Structural health monitoring of solid rocket motors using machine learning

What is it about?

Structural health monitoring as a means of determining both damage and defects in solid rocket motors is of significant interest to the defense industry because of the potential for cost savings and accident prevention. The ability to detect damage and defects in structures will help enable more timely, accurate, and reliable assessments of structural integrity. A fundamental understanding of stress and strain responses to storage conditions and to damage mechanisms is critical for the development of improved service life predictions for the next generation of propellants and motors. With the advent of data-driven science and engineering together with technologies of machine learning, it becomes possible to tackle and resolve these advanced engineering challenges. In this paper, we aim to address these challenges by focusing on structural health monitoring and damage detection using an integrated computational mechanics and machine learning principles. Specifically, finite-element models are used to characterize the stress sensor readings along the case wall for various groups of flaws. Using this information, the inverse problem is solved using concepts from machine learning and artificial neural networks. Quantitative characterization of defect sizes can be identified, based on the radial-stress data from embedded stress sensors in motor structures.

Featured Image

Photo by SpaceX on Unsplash

Photo by SpaceX on Unsplash