MSA - Should I compare my GRR to tolerance or to my sample?

T

Tenko

#1
Hello everyone,

I'll expose my problem: I do an internship in a French Medium Sized Company and I have to set up the capability of all the measurement gages (around 300). I have based my procedure to the 4th MSA edition. I have carried on measurement tests with 3 appraisers 16 parts and 3 trials. My problem is, when I start my analysis, should I compare my GRR to the tolerance or to the variation of pieces of my sample? I read in the forum that comparing to the tolerance is for inspection but I did't see that in the 4th MSA edition.

I have an ISO audit in 4 month to check how i perform MSA method so I won't be able to prove it if I choose this method. Moreover, my boss told me that on the short term, he needs to know what instrument can really distinct good and bad parts. The long term will be the SPC.

I'm really sorry for my english, I'm french and I got some difficulties.
 
Elsmar Forum Sponsor

Miner

Forum Moderator
Staff member
Admin
#2
The metric and criteria depend upon the intended usage of the gauge.

Inspection gauges should be evaluated using %Tolerance.

Gauges used for SPC, for capability studies, or any type of statistical studies such as DOEs should be evaluated using %Process Variation (if historical data on process variation is available) or %Study Variation or ndc (if no historical data is available) is used. If you use % Study Variation or ndc, the selection of samples is critical and must represent the actual process variation. Samples selected to represent the full range of tolerance, will likely inflate your results higher than they truely are, and consecutive samples that do not exhibit the full range of actual process variation will shrink your results giving percentages that are smaller than reality.
 
T

Tenko

#3
Thank you Miner,

You cleared up to me the ideas. Since I need to make capabilities, I chose to use the %Study Variation. Consequently, I will determine if an measurement system is capable (I don't know if it's said like that in english) with the %R&R. I don't understand what will the ndc bring for new information. For exemple, if I have 15% for %R&R and ndc=3 (and then <5), I will conclude the gage is aceptable regarding the %R&R but the ndc won't be a decision-making criterion.
 

Miner

Forum Moderator
Staff member
Admin
#4
Two comments:
  1. Do not use the study variation to assess capability. The samples size is too small for a valid capability. You need at least 100 units to get reasonable confidence limits around the Cpk. I may have misunderstood your intent. If you meant you need a passing MSA before performing a capability study, you are correct.
  2. %Study Variation and ndc are related metrics. The decision that you make for one will also be made for the other unless you are in a marginal situation where one barely passes and the other barely fails.
 
T

Tenko

#5
Hi,

1) I guess I was not clear about my goal, I'm sorry. The purpose of my internship is to carry on capabilities of measuring instruments. The MSA 4th edition is my reference. The problem is that my company can't make around 100 parts for my experiments. I thought that if I assessed the capability on 16 parts using the %GRR (compared to the %study variation) and the bias as decision-making variable, it would be enough to consider as an instrument as able to be used for SPC and inspection.

2) This point is okay.

Thank you very much for spending time for me.
 

Miner

Forum Moderator
Staff member
Admin
#6
The confidence intervals around Cpk are extremely wide with small sample sizes. You would get a number, but reallity could be very far from that number.
 
S

sc00by

#7
If you intend to use this gauge also for inspection then you should use Tolerance based GRR Tolerance with full tolerance range of your part feature.
Using extended range that slightly exceeds tolerance is a good practice as well so you can make sure that Gauge system is correctly able to pick up parts that are outside of the tolerance and flag them as bad ones.
 
T

Tenko

#8
Scooby, it is said in MSA 4th edition, as you explained, that using tolerance "when the measurement system is to be used to sort the process, AND the process has a Pp<1.0" (p121). I understand that if don't have any historical Pp, using tolerance is the wrong choice for me. The priority for me is to declare if the instruments used by the operators after manufacturing (inspection) but it requires 100% control if no SPC is carried on. It would slow the production and my boss doesn't want that. I guess that if my instruments are acceptable for SPC and statistical studies (%study variation), they will be aceptable for inspection. According to me, the capability of instruments based on %study variation is the start any capability process, SPC. Otherwise these studies, for exemple SPC, can't be reliable. Please correct me if i'm wrong.

Miner, I explained my boss that around 100 parts were required to evaluate the capability of my instrument with %study variation. It would cost too much money and time for my company. Further to this, I had an idea to simplify the sampling: I'll take around 15 parts, I check capability, if %R&R is lower criterion I declare this gage capable directly. I just made the hypothesis that the calculated part variation is lower than real part variation due to the process. Is it acceptable?
 
S

sc00by

#9
To my understanding you have no MSA done on these measurement systems and you would like to validate them as for a start, you would like to use it in production (or as I assume it is already being used in production) for SPC as well as inspection.
Few questions to you to ask yourself:
- you have no information whether gauge systems are valid or not for the features you are going to measure, your process variation may be low and so performing Study Variation based GRR will provide you only information about measurement system for this narrow spread only, now question is how are you going to pick up any changes within the process if you don't know how measurement system works/behaves on the whole spread of your tolerance?
- you don't have any historical Pp and you don't have any current Pp then why do you automatically assume your Pp>1.0 (or whatever limits your company has) and you would like to skip tolerance based GRR? Shouldn't you take into an account worst case scenario?
- your Pp depends on your sample rate and your sample size but unless you perform Pp on a continuous basis you are not able to say if there is a shift in the process over longer period of time i.e. shift due to machine changeover/retooling etc - SPC should have picked this up but how can you say if you haven't tested measurement system for shifted part spread?
- "when the measurement system is to be used to sort the process, AND the process has a Pp<1.0" (p121) - It is perfectly understandable that any Manager responsible for production will be against creating bottlenecks but in case of a quality risk such decisions have to be made and so this is in fact 100% inspection that should be considered in worst case scenario, don't you think?

