Advise on R Chart with Ppk of 17.85 for Ford

A

adamsjm

#11
Re: Need advise on R chart with PpK of 17.85 for Ford

Per your PPT presentation, you say "the sigma is below the size of the specification units and smaller than the MSA resolution." If this is the case, what is your Gauge R&R?
 
Elsmar Forum Sponsor

Steve Prevette

Deming Disciple
Staff member
Super Moderator
#12
Re: Need advise on R chart with PpK of 17.85 for Ford

Two things: You have a process shift in the data you must account for.

Second - consider that the control limits on the xbar - R are for the SETS OF FIVE. They are not the limits on the individuals. Consider the individuals chart I made. I think that is the source of confusion about the control limits are "too tight". Remember, the standard deviation of the average of five items is the individual standard deviation divided by the square root of 5.
 

Attachments

M

mlee97

#13
Re: Need advise on R chart with PpK of 17.85 for Ford

No, I don't want to account for the shift, that is the point. If the PPK is so good and the estimated PPM is no defects per 100 million parts, then it makes no sense to go looking for ghosts. It doesn't matter whether I can explain them or not or improve it or not. The variation is too small to indicate a problem with the gage or the machine. It would never effect the customer. (contract is for less than 5 million parts).
There is no risk to the customer at 17.8 PPK. The run chart is a simple snapshot predictor. But it can be wrong, even when the math is right. It is similar to when the Run chart shows a bad run of 5, but there was no reaction. If the data self corrected and ran another million parts with no out of control conditions, you wouldn't consider it still out of control. For an open ended SPC chart, it matters where the out-of-control condition occurred, earlier or later. Remember that the purpose of SPC is to protect the customer from defects. Once they are protected, there is no need for further reaction.

What I am looking for is a rule of thumb or a ratio which defines the point of diminishing return for SPC run charts to PPK.
 
Last edited by a moderator:
M

mlee97

#14
Re: Need advise on R chart with PpK of 17.85 for Ford

Per your PPT presentation, you say "the sigma is below the size of the specification units and smaller than the MSA resolution." If this is the case, what is your Gauge R&R?
Well I'm not at work now, but if I remember correctly, it had a SV of 6 percent and the average repeatable variation of same parts was around 0.400 RZ

I'll have to find it tomorrow.
 

rickpaul01

Involved in HankyPanky
#15
Re: Need advise on R chart with PpK of 17.85 for Ford

No, I don't want to account for the shift, that is the point. If the PPK is so good and the estimated PPM is no defects per 100 million parts, then it makes no sense to go looking for ghosts. It doesn't matter whether I can explain them or not or improve it or not. The variation is too small to indicate a problem with the gage or the machine. It would never effect the customer. (contract is for less than 5 million parts).
There is no risk to the customer at 17.8 PPK. The run chart is a simple snapshot predictor. But it can be wrong, even when the math is right. It is similar to when the Run chart shows a bad run of 5, but there was no reaction. If the data self corrected and ran another million parts with no out of control conditions, you wouldn't consider it still out of control. For an open ended SPC chart, it matters where the out-of-control condition occurred, earlier or later. Remember that the purpose of SPC is to protect the customer from defects. Once they are protected, there is no need for further reaction.

What I am looking for is a rule of thumb or a ratio which defines the point of diminishing return for SPC run charts to PPK.
Right, if the ppk is really 17.8, tell your customer to stop wasting your time.
 

Steve Prevette

Deming Disciple
Staff member
Super Moderator
#16
Re: Need advise on R chart with PpK of 17.85 for Ford

However - no matter what the PPK - you had some sort of process shift going on in your data. The process IS NOT STABLE - and therefore PPK is a MEANINGLESS CALCULATION. I would stronly recommend figuring out what the data shift was, and what caused it, and get the process stable. Then continue to monitor it so you know the process is stable.
 

Bill McNeese

Involved In Discussions
#17
Re: Need advise on R chart with PpK of 17.85 for Ford

I would be interested in knowing how you selected the 125 pieces out of the 300 plus piece run. From my perspective there are a number of ways ? process, specifications and customer - to look at this situation.

