hexagon logo
Register Sign in
  • Replies 31 replies
  • Subscribers 224 subscribers
  • Views 315 views
  • Users 0 members are here
Options
Related

Upper Management is looking for answers/Justification on probe breakage

SMace
Hello ALL GURU's:
Here's a great post for a lot of input on WWGD (What would Gurus Do)?
Lets start by saying there are NOT an excessive amount of probes being broken yearly. I would estimate less that $500.00 - $700.00 yearly max, if that.
However, The uppers want justification.
Now, it has been suggested to slow the rapid in the machine settings, which would cause more cycle time on check parts that are in a Pallet setting.
There is a CRASH attachment from Renishaw. Anyone currently use these?. And do they work. Say on a .7x20mm probe or a 1mmx27mm probe?.
In our shop, we have 4 Cmm's that all have touch probes and 2 have vision as well. What would a normal (realistic) probe consumption be expected
at your place?. Mind you that the specific probes that we have in question are 1.5x30mm. 1mmx27.5 and .7x20mm. Sometimes it may just be human error (can't help that).
Maybe a slight misload, part has been machined or molded not into spec. Other times, maybe doing a manual align and you forget the rapid is on and "Snap". It happens.
I was asked to figure the FORCE in which a probe breaks. Well, honestly what other factors go into that one?. Are you rapiding, your prehit/retract too close and you touch the part and snap?.
Seeing as we are a high volume molding/machine shop with multiple parts at any given time to inspect via DCC. What is the best course of action to go ZERO probe breakage a year?.
This would not include Set ups, Prove Outs as well.
I am just looking for some insight on this issue from a lot of people who know way more than me!. Hence, picking the brains of the Guru's. ANY input would be greatly appreciated and if you want to keep it short and sweet. great. I don't want to consume too much of your time writing.
Thank You All Slight smile

  • You're not working very hard if you don't break a few tips here and there. Also a CMM is just like any other machine in the shop, you are going to have to replace tips from breakage just like drills, taps, endmills, ect. Just do your best to keep it to a minimum.
  • This is a little late to the party, but this is what works for us. Our operators don’t break tips very often. The programmers break most of the tips.
    • Operators most frequently break tips when they either setup the fixturing/part wrong or are running the wrong program. Putting a picture of the setup into your program solves both of these problems. When they have it setup wrong, the picture pops up and it doesn’t look right, they can make corrections before the CMM starts to move.
    • All programs that are meant for measuring production parts should be stored in a write protected folder. This protects from operators/programmers accidentally making changes and then those changes get saved into the program. Furthermore, PC-DMIS is constantly saving and overwriting the programs. Occasionally through no operator interaction, it can save something dumb in the program that causes you chaos. Don’t let it overwrite your known good programs. Write protect the folder at the OS level.
    • Minimize the chance of the operator running the wrong programs. Have separate areas for “Production Ready” programs, “Archives” of older programs, and where you develop new programs. The operators should only ever be operating out of the “Production Ready” folder and it should only have the one program that they should be running. Old versions and experimental versions should be in the archive or developmental areas.
    • Fixturing -> Rough Alignment -> Part Program. Ideally the entire program runs in DCC. The less you ask an operator to do the less problems you’ll have. You need to find a way to ensure that the part is always in the same location on the CMM. The best way to do this is with fixturing. Then the program performs a rough alignment to find the part using larger approach and retract distances. Then inspect the part after the program knows exactly where the part is. All of our parts are roughly the same size, so we have a corner nest on the CMM that all the fixtures nest into. Our operators put the part into the fixture, put the fixture into the nest, and then run the program, and then a picture pops up and they verify everything looks right and click OK, and then the program just runs. If you really need to have the operator measure the part at all, use pictures to help them so they don’t ever have to guess.
    • Use clearance planes. A lot of problems happen when you mark some stuff out and then it drives through the part. If you use clearance planes it kind of modularizes your program, where you can mark stuff out without worrying about the move points between features.
  • Do you really need to use very small probes? As mentioned, they are very easy to break. I use a 3 x 30 probe as a standard, of coarse, what I measure I can get away with this. I would do as most above say, and if you can use a bigger probe, do that. If you think the readings will be different, do a study, it might surprise you. Bottom line is what the neon states.
  • We go thru several thousand dollars worth of styli each year. every now and then it will be when we prove out a program but the majority of our problems are operators loading the wrong operation program or entering the wrong op for a parametric program or not setting the part up correctly (not reading or following instructions). We currently have 1400+ programs in our working directory and another 6k plus in doc control and all have been used at one time or another but people can't seem to follow instructions with pictures.

    Management knows this is an issue and they do ask us to do what we can to help but when shop floor people run their own parts they will get broken and they understand that. We've got 13 CMMs and they are ran by approximately 200 different people each day.
  • Why do they break?

    [ ] Our alignments are incomplete and unstable, making the probe go to bad places
    [ ] We make really bad parts, the holes aren't where they should be
    [ ] We never test the programs in low speed for safety reasons before running them full tilt, que será será
    [ ] Operators are drunk / don't care at all
  • in reply to dwade
    dwade,
    In this situation I would ask them to invest into a scan barcode system.
  • Just an FYI if the shank is still intact, you can just have them replace the ruby costs like 17 bucks on average
  • Just keep telling them its on your to do list until they forget...
  • I've had some recent instances with ITP styli, where we think the ball just fell off. We lost about 3 in one week. They were contacted and replaced them FOC.
  • The TP20 module seems to be pretty solid haha. I accidentally dropped it today and no issues after re-running some part layouts and comparing to previous results.
    As for the tips, it's easy to break one when you are programming and rushed or impatient, so take your time if you're programming online and even if offline, be sure to test slowly once online.

    Also no matter how dummy proof you have made your programs,
    I always run the CMM slow during a new batch of freshly produced parts. I do this for the first part only, this way if anything is damaged from production or extremely out of spec that may have been missed by qualities first off, you can be more confident. It's good to leave a prehit and retract of 5mm in case there is a form oopsie from production that slips through.

    Just my 2 cents Slight smile

    Cheerios!
© 2024 Hexagon AB and/or its subsidiaries. Privacy Policy | Cloud Services Agreement