Your assumption that "if my instruments are acceptable for SPC and statistical studies (%study variation), they will be acceptable for inspection" is wrong.

Have a think about benefits of the study if you are going to have tolerance wide (or slightly wider) batch of parts:
- business can keep the samples to control gauge linearity (i.e. in case of gauge measurement characteristics change - I don't know what kind of gauge systems you use but this is very valid for any air mandrel gauges, for split ball probes and many more)
- business can keep the samples to control gauge correlation (i.e in case of having multiple gauges for the same feature)
- business can keep the samples to maintain operator quality performance (i.e. new operator on the line controlling the process and performing measurements)
- this batch is small

Question to you: you would like to calculate Process Capability Pp and you would require a batch of parts for that, I'm trying to understand why it would be expensive? This batch consists of a standard production parts that are being assessed only for the period of the study and you don't need to keep them any longer afterwards.

Regards,
sc00by
 
N

neyveliravi

#10
Hi Tenko..
Sorry for inturupting your discussion. If you want to prove the instrument and its handling method by three operators GR&R is ok with 10 samples, but remember the samples selected to be in the full range of your specification.
But, If you want to prove the process with SPC, 15 No's is not enough. If you want to go for control limits, as per the SPC Manual, min of 25 sub groups required.
 
Thread starter Similar threads Forum Replies Date
B For a measuring instrument, should MSA be done first or the Capability studies? Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 4
M MSA - According to the formula cg how much tolerance should be put in place? Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 1
M MSA - What method should we use for the stability of hardness? Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 1
Rameshwar25 Which tool should be used first - MSA or SPC Quality Tools, Improvement and Analysis 30
A MSA applicability for Hardness Testers - What type of MSA should be done? Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 6
G MSA Studies on heat shock tests - Should we just be calibrating the oven? Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 1
Nihls When the MSA results show no operator influence Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 3
S MSA for attribute relation gage Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 2
Nihls MSA Study Type 1 (CMM) Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 1
M MSA Study Type 1 not capable. We are at the limit. And manufacturing wants to continue producing. Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 7
I Overwhelmed with attribute MSA requirement for visual inspection IATF 16949 - Automotive Quality Systems Standard 8
H Need MSA 4th ed. compliant attribute MSA template General Measurement Device and Calibration Topics 4
P MSA for titration Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 6
P MSA - what exactly mean "system" and master sample Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 2
P MSA study for visual system with artifical inteligence Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 2
D MSA strategy Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 4
K Updated MSA with the Annual PPAP validations APQP and PPAP 8
E MSA for Push and Pull test Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 9
6 MSA for Calculated Clearance Dimension Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 1
Ron Rompen MSA on automated measurement system - Multiple Step Vision System Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 3
G MSA check list to audit IATF 16949 - Automotive Quality Systems Standard 10
R MSA for ATE (Automatic Test Equipment Embedded Software) Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 9
Bev D MSA Tools Misc. Quality Assurance and Business Systems Related Topics 3
R MSA studies for Heat Treatment Equipment Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 1
C Industrial scales and MSA (IATF 16949 requirement 7.1.5.1.1) IATF 16949 - Automotive Quality Systems Standard 30
M Difference between MSA and MSE? General Measurement Device and Calibration Topics 1
M Test method validation - Is MSA (MSA1, MSA2, MSA3 and linearity) a good solution? Medical Device and FDA Regulations and Standards News 1
A MSA results differences - Supplier results vs. My results Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 1
R Preliminary MSA and Sockets' correlation Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 0
I MSA requirement for 5 Micrometers + CP changes need customer approval? IATF 16949 - Automotive Quality Systems Standard 2
T Correct MSA study for an automated camera system which makes attribute inspection Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 1
Bev D MSA Misc. Quality Assurance and Business Systems Related Topics 46
Emran.mi Measurement system analysis - Can you help me about implementation MSA for CMM device Manufacturing and Related Processes 2
N MSA Study for a Leak Testing Device Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 6
N Justifications for not performing MSA (Measurement System Analysis) Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 6
D Do I need part variation while doing Destructive Variable Gage R&R MSA study Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 19
Prashant G MSA Study - AS 9100 and and our customer want us to do MSA study for their parts dimension Reliability Analysis - Predictions, Testing and Standards 4
D Compression Spring Force MSA Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 5
E MSA Study on MTS dynamic rate measurements Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 5
A The FILE recommended by AIAG MSA 4ed for LINEARITY Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 0
S IATF 16949 Audit MSA IATF 16949 - Automotive Quality Systems Standard 10
C Do Calibration Reference Standards require an MSA study? Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 2
A MSA When an Instrument Measures More than One Parameter Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 7
G MSA Type 1 - Cg and Cgk ranges Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 1
O MSA Study for a steel ruler Manufacturing and Related Processes 6
P IATF 16949 MSA Studies for CMM machines IATF 16949 - Automotive Quality Systems Standard 3
B 7.1.5.1.1 MSA for inline dimension vision inspection equipment IATF 16949 - Automotive Quality Systems Standard 5
G MSA on a Counting System Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 2
C AIAG's IATF 16949 section 7.1.5.1.1 - Torque MSA Requirement Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 10
S IATF 16949 7.1.5.1.1 Measurement System Analysis (MSA - Die casting) IATF 16949 - Automotive Quality Systems Standard 5

Similar threads

Top Bottom