How to do an MSA Study on Stopwatches

L

lauriewebb

Help! Can someone tell me how to do a Gage R&R study on stopwatches?
 
L

lauriewebb

We are a TS16949 company. We manufacture brake pads. During our audit in November the auditor wrote us up for not having MSA on a stopwatch that was referenced on our control plan that we use to measure conveyor speed in one of our processes. We also use a stopwatch to measure resin gel time and resin flow which is also listed on our control plan. We conducted initial GR&R using NIST time signal and Direct Comparison Method (Digital DUT). Characteristic of time as one minute interval. Performed as a 3 x 3 x10. Results > 30%. ANOVA shows that the characteristic needs more variation.
 
I

iamtroll

We are a TS16949 company. We manufacture brake pads. During our audit in November the auditor wrote us up for not having MSA on a stopwatch that was referenced on our control plan that we use to measure conveyor speed in one of our processes. We also use a stopwatch to measure resin gel time and resin flow which is also listed on our control plan. We conducted initial GR&R using NIST time signal and Direct Comparison Method (Digital DUT). Characteristic of time as one minute interval. Performed as a 3 x 3 x10. Results > 30%. ANOVA shows that the characteristic needs more variation.
There is not going to be much variation in the NIST time signal. I would suggest that you experiment with introducing variation by setting several time intervals as your characteristic. Although this is artificial, I don't know of another way to create variation in time. You just need to see what levels the ANOVA will accept. If you had a machined part with a standard of 1.000" and a tolerance of +-.005" you might expect a process variation of +-.002" among the ten parts. This would probably give enough variation to evaluate a standard micrometer. So if you have a characteristic of 60 seconds, you would need to experiment with a group of known different time intervals such as 59, 60, 61, 63, etc which would give you your variation "between parts" so that you can then evaluate how well the watch and the watch operator can agree on each different interval. It might be that seconds is not enough variation you might need ranges of 5 or 10 seconds. On the other hand your system might have enough variation with fractions of seconds, depending on the accuracy of your stopwatch and the reaction time of your operators. This is kind of general, I hope that it helps.
 

AndyN

Moved On
We are a TS16949 company. We manufacture brake pads. During our audit in November the auditor wrote us up for not having MSA on a stopwatch that was referenced on our control plan that we use to measure conveyor speed in one of our processes. We also use a stopwatch to measure resin gel time and resin flow which is also listed on our control plan. We conducted initial GR&R using NIST time signal and Direct Comparison Method (Digital DUT). Characteristic of time as one minute interval. Performed as a 3 x 3 x10. Results > 30%. ANOVA shows that the characteristic needs more variation.

I'm having a tough time understanding this one. I (and some of my colleagues) have never seen an MSA done on a stopwatch and, more's the point, the fact that the auditor wrote it up, doesn't necessarily, make it an issue...

I'd like to know how this timing makes an effect on the process - isn't there a controller for the belt speed? Isn't what you dial in more important than any secondary check?
 

Miner

Forum Moderator
Leader
Admin
I agree with Andy on this. But, on the chance that you must perform an MSA, consider that when using a stopwatch you have two possible scenarios that would impact your MSA:
  1. Event with warning: Example: Timing a runner on a track. You can see the runner approach the finish line and will unconciously coordinate the tripping of the stopwatch.
  2. Event without warning: Example: Timing a chemical reaction. When the reaction is complete, the chemical instantaneously changes color. You cannot see the reaction nearing completion. It happens suddenly, and the triggering of the stopwatch is dependent on both attention level and reflexes.
The latter scenario is likely to have greater bias and R&R than the former.

Take this into consideration when planning your MSA because one MSA may not cover all of your situations.
 
Top Bottom