Hi All
when can we pass from a 50% confidence curve to a 95% or 97.5% curve while constructing and testing for Wohler Curve ?
Dear Farzad Torabi ,
It totally depends on how your tests data is following normal or log normal distribution and how small is the standard deviation and variance. If you analyzed a tests data and calculated the level of confidence 50%, you literally need to perform more test and see how well your data will fit within a normal distribution. In my opinion this is the only way of increasing the level confidence interval. Please have look at the attached document.
Hope this answer helps.
But the problem is that in calculations, the confidence level is the input. How can I calculate it as an output ?
You can use Excel, "Data Analysis" function and use "Descriptive Statistics" option for calculating whatever confidence level you are interested in.
Incase the failure distribution follows constant or random failure behavior, you can use Chi-Sq relationship between MTBF, total test time/hrs/cycles, number of failures and CL. Keeping others fixed and varying total test time and no new failures, your CL will increase or vary. Hope this helps.
- Badges
- Science topic
- Similar topics
- Engineering
- Mechanical Engineering