Nondestructive imaging and modeling of mechanical effects of defects in composite materials

What is it about?

Advanced composite materials underpin the superior mechanical performance of many modern rotary- and fixed-wing aircraft structures, yet these intricate materials are frequently susceptible to forming complex manufacturing defects that can cause difficult-to-predict effects/limitations on operating conditions of composite structures. Typically, costly and time-consuming experimental programs are required to certify complex, non-conforming parts that are seeded with worst-case defects. In this paper, we present a digital-twin-like modeling approach that inputs nondestructive 3D imaging and then virtually tests and characterizes the realistic, as-manufactured performance of composite parts that contain industry-relevant defects.

Featured Image

Photo by Andrew Palmer on Unsplash

Photo by Andrew Palmer on Unsplash

Why is it important?

This paper demonstrates clear links between measurable defect characteristics and the degree/type of mechanical property degradation (e.g., stiffness, strength) of a complex composite part, without needing to experimentally test and destroy the part. The generalizable, machine-learning-enhanced methodology combines new computer vision techniques with advanced model meshing and virtual testing algorithms.