First, take a look at the process as described by the data with worrying about the specifications. Steve has run the numbers. I ran them also (see attached). The first question to answer is about the stability of the process. I have included the Xbar-R, individuals, process capability and histogram charts. As Steve points out, there has been a process shift that shows up great when you split the control limits. Something has happened to cause that shift. Note that it is a ?good? shift ? there is less variability in the process. If you can figure out what happened and ensure it continues to happened, you have improved the process considerably. You mention that you don?t want to account for the shift. As your customer, I would want you to ? it has improved your process and means less variation coming into my process.

Now look at the issue from specifications. Here, there is no debate. This is process is very capable of meeting specifications. That is seen easily from the high Ppk number. But there is no guarantee of keeping it there because the process is not stable.

Now look at the issue from the customer's perspective. As a customer, I want a product that is produced from a process that is in statistical control and is capable of meeting specifications. I want your process to be in statistical control because that means that the product I am receiving from you is homogenous - made from a consistent and predictable process. If your process is not in control, I cannot be sure of receiving the same homogeneous product from you in the future. And from your data, your process is not in control and you are missing an opportunity to ensure that the variation remains reduced.

I also want it with in specifications because that defines the limits of what works for me. However, I don't want you to send something that is near the USL one time and then near the LSL the next time - that could cause havoc in my process. That is why statistical control is so important ? that is the only way I have to ensure that you will continue to perform.

So, I tend to agree with the Ford fellow. I don?t know if you need to redo it, but you need to understand what happened because I want to get product from a process that is stable. And, in the end, he is the customer.
 

Attachments

bobdoering

Stop X-bar/R Madness!!
Trusted Information Resource
#18
Re: Need advise on R chart with PpK of 17.85 for Ford

I would be interested in knowing how you selected the 125 pieces out of the 300 plus piece run. From my perspective there are a number of ways – process, specifications and customer - to look at this situation.
This is a good question. In a machining operation, the data needs to in time-ordered sequence. My preference is taking a part every 3rd pc. Ford likes to see 5 pcs per sample, but that is only applicable when analyzing data from a normal distribution established from random, independent data - which is rare in machining if you are in control. In precicion machining, individual data - not averages - tends to be more meaningful.

First, take a look at the process as described by the data with worrying about the specifications. Steve has run the numbers. I ran them also (see attached). The first question to answer is about the stability of the process. I have included the Xbar-R, individuals, process capability and histogram charts. As Steve points out, there has been a process shift that shows up great when you split the control limits. Something has happened to cause that shift. Note that it is a “good” shift – there is less variability in the process. If you can figure out what happened and ensure it continues to happened, you have improved the process considerably. You mention that you don’t want to account for the shift. As your customer, I would want you to – it has improved your process and means less variation coming into my process.
There could very well be a "special cause" at the beginning of the process. With no other notes on the process (such as adjustments, et. al.) one might conclude that it was the special cause of warm-up/start-up. During this time, the process is not stable - but as a customer, you are getting those parts anyway. There maybe a need for additional inspection during that time - but only if there is risk of the parts getting out of specification. Can you improve the process by eliminating the warm-up/start up? Very low probability...

Now look at the issue from specifications. Here, there is no debate. This is process is very capable of meeting specifications. That is seen easily from the high Ppk number. But there is no guarantee of keeping it there because the process is not stable.
On the other hand, if all of the special causes were captured by this sample, then it is very likely to stay there. Of course, a 300 pc sample has very little predictive power over the life of the process, so these single-number capability indices are only academic, at best.

Now look at the issue from the customer's perspective. As a customer, I want a product that is produced from a process that is in statistical control and is capable of meeting specifications. I want your process to be in statistical control because that means that the product I am receiving from you is homogenous - made from a consistent and predictable process. If your process is not in control, I cannot be sure of receiving the same homogeneous product from you in the future. And from your data, your process is not in control and you are missing an opportunity to ensure that the variation remains reduced.
Homogeneous product only comes from a process whose output is random and independent variation. Not all processes (such as tool wear) offer that output, so to expect that from all processes defies physics and statistics.

What is even more important is the basis of the specifications. If the incoming component does not fit the processes it is being used in, the specifications seem to be lacking. If they really need to be tighter for the process to work correctly, quit kidding yourself and the supplier with wide specification, tighten them up and pay the price for the correct material.

