Corrosion

Predicting corrosion failures is a long-standing scientific challenge with very important implications for society. Labelled the biggest threat to materials’ sustainability, the failure of structures and industrial components due to corrosion entails a global cost of $2.5 trillion/year, equivalent to roughly 3.4% of the global GDP. Only in the UK, corrosion comes at a cost of 46 billion pounds per year. While great physical insight has been gained through decades of experimental corrosion research, the development of mechanistic, predictive models continues to thwart scientists and engineers due to the complex, multi-disciplinary and strongly coupled nature of the problem. For example, corrosion pits grow driven by the local chemical and mechanical fields but these are themselves dependent on the pit morphology. We have tackled this challenge and developed the first generation of mechanistic corrosion models by combining two new paradigms in the area of corrosion science: (i) phase field algorithms, which enable computational tracking of the corrosion front, and (ii) multi-physics modelling, establishing robust computational schemes capable of simulating coupled electro-chemo-mechanical phenomena. In this way, the physical processes governing corrosion can be explicitly simulated, enabling physics-based corrosion predictions that do not rely on strong assumptions (i.e., from “mesoscale first principles”). 

Furthermore, we have extended this class of electro-chemo-mechanical phase field-based models for material dissolution to the area of bio-corrosion (or biomaterial degradation). Mg alloys hold great promise as biodegradable materials due to their excellent biocompatibility and biodegradability. Screws and plates made of Mg alloys could provide stable implant materials that degrade in vivo, eliminating the need for a second surgical operation to remove the implant. However, this promise is being held back due to the rapid corrosion rate of Mg (degradation occurs before healing). We have recently developed the first computational model that is capable of predicting localised Mg corrosion in biological fluids. This work opens the door to the development of a new class of electro-chemo-mechanical models that can be benchmarked against in vitro experiments yet provide in vivo insight. Such a virtual tool has the potential  to efficiently test multiple potential solutions, materials and design strategies to develop biodegradable implants, from ad hoc alloying to Mg-based composites. 

A recent talk on this topic is given below: