Traditional reliability modelling focuses on using primarily statistical models to describe reliability characteristics. This involves common probability distributions (such as the exponential, normal, lognormal and Weibull distributions). However this is a somewhat inaccurate approach: true reliability modelling involves understanding exactly how things fail and modelling the underlying causes. An example of this approach is modelling crack growth (which causes failure) using physical variables and parameters.
This approach is called a physics of failure (PoF) or mechanistic approach. It is obviously inherently more complex, and subject matter expertise is required throughout all steps. However, the analysis that mechanistic modelling permits generates highly accurate results that you can extrapolate beyond the range of the data (all other techniques limit interpretation to interpolation only).
We can help you develop the necessary understanding of the physics of failure for your system’s components. Acuitas can support the development of the ensuing mechanistic model and subsequent analysis.