[Da...] Posted July 13, 2022 Share Posted July 13, 2022 This is a great topic and one most people working in measurement will come across frequently. So for me, this has a two-part answer. The first is Metrology (no not measuring stuff… science). Realistically, your measurement system should be able to resolve to <10% of your feature tolerance - so for a 0.010mm tolerance <0.001mm and so on. When it comes to performing any metrology i.e. uncertainty evaluation or any part of the MSA toolkit, a 10% feature resolution can often end up giving you ‘chunky data’ i.e. the resolution suppresses the ability of the evaluation by not seeing much variation. A nice set of analytic tools I like to use is EMP (Donald J Wheeler) where you perform tests to determine the Probable Error (an estimate of the uncertainty of any given value) and from that, you can determine an effective resolution. This, based on the data will tell you if you are resolving too little or too much... So I like this method because you can analyse what your ‘effective resolution’ is and then supply data that matches the capability of the system you are using. Metrology should be the driver not what people prefer, or how anyone would interpret a standard. You can have data without information, but you can’t have information without data.. (not my quote 😃 ) The second part is to ask the customer if they have a policy.. simple but effective. Link to comment Share on other sites More sharing options...
[Je...] Posted July 13, 2022 Share Posted July 13, 2022 Please sign in to view this quote. . Well said, David. Unfortunately, the driver in many cases, not all, is to use rounding as an exaggeration of the tolerance i.e. "get my parts to ship asap." . Link to comment Share on other sites More sharing options...
[Ri...] Posted July 13, 2022 Share Posted July 13, 2022 The argument typically isn't even about machine accuracy/resolution, it's just the drawing shows x amount of decimal places, so you should round to that. I can get behind accuracy/resolution discussions, but this non-sense with precision is silly to me. People just want to ship parts. The fact that we arguing over 0.0004" or 0.0004mm is silly because your process should be capable enough to not produce parts at the upper/lower limit either way. Link to comment Share on other sites More sharing options...
[Da...] Posted July 13, 2022 Share Posted July 13, 2022 Please sign in to view this quote. Process control now you are talking.. what a beautiful thing, but this does start with a good measuring system, rubbish in, rubbish out. But I 100% agree, and the metrology should deliver a ‘guard band’ thus making the upper and lower limits untouchable by measurement system variation. Link to comment Share on other sites More sharing options...
[Ri...] Posted July 13, 2022 Share Posted July 13, 2022 Please sign in to view this quote. Yeah, let's not rock the boat too much telling them that we will need to reduce tolerance allowance to account for measurement uncertainty. Lol. Link to comment Share on other sites More sharing options...
[Ke...] Posted July 14, 2022 Share Posted July 14, 2022 Please sign in to view this quote. I will (and have) in some cases; but, some tolerances are too tight to guardband. Link to comment Share on other sites More sharing options...
[Je...] Posted July 14, 2022 Share Posted July 14, 2022 Please sign in to view this quote. . Guard banding = We don't understand our processes = wasteful and expensive. . It's funny how the practice of guard banding gets such good PR. It makes the user look like some kind of sacrificial hero, when in reality it just means we don't apply discipline to understanding and controlling our processes. . Link to comment Share on other sites More sharing options...
[Jo...] Posted July 14, 2022 Author Share Posted July 14, 2022 Please sign in to view this quote. When someone ask me what my measurement uncertainty is, they expect a fixed value. So your CMM uncertainty(volumetric) is .0002" ? ok thanks. They usually are walking away before I can interject, that's not the only contributor to uncertainty. Link to comment Share on other sites More sharing options...
[Ke...] Posted July 14, 2022 Share Posted July 14, 2022 Please sign in to view this quote. This statement may be 80% accurate Link to comment Share on other sites More sharing options...
[Je...] Posted July 14, 2022 Share Posted July 14, 2022 Please sign in to view this quote. . 80% or 80.00000000000? . Link to comment Share on other sites More sharing options...
[Ke...] Posted July 14, 2022 Share Posted July 14, 2022 Please sign in to view this quote. 🤣 🤣 🤣 Link to comment Share on other sites More sharing options...
[Da...] Posted July 14, 2022 Share Posted July 14, 2022 Please sign in to view this quote. The precursor to understanding/controlling your process is to understand the measuring system used to monitor/validate, therefore you understand your measurement capability for free in my books 😁👍 Link to comment Share on other sites More sharing options...
[Ri...] Posted July 15, 2022 Share Posted July 15, 2022 I did some work for a company a while back. They chose to utilize Guard Banding. I can tell you this, they got their ass handed to them. The best phrase I recall from the engineering group: This is Aerospace, not wood shop... 🙄 🙄 🙄 🙄 🙄 🙄 (There were like six or so engineers) Link to comment Share on other sites More sharing options...
[Ri...] Posted August 5, 2022 Share Posted August 5, 2022 Sorry, boys - Zeiss told me to go pound sand. https://portal.zeiss.com/my-voice/softw ... 4f5c4baad8 Link to comment Share on other sites More sharing options...
[Jo...] Posted August 5, 2022 Author Share Posted August 5, 2022 Please sign in to view this quote. lol, even they understand rounding is dumb I'm sure it is more an issue with liability. Link to comment Share on other sites More sharing options...
Recommended Posts
Please sign in to comment
You will be able to leave a comment after signing in