I am running a GR&R on leak test equipment. I have two leak test equipment that test parts exactly the same:lol: or should be testing them the same. My question is when running a GR&R on these two machines how much of a deviation from one machine to the other should I expect.

The most that I have seen so far running stdev on excel was 7.35 and the least was 1.48 stdev. If any one could shed a little light on this it would be appreciated.

:thanks:

We really need more information to assist you.

What type of leak test equipment is this? Mass spec based using trace gases? In a sealed canister, a hand-held or robotic probe? Stationary or moving?

At a high level, you could perform a hypothesis test to determine whether the difference in results is statistically significant. The variation that you use in the equation should be the measurement variation, not the product variation.

However, before you do this, ensure that the gages have been calibrated, then determine whether each gage individually has an acceptable MSA. If they do not, correct the gage first.

We leak test engine blocks and heads that sit on a table and are transfered or rotated into 3 stationary stations. Each testing a different cavity. I have ran the R&R on them and they pass as an individual gage but when I do a comparision between the two there is a big difference.

We use the same parts and test them the same way on the equipment.

I have never really looked into the difference between the two machines we always were concerned with the individual gage.

Have the two gages been calibrated against a standard?

It would also help if you provide information about the type of leak test equipment you are using. Mass-spec, pressure decay? The more information that you provide, the less guessing we have to do to assist you.

Well they are calibrated to a part that is closest to zero on each circut in the test.

The test is pressure decay on engine heads and blocks. We test the leakage by filling up the cavity (low pressure oil, water jacket and high pressure oil) then letting the cavity stabalize for 30 secs then test the air loss.

When running the GR&R according to supplier specs is 10% but the equipment was never able to hold that so they accepted a deviation or a max allowable in each circit. We are within the max allowable but I wans't sure if I should be concerned with the variation that I see from the first leak tester to the second leak tester. We have two leak testers that check the parts exactly the same we just have two for more throughput.

Well they are calibrated to a part that is closest to zero on each circut in the test.

The test is pressure decay on engine heads and blocks. We test the leakage by filling up the cavity (low pressure oil, water jacket and high pressure oil) then letting the cavity stabalize for 30 secs then test the air loss.

When running the GR&R according to supplier specs is 10% but the equipment was never able to hold that so they accepted a deviation or a max allowable in each circit. We are within the max allowable but I wans't sure if I should be concerned with the variation that I see from the first leak tester to the second leak tester. We have two leak testers that check the parts exactly the same we just have two for more throughput.

Try the hypothesis test for significance between the two machines using the measurement variation determined by the MSA.

You may want to run an experiment where you vary the stabilization time to check the effect on the MSA.

If you can attach your data, we can assist in the analysis/interpretation.