Reporting alignment used to check true position

[giacal]

Hello, I am new to the forum and not familiar with PC-DMIS at all, so apologies in advance for the noob question.
I have to review CMM reports submitted by a supplier, the first thing that I noticed is that the information about the alignment used to check geometric tolerances is not reported. Also, nothing is stated about the number of points used to check the feature and their distribution on the feature itself. Without this information it's difficult to accept the report...
Is there a way for PC-DMIS to report the alignment used to check GD&T and the "probing strategy" followed during inspection?
Thank you in advance for your help, if you could point me where to find basic info like this about PC-DMIS (without having access to a license) I would be very grateful.
Cheers

[RandomJerk]

I'd personally be reluctant to send the information you're requesting unless the requirement was stated at the onset of the project and customer guidelines were laid out. If I check a critical bore that's also a datum, for example, during first machining operation and it's good, that actual measurement strategy may not be on the report as I've already proven the datum to be correct, I won't take as many hits on the second op program. There are too many factors that go into the decision making process for that to list.

However, to answer your question, ask your supplier to right click on the report window, select "Edit Object...", then check the boxes next to Show Features and Show Alignments. You may also request a pdf of the actual program (File...Printing...Edit Window Print) to show you how the named alignment was constructed.

[louisd]

Without having access to a license of PCDMIS, you are reading an excel file, text format, or a PDF file. It is absolutely impossible to extract or produce any "extra" information from these report outputs.
I must ask however:
What is mandating you to have a need to understand the alignment structure used to extract a 'geometric tolerance' (I'm presuming this means features without FCF (feature control frames) and defined datums)?
What is mandating your need to know the number of hits or points taken to extract the measured value?

These questions are not normally asked when reviewing, or even auditing measured data.
If you question the method of measurement or accuracy of the data, typically you just get a 3rd party involved to confirm measured values.

Does your drawing have any SPC dimensional requirements? If so, you MIGHT have some bigger ground to stand on, but otherwise, the reported numbers should be taken and honored at face value. An inspector's personal and professional integrity is represented on that report. If you feel they pencil-whipped the numbers, and want to catch them red-handed, you'll need objective evidence in order to question their integrity. This brings you right back to obtaining a 3rd party inspection report.

[giacal]

I'd personally be reluctant to send the information you're requesting unless the requirement was stated at the onset of the project and customer guidelines were laid out. If I check a critical bore that's also a datum, for example, during first machining operation and it's good, that actual measurement strategy may not be on the report as I've already proven the datum to be correct, I won't take as many hits on the second op program. There are too many factors that go into the decision making process for that to list.

However, to answer your question, ask your supplier to right click on the report window, select "Edit Object...", then check the boxes next to Show Features and Show Alignments. You may also request a pdf of the actual program (File...Printing...Edit Window Print) to show you how the named alignment was constructed.

Thank you for your actionable input, RandomJerk. Actually one of the requirement of the procurement is to define the measurement strategy for most critical dimensions and tolerances, so yes, it was defined at the onset of the project, but the supplier clearly overlooked it.

[RandomJerk]

Ah, in that case they should have provided info. I think you'll find many are gun shy about these things as more often than not they come up long after the fact.

[giacal]

Without having access to a license of PCDMIS, you are reading an excel file, text format, or a PDF file. It is absolutely impossible to extract or produce any "extra" information from these report outputs.
I must ask however:
What is mandating you to have a need to understand the alignment structure used to extract a 'geometric tolerance' (I'm presuming this means features without FCF (feature control frames) and defined datums)?
What is mandating your need to know the number of hits or points taken to extract the measured value?

These questions are not normally asked when reviewing, or even auditing measured data.
If you question the method of measurement or accuracy of the data, typically you just get a 3rd party involved to confirm measured values.

Does your drawing have any SPC dimensional requirements? If so, you MIGHT have some bigger ground to stand on, but otherwise, the reported numbers should be taken and honored at face value. An inspector's personal and professional integrity is represented on that report. If you feel they pencil-whipped the numbers, and want to catch them red-handed, you'll need objective evidence in order to question their integrity. This brings you right back to obtaining a 3rd party inspection report.