The variation displayed is already far within specification. If the variation at this point is still an issue, the specifications are fanciful at best.

So, I tend to agree with the Ford fellow. I don’t know if you need to redo it, but you need to understand what happened because I want to get product from a process that is stable. And, in the end, he is the customer.
I agree, you need to understand why the variation is there, but I also believe all the data you need is there, there just needs to be a better understanding -or explanation - of what occurred in the process at the time.
 

Bev D

Heretical Statistician
Staff member
Super Moderator
#19
Re: Need advise on R chart with PpK of 17.85 for Ford

well there is a difference between the statistical definition of stability and the layman's definition. Statistical stability is a measure of the homogeneity of the process stream (using the traditional approach as outlined by Shewhart). If you see shifts, cycles drifts or other patterns it indicates that the process stream isn't homogenous. The control limits for the subgroup mean (including the IMR chart) are calculated from the within subgroup variation. the theory is that if there is a homogenous stream (this is the essence of Shewhart's Normal bowl) the variation of the subgroup means will be equal to the population standard deviation divided by the square root of the sample size.

If the process stream is not homogenous, the control limits will correctly detect this. but the conclusion isn't necessarily that the process is unstable from a layman's definition. the layman's definition of unstable is unpredictable. When we say that the control chart must be in control (statistically stable) before any Ppk study can be 'meaningful'' we are equating the two definitions.

when we do this we ignore or discount the very real situation where the process stream is not homogenous but it IS predictable and therefore stable by any definition. IF we utilize rational subgrouping and be logical in our chart development (not rigidly stuck with the Shewhart charts just because) we can utilize analysis that demonstrates predictability.

A second collateral consideration of a 'homogenous' process is that when the relative variation from piece to piece (as is the case here - assuming that the sampling was correct and representative) is very tight, the other components of variation (lot to lot, set up to setup, hourly daily and weekly environmental cycles, operator to operator, raw material lot to lot, etc) are no longer 'overwhelmed' by the piece to piece variation and small shifts, cycles and trends can be seen. Remember that Shewhart and Deming were clear that no process was 'stable' for long periods of time. This does NOT mean that the process is unpredictable.

:soap:Frankly further reducing the variation of a process that has a Ppk greater than 17 is a waste (given that the sampling was correct and representative of the true process variation). MAYBE it is valuable to further investigate to confirm that the small non-homogeneity is in fact due to another component of variation has some value in redirecting the Ford guy and others who are stuck in the 'one shoe fits all' mentality of Cp-Ppk. all the more reason to abolish this abomination.
 
M

mlee97

#20
Re: Need advise on R chart with PpK of 17.85 for Ford

Here is the relevant snippet from the MSA (Company rules prevent the publishing of the whole thing.)
View attachment Surface Finish MSA box.xlsx

My SD of both Repeatability and reproducibility are higher than the shift. So you are asking me to investigate a shift that I can't reliably see with my gauge,

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?
 
