Nowadays, the additive manufacturing of multifunctional materials is booming. The fused deposition modeling (FDM) process is widely used thanks to the ease with which multimaterial parts can be printed. The main limitation of this process is the mechanical properties of the parts obtained. New continuous-fiber FDM printers significantly improve mechanical properties. Another limitation is the repeatability of the process. This paper proposes to explore the feasibility of printing parts in continuous carbon fiber and using this fiber as an indicator thanks to the electrical properties of the carbon fiber. The placement of the fiber in the part is based on the paths of a strain gauge. The results show that the resistivity evolves linearly during the elastic period. The gauge factor (GF) increases when the number of passes in the manufacturing plane is low, but repeatability is impacted. However, no correlation is possible during the plastic deformation of the sample. For an equivalent length of carbon fiber, it is preferable to have a strategy of superimposing layers of carbon fiber rather than a single-plane strategy. The mechanical properties remain equivalent but the variation in the electrical signal is greater when the layers are superimposed.

This paper deals with the numerical modelling of non-destructive testing of composite parts using active thermography. This method has emerged as a new approach for performing non-destructive testing (NDT) on continuous carbon fibre reinforced thermoplastic polymer (CCFRTP) components, particularly in view of detecting porosity or delamination. In this context, our numerical model has been developed around references containing internal defects of various shapes and sizes. The first novelty lies in the fact that the heat source used in the experimental setup is modelled exhaustively to accurately model the radiation emitted by the lamp, as well as the convection and conduction around the bulb. A second novelty concerns the modelling of the CCFRTP making up the benchmark used. Indeed, its thermal properties vary as a function of the sample temperature. Therefore, the actual thermal properties have been experimentally measured and were later used in our model. The latter then captures the material dependency on temperature. The results obtained by our model proved to be in close agreement with the experimental results on real reference points, paving the way for future use of the model to optimise experimental configurations and, in particular, the heating parameters.