Why Weibull distribution is most common used in engineering?
Reliability knowledge
Ernst Hjalmar Waloddi Weibull (18 June 1887 – 12 October 1979) was a Swedish engineer, scientist, and mathematician. (source Wikipedia)
The ultimate answer to the topic question, is flexibility. This distribution can take almost every shape. If we are looking on two parameters Weibull distribution, it can take shapes of exponential, normal and log normal distributions. If you hesitant which PDF to choose, to fit the your data results – Weibull can be an always not the worse choice. In the next graphs you can see the most common distributions used by engineers for reliability purposes and the Weibull 2p, that can mimic those:
As it seems from the graphs, it looks very close fit to the 3 major distributions, so what make it possible to happened? The secret ingredient is in the math of how Weibull 2p PDF looks like:
b - is what called, the shape of the distribution, see next graph with different b (beta) shape values.
Want to know more? Please comment what topics you are interested, in next chapter will cover the bathtub curve, widely used in the life evaluation of any product.
Asset Management Consultant | Reliability Engineering | Predictive Maintenance
2yIn addition to flexibility, Weibull distribution gathered prevalent usage as it worked even with a few failures samples (2 or 3). Weibull, himself, did not foresee the adoption by the Reliability community but pioneers in our field experimented with his technique and found informative results. Good summary article! Interested in the topic: goodness-of-fit test for choosing the right distribution.