Physics in a Digital Twin World

Abstract

This chapter explores the use of physics-based modeling & simulation technologies as part of Digital Twin solutions. It describes how an understanding of physics-based phenomena at microscale can be the basis for avoiding failures at a systems level, not only to improve safety, but to also reduce operating costs dramatically. The use of physics in the Digital Twin ecosystem enables the capability to see beyond the limitations of sensors and extends the ability to reasonably predict the future health state of complex mechanical systems. In mechanical systems under operation, components break when materials fail. Fortunately, the progression to failure can be modeled and anticipated using a physics-based approach to analyze the accumulation of damage due to microstructure-level stresses and eventual initiation of component failure. This capability is critical to forecasting the ability of a mechanical design to withstand operational loading conditions and for predicting component stresses that will ultimately determine the remaining useful life of the system. It is also important for establishing inspection and remediation procedures to avoid failures. Integration of physics-based modeling offers a complementary prognostics capability that, combined with advanced sensor data analytics, can be used to predict the probability of failure of key components at any point in the lifecycle - even before failure initiation. The Chapter further describes a hierarchical multi-level framework that prescribes how detailed physics- based modeling can be extended to account for behaviors at a systems level. The approach is illustrated through a use case involving aerospace – specifically, helicopter drive systems – and addresses how the multiscale framework approach that combines physics and data science can maximize health state awareness in specific applications.