An innovative fuzzy-inference system for predicting the mechanical behavior of 3D printing thermoset carbon fiber composite materials

Creating a complicated fiber reinforced composite structure with high mechanical performance is now possible utilizing the additive manufacturing technology. It is observed that the amount of fibers has a tremendous impact on the mechanical behavior of composites. However, additive manufacturing faces technical difficulties as with nozzle blockage and fiber aggregation at high load usage. This study aims to predict and assess the inherent mechanical characteristics of any carbon fiber-based composites fabricated by additive manufacturing at various fiber fractions with high level of accuracy. A robust adaptive neuro-fuzzy inference system (ANFIS) model for predicting mechanical performance has been constructed to compensate the experimental work assessments. The model was built by integrating the artificial neural network and the fuzzy inference system functions based upon real experimental data of thermoset carbon fiber composites fabricated by additive manufacturing technique. The precision of the computations in the model depends on the ANFIS’s four fundamental structures (fuzzification, rules, inference engine, and defuzzification). The created ANFIS model was capable of predicting all of flexural stress, flexural modulus, and elongation at break properties of the fabricated carbon composites. Results indicated that the model's predictions were substantially in accord, verifying the approach used in this work, which can enhance the technical assessment of such carbon fiber composites and expand their usability at larger sizes, opening the path for their use in a variety of technological applications requiring high levels of predictability.