Characterization and modeling of continuous carbon fiber-reinforced polycarbonate under multiaxial loads

The presented work considers the characterization of a continuous fiber-reinforced thermoplastic (FRTP) with unidirectional plies under quasi-static, multiaxial loading conditions and the modeling of the non-linear mechanical behavior. For the experimental investigation, hoop wound tubular specimens fabricated from carbon fiber-reinforced polycarbonate (CF-PC) tape were tested on an electromechanical static testing machine with an axial and torsional drive. The multiaxial material response was determined for an extensive set of tests performed at several load ratios of in-plane shear to transverse stress. For modeling the non-linear mechanical response, an elasto-plastic framework is used which is kept preferably simple in order to facilitate the application in numerical analysis. Based on the generated experimental data, suitable criteria for fracture and yielding are proposed. The defined yield criterion respects the special characteristics of a FRTP such as the different mechanical behavior under transverse tensile and compressive loading as well as a linear material response in the direction of the fiber reinforcement. Furthermore, the implementation of the proposed model as a user-defined material (UMAT) routine for implicit finite element analysis (FEA) using Abaqus/Standard is described and the calibration of the plasticity approach using only the data of the in-plane shear and transverse compression tests is discussed. The model predictions for multiaxial loads are validated using the determined biaxial material data. [Display omitted] •Multiaxial experimental data for a continuous fiber-reinforced thermoplastic•Full dataset of the arithmetic mean stress–strain curves is provided•Selection of a suitable fracture criterion and definition of a yield criterion•Development and implementation of a non-linear elasto-plastic material model•Calibration and validation of the material model using the determined multiaxial data