Poor GR&R results due to Low Part Variation

E

eturkan

#11
Your gage discrimination is not good enough.

I do deal with parts with +/- .010" parts and we use micrometers which has 0.00005" discrimination. Our Gage R&R is less than 10%.

Look at the minitab results.It is not just about GRR%

-You only have 3 ranges..
-Less than 50% points are out of control in XR chart which means your measurement deviation is high.
 

Attachments

Last edited by a moderator:
Elsmar Forum Sponsor

Ninja

Looking for Reality
Staff member
Super Moderator
#12
Keep in mind through all of the pressure you're under that GRR means different things to different people and organizations.

"True" Anova GRR is one thing, what is acceptable to a customer may be a bit different.

Many customers (including automotive) will accept GRR as a % of spec tolerance...have you tried this yet?

At the same time, have you asked for guidance from the customer? Find a person who has a 'reasonable streak' in them and explain the situation. While they are insistent on GRR, they likely have allowed exceptions before...they know what exceptions may have been made whereas you don't. Look for guidance from both quality and production engineering.

As demanding as automotive can be (much of it valid), they also run production lines in the real world...In my experience, the more senior folks tend to have a pretty good grip on what makes sense and can recognize an unrealistic demand and work with you to find a mutually acceptable approach.
Look for the older guys...they've been through the fire drill a few times.
 
X

xavierthomas

#13
Thanks Ninja.

I have found most automotive customers are realistic and many will accept the %Tolerance method but two demand the ANOVA method and basically have no clue what the part is, its function nor do they care that they haven't had a reject from us ever.

We are a small supplier to them and don't get much face time, yet this product is our livelihood. We will improvise, adapt and overcome!
 
E

elger

#14
If calculating the GRR from the Tolerance does not improve results, the source may be that the samples selected have too little variation. Selecting samples from different runs may help.
A similar situation was found with a stamping process when the parts used for the GRR Study had been taken for a production run with a Cpk of 13.
If you will send me the data, I can compare the result to that study.
 

Bev D

Heretical Statistician
Staff member
Super Moderator
#15
OK so this is for a quick floor check gage - these kinds of details are always helpful so we can provide advice that is on-point. SPC is not a concern here.

And in the grand scheme of things your specs aren't really a concern with this gage either as it isn't the final acceptance gage. you may produce scrap (or rework) but won't ship out of tolerance parts...

I have plotted the data on a Youden plot. I added a spec box so you can see the relative variation of the gage in comparison to the tolerance. (which is really pretty small). I think you are good to go. (I did have to randomly place the box as you didnt' state the actual limits only the tolerance. none the less it is really the comparison that matters. you can adjust the box to the correct LSL/USL in the attached spreadsheet.

I don't know about how reasonable your customer is, but I've always found that repeated discussion (on different issues) can wear them down to a state of reasonableness when your evidence is compelling that you are not trying to get away with something.


...As Deming said: "plot the bloody dots"
...As Ellis Ott said: "Plot your data, look at your data and THINK about your data"
 

Attachments

E

elger

#17
Using the tolerance to calculate GRR, the GRR is 0.42 and the ndc is very high. However, there are two points outside the control limt on the range chart.

You need to either select samples over a greater production time or use the tolerance method to calculate GRR.

elger
 
X

xavierthomas

#18
It appears my biggest challenge is to convince a few Statistibots that they need to look beyond the numbers at times and consider the big picture.
 

Bev D

Heretical Statistician
Staff member
Super Moderator
#19
statistibot! I love it.

too often people look only at the statistical output without any insight ot the structure of the test or the reality of the data. sad but true.
 
C

Coleman Donnelly

#20
The ONLY time it is permissable to cherry pick parts or to deliberately select parts out os specification is when your gauge is used as an inspection device and the only metric is %Tolerance.

If your gauge is used for SPC or capability studies it is never acceptable to select anything other than random parts that represent the process. Doing so is no different than deliberately falsifying test results. See my blog on MSA, specifically the section on Part Selection.
Sorry - I?m just now getting back to this :)

I disagree with this entirely... well sort of...

I think that it is absolutely necessary to cherry pick your samples! I think that your MSA should be an iterative process and the first iteration of that process should include your ability to check non-conforming product at both ends of the spectrum... I have seen systems that are very capable of inspecting nominal parts but error stacks up with deviation from nominal.

Secondly I think when you are trying to asses NDC for purposes of SPC and process control, it is important to ensure that your samples span the distribution of your process. Grabbing 5 random samples may give you a false failure for NDC because all 5 of the samples you happen to grab fall within +/- 1 sigma of your process mean. Spreading this out to +/-3 sigma should help here, but grabbing samples arbitrarily is relying on luck of the draw...
 
