Modeling materials under coupled extremes: Enabling better predictions of performance

Materials for the next generation of electric power infrastructure will be subject to harsh service environments featuring extreme levels of stress, temperature, irradiation, and corrosive attack, often simultaneously. In this article, we review the overarching technical issues involved in designing/certifying materials that can withstand these conditions, with specific examples given for fusion plasma containment, molten salt reactors, and thermoelectric devices. Then, we examine the new advances and broad persistent needs for modeling tools capable of both accelerating the discovery of improved materials for such applications, and predicting how materials will evolve, degrade, and eventually fail. Particular emphasis is given to the need for advancing materials informatics and machine learning capabilities in concert with ever more comprehensive multiphysics simulations at intermediate time and length scales to understand how coupled extremes affect properties and performance differently than single extremes in isolation. Graphical abstract