Fundamentals of Monte Carlo Analysis

podcast episode

The survival of an individual item is difficult to determine. Use and environmental conditions, along with material and assembly differences, count. Let’s explore how to model and use the naturally occurring variability to improve reliability performance.

The Monte Carlo method is a relatively simple process that permits you to create models that include naturally occurring variability. The work of reliability engineers involves a wide range of topics and questions, two of which benefit from using Monte Carlo modeling: tolerance analysis and product life estimation.

Let’s first define the Monte Carlo method and provide a basic overview of what is necessary to create such a model for any situation. Then, let’s explore a couple of examples for life estimation and tolerance analysis. Finally, we can dive into where and how to get the necessary variation and distribution information to make this work.

There is a spectrum to engineering modeling work, from educated guesses to deterministic modeling. Along with this spectrum, the need for more information and understanding increases. We can select a number ‘out of the air’ for a guess, yet we may need a precise polymer chain length for a diffusion rate model. The Monte Carlo method is flexible enough to incorporate engineering guesses when no better data is available, yet permits the use of detailed observations and deterministic models of failure mechanisms when available as well.

This is a tool that you should use when it makes sense. Understanding how the method works and when and where it applies will help you apply this powerful approach when it is the right method.

This Accendo Reliability webinar originally broadcast on 9 April 2019.