot on them. when i went to another company after they lost a head programmer i ended up scrapping 6 of the stylus tips some you could see Flat spot on tip & Around the equator where they were scanning brass rings I.D. & O.D. - - - the machine shop complained the CMM's didn't render same results as Deltronic Pin Gauges i knew right off i needed to check the stylus tips and make sure the programs were using the proper evaluation settings . . . . 😮

[[Wo...]]

Please sign in to view this quote.

This is my approach as well - if in doubt - replace it. The risk is just not worth it.
I also managed to flat-spot a probe. Curiously, the sigma was perfect as the Tensor calibration is not hitting the top of the sphere, and that's where a flat spot was (about 0.002" deep)

[[Mi...]]

For regular steel parts our probes often last years and that's in a high volume environment. Generally they get replaced because they are broken, or glue has failed.

We do have a a nasty scanning job on a rough ceramic disc that eats up probes fast, like above we considered diamond probes, but the decision was made to get the shortest, cheapest ruby probe and just replace them every few weeks. The job only measures flatness, which stays accurate enough for it's purpose even with a noticeable flat spot.