Thread starter Similar threads Forum Replies Date
Y Poor Gage R&R Results - High R&R result and poor ndc value Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 5
M Poor Gage R&R results (30%RR to 80%RR) - What Next? Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 6
S Cost of Poor Quality (CoPQ) - Scrap element of calculation includes lost margin? Misc. Quality Assurance and Business Systems Related Topics 7
qualprod Addressing poor commitment in people Misc. Quality Assurance and Business Systems Related Topics 23
qualprod Best criteria to measure Corrective Action effectiveness - Poor Maintenance ISO 9000, ISO 9001, and ISO 9004 Quality Management Systems Standards 10
C What is your Cost of Poor Quality (COPQ)? Misc. Quality Assurance and Business Systems Related Topics 8
V For screening a design, can we live with the low/poor predicted R-Sq.? Statistical Analysis Tools, Techniques and SPC 3
M Reducing Cost of Poor Quality - Your opinion requested Quality Manager and Management Related Issues 14
M Contractors with poor Quality Records ISO 9000, ISO 9001, and ISO 9004 Quality Management Systems Standards 9
M Customer Satisfaction Survey - Poor Response ISO 9000, ISO 9001, and ISO 9004 Quality Management Systems Standards 11
S Difference between COQ (Cost of Quality) vs COPQ (Cost of Poor Quality)? Misc. Quality Assurance and Business Systems Related Topics 8
M Which metric is better for Cost of Poor Quality (COPQ)? Quality Tools, Improvement and Analysis 17
N Improving Poor Communication with Customers Customer Complaints 6
P Smart or Sexy (CAR) - Poor Root Cause Analysis and Poor Effectiveness Check Nonconformance and Corrective Action 8
Y FDA News bulletin: It's the FDA's fault! FDA at Fault for Poor Quality Devices US Food and Drug Administration (FDA) 6
F Poor Containment - Supplier Nonconformance Problem Solving, Root Cause Fault and Failure Analysis 7
R Development Rejection Costs in COPQ (Cost of Poor Quality) Calculations Misc. Quality Assurance and Business Systems Related Topics 6
AnaMariaVR2 Poor Writing Skills Overshadow Good Content, Credibility & Revenues Records and Data - Quality, Legal and Other Evidence 19
1 Cost of Poor Quality form or template wanted Quality Tools, Improvement and Analysis 1
S COPQ (Cost of Poor Quality) Vs PONC (Price of Non Conformance) - Explanation wanted Quality Manager and Management Related Issues 5
P Measuring Costs of Poor Quality Quality Manager and Management Related Issues 2
A Chaotic Industry, Poor Service From Almost Everyone Coffee Break and Water Cooler Discussions 12
M Cost of Poor Quality Reports IATF 16949 - Automotive Quality Systems Standard 15
E Poor Manufacturing Planning fixes Misc. Quality Assurance and Business Systems Related Topics 4
K Generic worksheet for calculating Supplier Chargeback and Cost of Poor Quality (COPQ) Supplier Quality Assurance and other Supplier Issues 2
N Personnel COPQ (Cost of Poor Quality) Quality Manager and Management Related Issues 6
P Evidence of Supplier Corrective Action - Poor Responses by Suppliers ISO 9000, ISO 9001, and ISO 9004 Quality Management Systems Standards 6
Marc Poor Hungry Kitty Funny Stuff - Jokes and Humour 8
S COPQ (Cost of Poor Quality) - Can we add Premium Freight? Misc. Quality Assurance and Business Systems Related Topics 6
S COPQ - Cost of Poor Quality - COPQ course for non-financial people Misc. Quality Assurance and Business Systems Related Topics 1
N Identification of COPQ (Cost of Poor Quality) aspects Quality Tools, Improvement and Analysis 2
bobdoering Consumer Reports: Poor Man's Opera Coffee Break and Water Cooler Discussions 0
S Examples of major problems caused by Poor document control Document Control Systems, Procedures, Forms and Templates 14
T Poor Capability - Electric torque screwdriver Capability, Accuracy and Stability - Processes, Machines, etc. 2
H Cost of Poor Quality (COPQ) - All returned/defective parts (plastic) are recycled Training - Internal, External, Online and Distance Learning 7
Manix Cost of Poor Quality (COPQ) and Reputation Quality Tools, Improvement and Analysis 25
Z Consequences for Poor Performance based on a Monthly Scorecard - Actions? Supplier Quality Assurance and other Supplier Issues 15
I High COPQ (Cost Of Poor Quality) from Our Quality Department IATF 16949 - Automotive Quality Systems Standard 7
N Poor control of testing chemicals ? which ISO 13485/9001 clause does it contravene? ISO 13485:2016 - Medical Device Quality Management Systems 8
S New Auditor wants Mould tool repair bill included in Cost of Poor Quality! IATF 16949 - Automotive Quality Systems Standard 18
Z Creating and Maintaining a Top Ten Poor Quality Supplier Performer List Supplier Quality Assurance and other Supplier Issues 16
S Operator Capability - Gage R&R Poor Repeatability Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 8
M One weakness of audit non-conformance reports is the poor writing Miscellaneous Environmental Standards and EMS Related Discussions 10
S Is there any standard for "Cost of Poor Quality" (COPQ) by ISO Various Other Specifications, Standards, and related Requirements 5
J Cost of Poor Quality (COPQ) - Iceberg picture Quality Tools, Improvement and Analysis 4
D Scary stories - the REAL consequences of poor quality Misc. Quality Assurance and Business Systems Related Topics 3
B Machine failure defects due to no/poor maintenance for FMEA - Looking for examples FMEA and Control Plans 6
M Need input - Potential poor external auditor? My first ISO 14001 registration audit ISO 14001:2015 Specific Discussions 4
P Acronym COPQ (Cost of Poor Quality) and COQ (Cost of Quality) - Differences Definitions, Acronyms, Abbreviations and Interpretations Listed Alphabetically 26
L TS 16949 Clause 5.6.1.1 - Cost of "POOR" Quality vs. Cost of Quality IATF 16949 - Automotive Quality Systems Standard 4

Similar threads

Top Bottom