Weibull parameters why not mean and deviation

[Gerard Westhoff]

Dear all,

I'm a student mathematics at Leiden University. And very much interested in statistics. Now, the other day I had to simulate (with Matlab) an electronic system in which crucial components had a Weibull-distribution. Alas, the Matlab function WEIBRND accepted only parameters aplha/etha and bèta. NOT the meanlifetime and deviation, which were given (rather than alpha and bèta). So, I got stuck - how to solve alpha and bèta from the formulas mentioned in http://www.weibull.com/SystemRelWeb/the_weibull_distribution.htm ??

So, here my questions:
* Are there simple rules of fist that enable me to guess mean and deviation from alpha and bèta?
* Why are aplha/etha and bèta more important than mean and deviation?
* What is the physical interpretation of these parameters? And how are they used in reliabiltystudies?
* How do manufacturers (empirically) estimate alpha and bèta from the physcial properties of a component?

Help is very much appeciated!! Thanx,
Gerard Westhoff

[RS Support]

You do not guess the mean or standard deviation but compute it. Formulations are given at: http://www.weibull.com/LifeDataWeb/weibull_statistical_properties.htm


As for solving for beta and eta see: Weibull Parameter Estimation http://www.weibull.com/LifeDataWeb/estimation_of_the_weibull_parameter.htm

There is also an online calculator for these parameters at: http://www.weibull.com/itools/index.htm