Influence of residual stress and fluid–structure interaction on the impact behavior of fused filament fabrication components

Abstract

Despite the proliferation of cellular fused filament fabrication (FFF) polymer components for a variety of industrial applications, few studies have investigated their fluid–structure interaction (FSI) behavior during loading, particularly under dynamic impact conditions. Furthermore, the extent to which residual stresses from the FFF build process affect the dynamic load bearing characteristics has not been addressed. In this work, simulations and experiments are conducted for cylindrical nylon specimens fabricated with two different internal closed-cell cavity structures to assess the influence of the entrapped fluid and the FFF residual stresses on the state of stress during high strain-rate impact. The demonstrated 2-stage computational approach includes a thermomechanical model of the FFF build to calculate residual stress and distortion, which forms the initial state for a subsequently executed dynamic impact model using smoothed particle hydrodynamics (SPH) to capture the effects of air within the internal cavities. Dynamic displacement boundary conditions for the FSI simulations are identified using digital image correlation (DIC), obtained from impact experiments on the FFF specimens performed using split Hopkinson pressure bar (SHPB) tests. Findings reveal that FFF residual stresses significantly influence the stress–strain response during dynamic impact, even at strain rates of 500–600 s−1. In addition, while the influences of both FFF residual stress and FSI vary with internal cellular structure, the study reveals that their coupled effects must be considered to accurately characterize the impact behavior. Validity of the 2-stage numerical approach, as well as significance of FFF residual stress and the influence of FSI, are justified by comparing numerical predictions with experimental measurements, and observing root-mean-square stress errors within 12.77% and 11.87%, and peak stress errors within 1.93% and 1.34% for the two specimens.