Thank you for your insight, louisd. When I say geometric tolerance I mean, for example, a true position tolerance with respect to a defined datum system, therefore the alignment is quite critical for the outcome of the measurement. Both the alignment and the probing strategy (actually the whole measurement procedure) must be reported as part of a contractual requirement, my question is more related to understand if PC-DMIS is capable to report this data. We cannot take the numbers at face value not only because the manufacturer himself is producing the report (and we have reasons to question its skills in interpreting and inspecting GD&T), but also for traceability and audit reasons: one day, ten years from now, we could be asked for whatever reason to produce evidence of how that part was inspected.

[Mike Ruff]

1. Has your inspection dept. measured the parts as failing? If not, why is this an issue?

2. Do you require a specific measuring strategy? If not, why is this an issue?

This does not seem like grounds for rejection IMO, especially if the parts are conforming. Even if a specific measuring strategy is required, there is a lot more to it than that, such as filters, fitting algorithms, etc. that can vary your results.

We have made a LOT of aerospace and medical parts, I have never had this type of requirement before, even on flight-critical components or instruments used for open heart surgery...

[giacal]

1. Has your inspection dept. measured the parts as failing? If not, why is this an issue?

We don't have such facility, it's the manufacturer's inspection dept which has measured the part. The issue here is to have full traceability of the results, and without information about the alignment it's not possible to demonstrate it.

2. Do you require a specific measuring strategy? If not, why is this an issue?

We do not require a specific measurement strategy, but we have to make sure that the part is inspected according to metrology good practices. As I mentioned, we have reasons to question the manufacturer's metrology and inspection skills. We are not asking for high-level specifications here, we just want the report to show that features were probed with a sufficient number of points well distributed, and the alignment was done properly.

[louisd]

the alignment and the probing strategy (actually the whole measurement procedure) must be reported as part of a contractual requirement,

We do not require a specific measurement strategy, but wehave to make sure that the part is inspected according to metrology good practices.

Contradicting? --If you are referring to GMP per FDA, this is heinously overkill.

If you actually have this as a contractual requirement with the supplier making your parts and inspection report, reject the product and send parts back, stating they didn't comply with the quality reporting obligations of the PO. Let supplier deal with how they want to meet those requirements.

-Yes PCDMIS can output number of points, with as stated above, or via a ".numhits" output.
-I don't know how the distribution of hits about the feature can be effectively communicated without supplier manually assessing each measured feature or component and inserting a comment describing the distribution of hits.

Hint: there is not a single document that will tell you "thout must use x number of hits to measure this feature (or shape)", unless you have defined that internally, and communicated it to supplier. So, to try to hold supplier accountable for measuring with a point density not to your liking, is going to be a real knee-slapper.

Again, if you don't trust suppliers measured values, 3rd party inspection needs to get involved, so you have objective evidence of negligent acts and falsification of data, or at least go back to supplier with your concerns being quantifiable and backed up, so you can "develop" that supplier to meet your company's expectations of PO fulfillment..

[MikeD6]

Sounds like you need to visit them and audit their measurement process if you're that worried about it. That's the only way you are going to get the information you are looking for. If they let you, I don't know what kind of quality clauses you have in your purchasing contracts. Things like probing strategies, the number of hits used, are buried in the code. Like others have also said, I've never had a customer request the type of information that you are asking for, unless there is an issue. Not saying what you're asking for is unreasonable, just unusual.

Bear in mind that the alignment used to physically measure the part and the alignment used to generate the measured values on your report can be, and often are, different. What's important is that the alignment used to measure the part is consistent and the fixturing is stable so that the parts are measured consistently. Once that is complete they can then use the measured features to create alignments that match the datum structures on the prints. My point is that their fixturing can be just as important as their measurement strategy in obtaining accurate results and the only way to know what it is is to ask to see it.

For example, I measure primarily sheet metal parts, so most of my "measuring" alignments are created by measuring the fixturing that holds the parts rather than the parts themselves. I can then use measured features to create any alignment I please to report out values. Hope this helps.

[giacal]

Hint: there is not a single document that will tell you "thout must use x number of hits to measure this feature (or shape)", unless you have defined that internally, and communicated it to supplier. So, to try to hold supplier accountable for measuring with a point density not to your liking, is going to be a real knee-slapper.

