A strength prediction model for 3D-printed continuous fiber reinforced composites

Abstract This study presents the development of a predictive model to determine the tensile strength of continuous fiber thermoplastic composites produced through 3D printing. The proposed model utilizes the Curtin model to predict composite strength by incorporating statistical Weibull parameters related to composite filament strength and fiber volume fraction. Moreover, the model incorporates a welding function that captures the effect of the temperature history on the tensile strength, which is influenced by the interfacial shear strength at the filament-matrix interface. The proposed model provides a direct link between the printing process parameters and the tensile strength of the 3D-printed composite, while the spatial variation of strength is captured. The development of the model is based on a comprehensive experimental campaign, which includes single-fiber tensile tests to determine the fiber stiffness, tensile tests on composite filaments with different gauge lengths to extract the Weibull parameters of filament strength, and tensile tests on annealed 3D-printed composite samples to determine the welding function parameters. To quantify the thermal history during the 3D printing process and subsequently calculate the welding function, a finite element model is implemented in the commercial FEM software Abaqus through the utilization of advanced subroutines. The accuracy of the strength predictive model is validated against tensile tests conducted on 3D-printed composite coupons. The framework was calibrated and validated for PolyLite PETG reinforced with Anisoprint CCF 1.5 K carbon-fiber composite filament. The fitted welding-function parameters were A = 6.41 × 10 − 11 s and E a =11.202 kJ mol − 1 . For the withheld validation conditions, the model predicted the tensile strength with RMSE = 5.67 MPa and MAPE = 0.57%. This model, therefore, enables a cost- and time-efficient optimization of the manufacturing process to achieve the desired mechanical performance, reducing costly trial-and-error-based 3D printing campaigns.