establish knock down factor with R=99% and confidence level=95% for fatigue mean data? 1

[JBlack68]

To all

What is the knock down factor I need to apply to a mean stress to get reliability = 99% with a confidence level = 95%?
I am mixing up my basic stat an need a refresher on this one!

Assume I have a sample size n = 1 (only going to test the system once!)

Thanks
Regards

 

[SWComposites]

Suggest reading the MMPDS data analysis chapter.

But with one specimen, you have insufficient data to estimate the fatigue scatter, so no way to calculate statistical lower bound.

 

[JBlack68]

I am reading it and using a set of data/curve from MMPDS as a 1st pass
I am trying to see what (statistical) parameters I need to know if I have 2 points (log-log)r and "correct" for say reliability = 99% with a confidence level = 95%.
Realised that stating n=1 was misleading. I meant to state that I will only test 1-off structure.

Basic stat states standard error = sd/sqrt(sd = standard deviation, n is the number of test specimen used for SN data gathering). This is stated in the MMPDS data but never in generic textbook

so if I have SE = Standard Error of log(N)
correction of the curve would be
Log (N99%) = Log(Nmean)+SE*SD where SD = -3

I think I am getting confused between reliability and confidence level