Experimental and numerical investigation of the laser shock-based paint stripping process on CFRP substrates

Various laser shock wave applications have been proposed for metals; the most popular is the laser shock peening. In composites, the use of the laser shock wave technology is still at an early stage. Laser shock paint stripping (LSPS) is a method where a laser-induced shock wave is used to remove a paint from a substrate. In this study, a combined experimental–numerical study was performed on the LSPS on Carbon Fiber Reinforced Polymers (CFRP) substrates. Two types of specimens were investigated: specimens with an epoxy structural primer covered by a polyurethane base and a clear coat and specimens with the above layers plus an epoxy exterior primer between the epoxy structural primer and the polyurethane layer. The specimens were shot using various laser intensities aiming to obtain the stripping threshold and the stripping percentage as functions of the laser intensity. The diameter of the stripped area was measured by a profilometer. An electron microscope was used to evaluate the stripping efficiency. In parallel, a numerical model was developed using the LS-DYNA explicit Finite Element (FE) software. For the modeling of the CFRP, a progressive damage model with strain rate parameters was used while for the epoxy an elastic–plastic-hydrodynamic material model combined with the Gruneisen equation of state. The interface between these two materials was modeled using cohesive zone elements of zero thickness, which follow a bi-linear traction-separation law. Calibration of the modeling of the shock wave propagation inside the composite material was achieved through the comparison of the back-face velocity profiles, measured by a velocity interferometer system for any reflector system. The results from the multiple tests conducted revealed the repeatability of the process and the significant effect of the laser intensity. On the other hand, the model managed to predict well the back face velocity–time profile but in most cases, it failed to predict the stripping pattern. This is mainly due to lack of crucial material properties and the assumptions made in the modeling of the applied pressure profile.