Prediction of deformation and failure behavior of continuous fiber reinforced composite fabricated by additive manufacturing

Additive manufacturing (AM) of continuous fiber reinforced polymer (FRP) aims to achieve rapid prototyping of complex structures with superior properties. Among the conventional AM technology, the most widely used method, an extrusion-based way, causes poor fiber-resin bonding, which leads to undesirable structural performance. Particularly establishing an accurate algorithmic model to predict the overall mechanical behaviors is challenging. The existing models lack the description of the three-dimensional progressive failure process under impact, especially the reflection of damage between printing layers. Here we overcome the inherent defects of printed continuous FRP, and characterize and predict the three-dimensional progressive failure behavior of structures with physics-based modeling under impact. These aims are achieved by (1) preprocessing to enhance printing qualities; (2) modeling with three-dimensional anisotropy to reflect printed structural characteristics; (3) validating to demonstrate model adaptabilities to deformation. As an example, the crashworthiness and damage of load-bearing structures are predicted. The model shows relative errors of 8.23% specific energy absorption and 5.08% crush force efficiency compared with experimental results. The printing layer separation controlled by inter-layer damage shows large deformation distribution. The work demonstrates comprehensive understandings of property improvement during manufacturing and performance prediction by failure model for AM fabricated FRP.

Automated summary

This study focuses on improving printing quality, establishing a three-dimensional anisotropy model, and validating the adaptability of the predictive model for continuous fiber-reinforced polymer structures under impact. Through validation and comparison with experimental results, the model shows relatively low relative errors in predicting structural performance.