Modelling crack growth.
Understanding variability.

The research question

Fatigue cracks grow under repeated loading. A computational model can describe that growth, but material behaviour and experimental data vary. How can a model account for that variation rather than provide only one predicted life?

The approach

The work improved the UniGrow fatigue crack-growth model and combined it with Monte Carlo simulation. It considered constant and variable-amplitude loading and compared calculations with experimental data from aluminium and steel alloys.

The model connected fracture mechanics with local elastic-plastic stress and strain near the crack tip. Simulation provided a way to study how variability in the inputs affected the distribution of predicted fatigue lives.

Why it matters

The research connects a mechanistic model to a probabilistic description. It offers an example of investigating physical behaviour, translating a theory into software, and comparing predictions with observations.

This section documents historical academic research. The thesis and publications contain the methodology, assumptions, and validation details.

Source: University of Waterloo thesis record