I’m doing some testing at work and need to see what the Ra, Tp, Rz and Rp values are for a part. These parts need to be held to a certain value of each but the profilometers I’m using are giving me readings all over the map.
My theory, is that you can have a good surface finish and get a bad reading, but you can’t have a bad finish with a good reading. It will always either give you the actual surfaces finish or worst. I just don’t see how it can get a reading that’s better than it actually is.
Is this true? If it is, I can feel better when I find a good reading and go with it. Another reason I think this might be ok is because the Ra values are very consistent, but it’s the other 3 that are changing quite ridiculously.
It takes an average. It really depends on the spot you check. Think of a road with pot holes, some areas are great and others are not. Are these parts machined? I can check the parts here if you think you may be getting bad readings. I believe it will check up to a 4RMS finish. Don’t know if this helps you at all.(?)
I know it takes the average (for Ra) but there is no way in hell that these parts are as bad as this thing is saying they are sometimes. They are superfinished parts. Plus it’s also changing quite a bit when testing the same spot more than once. I can’t bring them outside the building let alone somewhere else, but thanks for the offer.
That does a good job of explaining what the parameters are, but it doesn’t quite answer my question.
That’s my next step. This was just a thought. I’m going to have them re-calibrated. If that doesn’t go, I’m calling the manufacturer of the profilometers to see if they have any suggestions.
I’ve been cleaning the parts with alcohol before testing them, and leveling every time. So I’m running out of ideas. I need to figure this out before passing the process down to other people.
Check them out and let us know, then we can go from there. If you don’t know if it is working correctly the rest is just conjecture.
Are you developing a new test, or just unsure of the results you are getting on an existing TM with new parts? Has an R&R study been run to confirm the suitability of the test method and effectiveness of the training?
R&R is going to happen soon if this doesn’t cooperate. I’m just verifying surface finish on this part to see if our process is meeting specs that might be put on a similar part in the near future. We want to make sure we are before committing to it. So this isn’t going to be a permanent operation, just long enough for me to collect sufficient data to say yes we are, or no we aren’t. So I’m just unsure that this profilometer is giving me the actual values.
Took the parts back out and re-tested them when no one was there so no other machines were running around me. I got much more consistent results finally. So I ran it a few times while tapping my hand on the table and bam, there goes the tp and Rp. I guess it’s so sensitive that when other machines are running, it causes enough vibration to throw off the reading. Which is surprising because this table is a pretty thick slab of marble.
Looks like I might be taking those profilometers out of the area soon, we’ll see. I wouldn’t have guessed it could have effected it that much.
Another trick is to change the stroke value of the gage. If your control process does not allow that than have yourself and another associate perform 5 tests independently. When was the last time you did a documented gage R&R on the test equipemnt?
