Data-Driven Physics-Based Modeling Tools to Determine Effective Mechanical Properties of As-Built Composite Structures

This opportunity targets the development of a software toolkit to automate the generation of nonlinear, anisotropic mechanical properties for as-built composite structures, including the effects of defects, to accelerate finite element (FE) analysis for fleet repairs and aircraft production non-conformal dispositions.  To meet this objective, Navy wants to implement a data-driven, physics-based, modeling approach including material structure and full-field deformation by analyzing the parts in the as-built condition with their own unique configuration, including manufacturing defects and service damage. Phase I effort by the Numerical Technology Company, LLC (NTC) teaming with the University of Texas Arlington (UTA) successfully demonstrated the feasibility of the proposed concept.  The NTC – UTA Team developed a unique working prototype of the toolkit for measuring in-situ mechanical properties of as-built composite structures – the biggest challenge of this SBIR topic.  The proposed concept addressed nonuniform test article (coupon) geometry; complex material structure / irregularity; multiple materials and complex stacking sequences present in the same coupon; and implemented effective small test articles attainable for machining from complex as-built parts, viable for automation.  Such articles are a practical building block for analyzing larger structural elements, up to full-scale components.  NTC – UTA Team demonstrated transition of the X-Ray CT data to FEM and comprehensive progressive damage and failure analysis (PDFA) capturing the multiple failure modes and their interaction in composite structure; assessing nonlinear properties for multiple materials; and verifying the predicted material properties through PDFA.  The material and manufacturing complexity of the as-built parts clearly showed the need in physics-based approaches capturing the failure modes and their interaction – which could not be replaced by black-box methods.  The proposed concept has been demonstrated on the V-22 composite proprotor blade structure.  This concept is relying on the physics-based, data-driven methods which bridge multiple modeling scales and increase throughput of the in-situ material characterization through automation and machine learning (ML). The Team will also apply machine learning to generation of the in-situ properties especially as relevant to multiscale approaches and interrogation. Microstructural measurements and data-driven models will be used for training the models at higher scales, and exploiting data collected from testing at the macro-scale, by prototyping and utilizing a cascaded homogenization computational framework.  Such multiscale interrogation and modeling will be utilized for the models that contain the as-manufactured features of composite parts.