Advise on R Chart with Ppk of 17.85 for Ford

[mlee97]

Re: Need advise on R chart with PpK of 17.85 for Ford

Some additional info..

The operators were instructed to take the samples from the samples run (324 pcs) spread out evenly. They numbered the parts as they sampled so they should be fairly close.

In production the surface finish is machined by a large drill that just spots the face. In the plant's history we have never had a general surface finish failure reach our customer in over 15 years of my history on any production lines. We only have special cause deep scratches, dings, or missing spot faces, never normal tool wear related rough finish on the aluminum.

Our Machining procedure requires a surface finish check on these parts every two hours, so despite other misgivings, the customer is protected from variance creep. However, If I don't get this resolved, we are bound to 100% test.

 

[Bev D]

Re: Need advise on R chart with PpK of 17.85 for Ford

In effect you are saying that I should stop the line and invest real money in searching for something minute that is not a trend, and that my high PPK amounts to nothing because of a tiny shift that had it been a hair smaller would have been all Great, fabulous machining lines, but is now a showstopper because I can't get PPAP acceptance. This idea that the 3 sigma control line is some magical decider of fate with no leeway so that the PPK values suddenly become meaningless at a single line in the sand goes against my understanding. Why does a great PPK at an control line SD of 2.99999999 have great value, but a 3.00000001 is suddenly completely worthless?

I - Bev D - am agreeing with you. it is crazy...we need to engage our brains and stop black box thinking

 

[mlee97]

Re: Need advise on R chart with PpK of 17.85 for Ford

Is there no ratio of tolerance based PPK to 6sd too high for a rule of thumb to pay less attention to the run? I mean, my example is strong but you could make up examples that take it to ridiculous lengths.

 

[Bill McNeese]

Re: Need advise on R chart with PpK of 17.85 for Ford

Our Machining procedure requires a surface finish check on these parts every two hours, so despite other misgivings, the customer is protected from variance creep. However, If I don't get this resolved, we are bound to 100% test.

There is no way you need to do 100% inspection with this Ppk. I don't think anyone is suggesting that. If the Ford fellow is, then I definitely disagree with that. He needs to come back to earth and be realistic. A "in control" process will have out of control points sometimes - entropy happens. With that Ppk, you should sample very little.

My point is, if you are taking the data anyway and have evidence of a special cause, taking some time to see what it might be is a good approach. As Dr. Deming said - improve constantly and forever - constantly strive to reduce variation.

 

[adamsjm]

Re: Need advise on R chart with PpK of 17.85 for Ford

So you are asking me to investigate a shift that I can't reliably see with my gauge,

If the shift is less than gauge R&R then is the data collected valid?
Your raw data shows shows 110 different values for the 125 samples collected, but your R&R study shows only 16 ndc's.
Thank you for your R&R analysis, but what is the GR&R numeric value?
Your measurements may be giving you a false Ppk value.

 

[Bill McNeese]

Re: Need advise on R chart with PpK of 17.85 for Ford

Your Gage R&R has a standard deviation for the part-to-part variation of 7.75. How is that possible when the 125 samples have a standard deviation of about .16? Something doesn't seem right.

 

[Steve Prevette]

Re: Need advise on R chart with PpK of 17.85 for Ford

The data look good to me, I do agree with the others that something must be wrong with the gage R&R or you couldn't be getting data down to the precision you are getting in your measurements.

I'm probably due to visit New Brookland Hobby Shop in West Columbia - are you any ways near there?

 

[mlee97]

"I'm probably due to visit New Brookland Hobby Shop in West Columbia - are you any ways near there? "

It's about seven miles from me. Let me know when you're coming, maybe we could go out to eat and talk shop..

Thanks,

Mitchell

 

[Steve Prevette]

"I'm probably due to visit New Brookland Hobby Shop in West Columbia - are you any ways near there? "

It's about seven miles from me. Let me know when you're coming, maybe we could go out to eat and talk shop..

Thanks,

Mitchell

Sounds good. I sent you a private message with my home email. Probably can make it over this weekend or next.

 

[mlee97]

Well there were problems with the GRR/MSA, The biggest was that one cannot get tight repeatable results because the soft aluminum is deformed by the probe so one cannot test in the exact same place very time, yet the area for the test is highly constricted due to the part and fixture shape. The track from the previous test is so small and invisible to the eye, that it cannot be reliably avoided, so even treating it as a destructive test is adulterated.

Secondly, because true process variation is so small in on our process, Ford suggested that we artificially rough up the surface on some of the MSA parts. When this was performed it created parts past the upper spec limit for this spot face but within tolerance for other surface finish measurements on the part. The Ford SQA bought off on the MSA as is, despite it being as ugly as a dirt clod.