Long OCI's and Wear Numbers

[Gary Allan]

Okay ...really concentrating (not easy, at times) ...I see what you're saying ...but any and all of this falls into a fairly narrow range and conforms more to a linear profile than non-linear. I think you could argue from the other end of it with more weight.

 

[1sttruck]

k miles
Fe Increment
1 10
2 2
3 3
4 7
5 7
6 6
7 1
8 14
9 7
10 -3

 

[dcoyne78]

You seem to be putting a lot of weight on a single study. Keep in mind that the reproducibility for FE is +/- 3 PPM. So your increment idea may sound nice, but you are reading too much into each sample at 1000 miles. You might also want to look at the Amsoil test done in the same study, I think your result will look different. The person who did the study suggested that the engine may not have been completely broken in during the Mobil 1 phase of the study. Or Amsoil is much better oil. In either case I don't see that 3000 mile oil changes are warranted. If you only look at the Mobil 1 part of the study, then you could argue that 6000 mile changes would be safer than 9000 miles. If you compare 3000 miles at 15 PPM Fe vs 6000 miles at 35 PPM and factor in the fact that reproducibility is +/- 3 PPM, then I am not convinced that the 30 PPM Fe over 2 3000 mile ocis is a significant improvement over 35 PPM for 6000 miles (or 36 PPM over 7000 miles).

Dennis

 

[1sttruck]

You are reading too much into the average Fe over miles driven :^)

If you do SPC (statistical process control) on a process you react to each sample, not the average of all samples up to each point. The reason for this is that as stated before averaging is designed to minimize variation. The average is used for multiple samples at a data point, and to develop limits to react to. If the question is 'at which 1000 mile increment should I change the oil', why do you want to ignore the more sentitive data for each 1000 mile increment ?