# Determining Uncertainty from Gage R&R

#### RichardJDrexler

##### Registered
Folks,

Let me preface this with the fact that I am not six sigma certified.

Process:
1. Weigh tin (Pre-weight)
2. Add solvent w/ contaminate to tin
3. Evaporate solvent; leaving contaminate in tin
4. Weigh tin again (Post-weight)
5. Delta of weights is weight of contaminate

I am trying to determine the uncertainty of a measurement taken with a balance. Given the nature of the work it is extremely important to over estimate the weight of the contaminate rather than under estimate it.

Results from Gage R&R:
Equipment Variation: 0.00003g
Appraiser Variation: 0.00001g
Interaction Variation: 0.00001g

Given my variation results, is it safe to call the sum of my variations the "error" of a possible measurement?

Right now the assumption is that the balance is accurate down to 0.00001g since that is the resolution displayed on the screen. So if the pre-weight measured 2.00000g and the post-weight measured 2.00001g we would say there is 0.00001g of contaminate. I don't think this is accurate given how much a reading on the balance fluctuates after it decides the reading is acceptable.

My thought is that we need an error around the measurement to be conservative. So if the pre-weight measured 2.00000 +/- 0.00005g and the post-weight measured 2.00001g +/- 0.00005g then the amount of contaminate I could have in the worst case is actually: 2.00006g - 1.99995g = 0.00011g. Therefore to be conservative I should use 0.0011g as my result. Additionally, this means I would never be able to claim that I have a contaminate with a weight less than 0.00010g.

I have attached the spreadsheet with my work.

Cheers.
Rich

#### Attachments

• 42.2 KB Views: 139

Staff member