Last edited by a moderator:
Thread starter Similar threads Forum Replies Date
H MDD Article 12 Labeling for Class IIa Medical Device - Please Advise EU Medical Device Regulations 3
T New role Quality Manager IATF 16949, new company, advise how to get started IATF 16949 - Automotive Quality Systems Standard 2
L ISO 9001 - 2015, is the Quality Manual needed? Please advise ISO 9000, ISO 9001, and ISO 9004 Quality Management Systems Standards 5
C Advise on Oral Spray Pad Printing and Compliance 21 CFR Part 820 - US FDA Quality System Regulations (QSR) 2
F Need advise on Root Cause analysis for printing issue Problem Solving, Root Cause Fault and Failure Analysis 3
D Include as an attachment or make a separate document? Advise on Document Numbering Document Control Systems, Procedures, Forms and Templates 1
P ISO13485 Clause 7.2.1 (d) Additional Requirements - Need advise ISO 13485:2016 - Medical Device Quality Management Systems 2
K Customer Satisfaction - Would you advise to investigate the trend or the incidence? Customer Complaints 9
D CE Mark advise sought for medical diagnostic Software CE Marking (Conformité Européene) / CB Scheme 4
C Need advise to make judgement to take job offer Career and Occupation Discussions 9
F What training and certification path would you advise me to take? Professional Certifications and Degrees 13
V Can anyone advise a method to research 'Measurement System' for testing a tire? Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 3
S I have been assigned to implement ISO 14001 - Please Advise ISO 14001:2015 Specific Discussions 11
E Digital Height Gauge Gage R&R - Need Advise and Help Quick Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 5
J What is Quality for me?... How to do it? My advise for Teachers Philosophy, Gurus, Innovation and Evolution 7
G Need Advise on upcoming ISO 9001:2000 Surveillance ISO 9000, ISO 9001, and ISO 9004 Quality Management Systems Standards 7
M Advise: Cost and suggestion of APQP planning software APQP and PPAP 2
D ISO/TS 16949 Books - Please advise on the best books available IATF 16949 - Automotive Quality Systems Standard 2
M Internal Document Numbering System - Please review and advise Document Control Systems, Procedures, Forms and Templates 7
H Need advise on formats to use when writing procedures Document Control Systems, Procedures, Forms and Templates 14
U Can any body advise on ISO 9001:2000? Textile Company ISO 9000, ISO 9001, and ISO 9004 Quality Management Systems Standards 6
Howard Atkins Advise on applicability QS-9000 - American Automotive Manufacturers Standard 1
A SPC procedure / flow chart Statistical Analysis Tools, Techniques and SPC 2
optomist1 Paynter Chart Example APQP and PPAP 1
A Fabric roll inspection - What type of Control Chart to use? Inspection, Prints (Drawings), Testing, Sampling and Related Topics 2
Z Putting back excluded rows/data points in a control chart Using Minitab Software 0
B Control chart and sample time Statistical Analysis Tools, Techniques and SPC 1
J Control chart for huge sample size Statistical Analysis Tools, Techniques and SPC 9
R Capability analysis - What is going on this chart? Manufacturing and Related Processes 15
S How MR (Moving Range) in Z-MR chart is calculated Using Minitab Software 5
M What is "setup control chart"? Statistical Analysis Tools, Techniques and SPC 2
D Virtual Documentation via Master Flow Chart? Process Maps, Process Mapping and Turtle Diagrams 4
A Redesigning our process flow chart, PFMEA and Control Plan Process Maps, Process Mapping and Turtle Diagrams 4
K Vigilance process flow chart EU Medical Device Regulations 1
D Control chart applicable? Percentage of compliance with a standard Statistical Analysis Tools, Techniques and SPC 2
G I-chart form for GM GP-12 Customer and Company Specific Requirements 2
D Link between FMEA, flow chart and control plan FMEA and Control Plans 10
B How to do a control chart - Machining product that requires flatness control Capability, Accuracy and Stability - Processes, Machines, etc. 1
D China - Where I can get NMPA complete structure/organogram chart and GB/YY format list China Medical Device Regulations 1
K Turtle diagram or process interaction chart - Making it easier for an auditor Process Maps, Process Mapping and Turtle Diagrams 23
D Should "Waste" be included as Output in SIPOC Chart? Process Maps, Process Mapping and Turtle Diagrams 8
S Assembly House Process Chart ISO 9000, ISO 9001, and ISO 9004 Quality Management Systems Standards 4
K Why calculate LCL in P-chart? Statistical Analysis Tools, Techniques and SPC 4
B GM GP-12 Pre-Launch - They asking for a I chart Statistical Analysis Tools, Techniques and SPC 2
bobdoering SPC Charting frequency calculator for X hi/lo chart Imported Legacy Blogs 0
K Operating characteristic (OC) curve S chart Statistical Analysis Tools, Techniques and SPC 1
K What is the Beta RISK for S chart OF XBAR-S chart Statistical Analysis Tools, Techniques and SPC 0
S Understanding control chart and measurement capability Statistical Analysis Tools, Techniques and SPC 2
F Appropriate Control Chart Statistical Analysis Tools, Techniques and SPC 5
S used excel formula calculated the ARL's with Rule1&Rule2 for Shewhart control chart Statistical Analysis Tools, Techniques and SPC 0

Similar threads

Top Bottom