I agree with you that no document mandates a measurement strategy, but there are good practices like the ones described in NPL's GPG41. Based on those, I can question the supplier's strategy if it's measuring a 50-mm deep hole with 3 points on a single cross section near the top surface, for example. Now we can discuss if this ground is solid or not, but at least it's a baseline to check if what the supplier is doing is sensible or not. The context here is that we are not dealing with metrology service companies whose core businesses are 2nd and 3rd party inspections, otherwise I would have gone there to witness some random inspections (and learn a lot about PC-DMIS and CMMs in the process), and be at ease.

[giacal]

Sounds like you need to visit them and audit their measurement process if you're that worried about it. That's the only way you are going to get the information you are looking for. If they let you, I don't know what kind of quality clauses you have in your purchasing contracts. Things like probing strategies, the number of hits used, are buried in the code. Like others have also said, I've never had a customer request the type of information that you are asking for, unless there is an issue. Not saying what you're asking for is unreasonable, just unusual.

Bear in mind that the alignment used to physically measure the part and the alignment used to generate the measured values on your report can be, and often are, different. What's important is that the alignment used to measure the part is consistent and the fixturing is stable so that the parts are measured consistently. Once that is complete they can then use the measured features to create alignments that match the datum structures on the prints. My point is that their fixturing can be just as important as their measurement strategy in obtaining accurate results and the only way to know what it is is to ask to see it.

For example, I measure primarily sheet metal parts, so most of my "measuring" alignments are created by measuring the fixturing that holds the parts rather than the parts themselves. I can then use measured features to create any alignment I please to report out values. Hope this helps.

Yes it helps a lot indeed, thank you! In fact what I would like to see in the report is exactly how they create the alignments and how they are using them to generate the measured values. I understand that the "measuring" alignments are important in operating the machine, but less so in assessing compliance with specifications, because they are overridden by the alignment corresponding to the datum systems defined in the drawings. Is my understanding correct?

[MikeD6]

Yes it helps a lot indeed, thank you! In fact what I would like to see in the report is exactly how they create the alignments and how they are using them to generate the measured values. I understand that the "measuring" alignments are important in operating the machine, but less so in assessing compliance with specifications, because they are overridden by the alignment corresponding to the datum systems defined in the drawings. Is my understanding correct?

That is correct. Compliance to the print per the datum systems is what makes a part good or not.

[Quality ish]

We don't have such facility, it's the manufacturer's inspection dept which has measured the part. The issue here is to have full traceability of the results, and without information about the alignment it's not possible to demonstrate it.



We do not require a specific measurement strategy, but we have to make sure that the part is inspected according to metrology good practices. As I mentioned, we have reasons to question the manufacturer's metrology and inspection skills. We are not asking for high-level specifications here, we just want the report to show that features were probed with a sufficient number of points well distributed, and the alignment was done properly.

You don't have the facility to check the part yourself, but want verification from the supplier that they did it correctly...

That's just asking for problems, if you ask me. Sometimes, when you're playing cards, you've got to call the other player to keep them honest. In this aspect, you've either got to be able to measure the part occasionally yourself to ensure it conforms to print, or submit it to a 3rd party for verification.

[kingsld1]

If you have to see where the points are being taken there are really only three alternatives in PC DMIS.

One is that the program be written only using points instead of circles, lines, planes, etc. Then the circles, lines, planes would be constructed from the measured points. A screen capture could let you see all the points you want. This would mean a complete rewrite of their programs. Your vendor will scream and use many many bad words at you and about you.

The other possibility is to display path lines of the individual features. He might be able to do a screen capture with path lines displayed. Depending on how complex the part is, this may take a few minutes or several days.

The only other thing I can think of is to grab a video camera and make a movie of the part being run

[NinjaBadger]

I'll agree with what most others have said but add this...


Having this information provided to you is of limited use, especially if you don't trust your supplier. I could add all sorts of information to a report and still fudge the results if I were so inclined (I must stress, I'm not and I don't!)

If you can't check the parts yourself you need to send it out somewhere and perform a correlation study to get confidence in your suppliers results.


Regardless of if I think it's a good idea or not, to actually answer your question however you could do the following...

1) As previously mentioned get them to switch on Features and Alignments on the report (Right click in Report window > Edit and tick Features and Alignments
2) Ask them to report on alignment features immediately after each alignment in the program, i.e. so if you have Alignment ABC report on Datum plane A to show it's zero in whichever axis.

Oh, I like the screen shot with path lines shown idea as well!