Tension Testing "Sampling Plan" 3

[roscoe1]

I am testing a braze pre-form on Inconel 600 base metal. I will be conducting a series of tension test to E8.

The test will be utlimate tensile after melt, after diffusion, which will be compared to bars which are untouched.

I am looking for the min. amount bars per group which would give me a good avg. sampling plan.

Would someone be kind enough to comment.

Thanks John

 

[stel8]

If you only need to compare those two or three treatments (untouched, after melt, and after diffusion) ANOVA seems like the best method to find out if there is any significant difference between treatments. The larger the number of tests (if you can afford it), the better.

Once the experiment is done, and you have the data, for future testing there is a way to determine the sample size based on the variance of each treatment done in the initial experiment.

 

[cab1990]

try MIL-HNDBK-5 for a guideline on what it take to produce mechanical properties with various confidence levels

 

[GregLocock]

Check out hypothesis testing in any statistics book.

Basically, you need to find the mean and standard deviation for your two groups, and then apply an appropriate test to see if they are different, to whatever confidence level you need.

http://www.itl.nist.gov/div898/handbook/prc/section3/prc31.htm

Explains how to determine whether the means of the two groups are significantly different.

Or see example 10.9 in Schaum's Outline of Statistics.



Cheers

Greg Locock