Accuracy of a Prismo Navigator MASS

[[Li...]]

A Quality Engineer just asked me how accurate is our most accurate CMM. The Prismo we have is our most accurate CMM, but what spec do I give him? I see U=0.1 + L/843 (Uncertainty of measurement) in the C99/II tech data book for the machine.

Can someone explain this to me so I can explain this to him. My guess is that a customer has requested this info.

[[Mi...]]

Please sign in to view this quote.

The larger the distance you're measuring, the larger the uncertainty.

L is the feature size in millimeters, and I think the uncertainty result is micrometers, so for a 10mm feature, the uncertainty is 0.1μm + (10/843)μm = .112μm.

[[SH...]]

Please sign in to view this quote.

I think it should be 1+L over 843.

IMG_20200819_231303.jpg

[[Er...]]

My personal opinion is that MPEe is a useless number. (Other then a quick argument with a hexagon sales person, "Mines better") How many of you are measuring gauge block like parts?

Give him numbers for RONt and MPE-THP or something that more reflects the parts you are measuring.

Or run a VCMM to get some estimations of what you can achive.

[[Ke...]]

I can hear the customer now: "Great, that is what we want *all* of our parts ran on" 🤣

[[To...]]

Here's something from Measuring Hero Ryan Stauffer.

[[Da...]]

Please sign in to view this quote.

I don't want to say you are not the best, because you are, but my 2 cents would be: "look QE, ALL those numbers mean nothing. Because we only took minimal formal training and our measurement strategies and GD&T knowledge are questionable. We shouldn't even get into how company stimulates our will to work".

Again, that is nothing to take personally.
Stay sharp and fight hard like a metrologists have to these days.

[[Is...]]

Better if you check the capability of your Methods in each characteristic..
Using VDA5 studies you can calculate QMP and QMS according to your design specifications.
You can include all the uncertainties you want; Normally we only consider MPE, RONt, resolution and the temperature as fixed uncertainty values additional to the Bias and the repeatability (using different days, parts and clamping as factor of reproducibility).
in the automotive industry this is a requirement from german customers and capability tolerances are agreed with them.

These screenshots are from google (using Qdas ):

qdas.png

[[Cl...]]

Please sign in to view this quote.

Maybe a simple measurement of a class XX ring gage and grade 00 gage blocks would be enough for the customer.

[[Ke...]]

Seriously, It really does depend on how deep you want to dive. Knowing the uncertainty of your measurements is definitely important; moreso, if you are actually working with tight tolerances. As Israel pointed out, different evaluation methods (2D vs 3D; scanning vs single point) all have different uncertainty values. HN Metrology Consulting has a great (and affordable) online training module for learning how to calculate measurement uncertainty.
https://www.hn-metrology.com/

[[Ch...]]

Keith, thanks for the article and look who wrote it. Who know that guy from the meme was also a metrology consultant.

meme_0.png

[[Ke...]]

🤣 🤣

[[Aa...]]

Please sign in to view this quote.

Eric (or anyone who knows can chime in),
I would like to know more about this Virtual CMM software. Did you have to pay extra for it? Is it restricted in its availability? How do you access it?

[[No...]]

I don't have it, but it's of course an option you have to pay for. And it's not plug and play. It needs some kind of preparation. I don't remember exactly, but it needs accuracy data that must be collected on the physical CMM you want to use it for, since this is what its calculations are based on. That data collection is probably also done by Zeiss, for extra cost.

[[Je...]]

Please sign in to view this quote.

.
.
.
Thanks, Dane, for providing sober insight on the role programmer skill plays in measurement uncertainty. The struggle is real and all-too-often overlooked.

Although I've been with companies with a similar culture to what you described, my present company is a stark contrast. They prioritize employee training and ensure metrologists understand GD&T. Overall, it is simply a great company to work for. Everyday, I feel engaged in my work and challenged to get better. These conditions lower (but don't eliminate) the risk of programmers introducing measurement error.

I just wanted to share a success story. Although rare, there are great companies out there for metrologists. I had to leave a previous employer to find this one. I realize that's not always possible for people, but the current employment landscape is screaming for anyone with Calypso experience.


Jeff Frodermann
Meier Tool & Engineering
Anoka, Minnesota
.