33 Collision Error: Lost of commands possible

[[Mi...]]

To give a little context and background, I've been with the current company for about 4.5 years, and about 2 years ago was "voluntold" once the sole programmer left that I was going to be replacing him, which I did without question. I had to obviously use the current library of programs and some of the things done in these programs are just astonishing. I've been able to deal with and overcome most of the issues whether through full on reprogramming or dissecting programs and touching them up.
One thing I haven't been able to fix is the extremely annoying "ConturaXT: 33 Collision: Lost of commands possible" error. The strange thing about this error is that it comes across misleading. It registers as a "collision" through everything, but it is NEVER actually touching anything. Machine can be turned down to 25% crawling and it will still produce this error. We have a part right now that we have never had a problem with, but all of a sudden now this error is infesting this program.
We recently updated to 2023 and updating versions has caused problems in the past (as far back as 2019 from my experience, I can find screenshots of the error earlier in time), but this error was around before that. It may be new to this program after the upgrade, but the error itself happens across all 3 machines we have. Almost all of the air bearings were replaced last calibration, but again the error existed before that happened. We used to be able to run 16 parts in about an hour 8 time span, and the program doesn't want to make it past the 1st part it seems now. It makes it through the base alignment no problem, almost makes it through the whole part basically. It is measuring a circle/cylinder of a coaxial set of holes, moves from one side to the other, positions itself to start taking hits, and then crashes. Velocity 0 and everything else indicates the probe was not actually moving and doesn't actually make contact in any way. Machine speed is turned down both at the Run Window as well as individually through features. We run these machines slower than they are capable of because they tend to be very jerky. I will answer any other questions or look into whatever is suggested, just trying to narrow down this issue. The tech that was here previously was confused when it happened on a different machine and I showed him the smart services report that there was 0 velocity and the move speed and everything didn't add up to a collision; I said your machine is registering a collision and making it look/appear as if we are beating this machine up so when you come in and look at it, it looks like on paper it was in a street brawl. It is not always a circle/cylinder that makes this crash, it can be various features.
I realize the quick and easy method is to reprogram and phase out these older programs, but I'm more curious if anyone has run into this exact error without collision and had a cause or anything narrowed down. This error comes out of nowhere, and as of right now anything that has been reprogrammed in the last year and a half to two years doesn't seem to produce the error.

*Edit: I should probably add, machine is a Contura G2 with a Vast XT. Either 2007/2008/2014 (One of the older is a XXT, other 2 are XT's). Running 2023. Non RDS or XTR as it has not been retrofitted in any other fashion than having FutureFit upgrades installed recently. Problem still existed before/after that upgrade.

[[Ma...]]

We have VAST XT and experiencing this type of error only on curves where is sometimes inverted vector into material.

If probe is not touching a part and error occurs, then i would suspect head. If it's in contact, then carefully check measured feature.
Or call ZEISS for help.

[[Je...]]

Hello Michael,
I can see you are quite frustrated with the 33-Collision error. Hopefully I can alleviate some of that frustration by revealing some information.

32-Collision = Physical contact - CRASH - Check for damage then fix the program
33-Collision = No physical contact - AIR PROBING - No damage. Fix the program.

Now this isn't a comprehensive explanation but that should help you to determine where the problem lies. For each, if the occurrences are "random" or unpredictable then first place to look is your base alignment to review if it is robust enough to repeat. If the occurrence is predictable and always happens at a specific place, review the nominals and strategy for that feature specifically.

[[Mi...]]

We have gone through several probe head changes over the past year and some change on this machine in specific. The problem was present both before and head all head changes, and again this problem is across 3 machines. This machine in specific has been having head problems, but it was just replaced a few months ago and is not showing any symptoms of being off. Zeiss has had this problem brought to light many times as they have been here on several services calls and have not narrowed down anything specific. This is just a cylinder to cylinder in a coaxial fashion. There are lots of weird movements and multiple features checked multiples times for different methods of reporting, the program could definitely be cleaned up. But even on other programs that have been "cleaned up", the problem still seems to exist. I am going to have my programmer look into making changes aside from what we have already tried. The program has ran flawlessly for almost 3 years, it's odd that it comes out of nowhere after an upgrade to software but wouldn't be the first time.

Thanks for the information, I was unaware that 33 meant a no contact collision.

[[Ja...]]

I have not run into such problems yet. But I would suggest you start with CAD healing and CAD model comparison to reapply features to CAD model. Any features that are out after reapplying might need some special attention. I usually extract the feature again, copy strategy and other settings from old one, delete old one and rename the new feature same as old one.

[[DW...]]

Please sign in to view this quote.

Please sign in to view this quote.

It would seem to me that this statement would be indicative of a problem with the programming.

Can you post some screenshots of the feature(s) strategy windows where this error occurs?

I am assuming there are multiple CMM technicians, and only one CMM programmer. Is it possible for technicians to save any changes they make to a program? This can be found under "Change Settings">"User">"Privileges" tab. Perhaps a technician saved a space point inadvertently placed inside a features strategy? Or as a feature itself?

[[Ni...]]

I usually got this error when running the part and then manually lowering the speed. That causes the machine to jerk which registers as a crash.
So a fix would be to just let it run at the current speed without touching the speed knob. You'll need to probe out the program first to make sure you can run it at the max knob speed.

Another fix I used to implement was to change the "Start Angle". I noticed this would only happen with circle and cylinder features. So, adjusting the Start Angle would usually fix it.

Last thing I did was raise up the probing Dynamic (I don't think the Vast XXT has it) and change the probing behavior of passive sensors from Sensitive to Normal.
I would change it from sensitive to normal when qualifying the probes too.

That would usually stop it from giving those errors.

[[Je...]]

Please sign in to view this quote.

I would be interested in seeing the features and associated strategies applied for which this is occurring. Based on your description of the symptoms, I can say with strong confidence that the issue within your inspection program and likely is within the feature strategies. Be it an insufficiently defined Base Alignment, a local/secondary alignment that is improperly defined, clearance/retract distances that are insufficient, incorrect nominals or even non-compliant parts (machine can't measure it if it is outside the expected range of tactile search distance).

I am also curious if you notice a message "***SCANNING OPTIMIZATION***" across the top of the user interface window just before the "Collision 33" error occurrence. This would only happen on the XXT enabled machines. This is generally a symptom of poor measurement strategy relative to expected part conditions. Generally an excessive form error or delta from nominal size/location (actually hitting the high/low limits for force on the passive sensor due to scanned surfaces deviating from nominal location). Fixing the program is the correct resolution but there is a setting to eliminate the repeated scan option as well. I advise visiting the KNOWLEDGE BASE and searching for "Scanning optimization" to learn more.