Resin-based composite material (CFRP) surface coatings have always faced the risk of morphological damage and substrate overheating during laser cleaning. The key for solving these problems lies in the need for sufficient process experiments to establish a reliable relationship between the cleaning parameters and characteristics. Cleaning depth (H), surface roughness (S_a), and cleaning temperature (T) are the three most important cleaning indicators. H represents cleaning efficiency and effectiveness, S_a is related to the quality of re-coating, and T reflects the trend of thermal damage. Therefore, this study uses an infrared nanosecond laser to remove paint from a CFRP surface and uses laser power (P), scanning speed (V), overlap rate (η), and repetition frequency ( f ) as variables to study and statistically analyze the H, S_a, and T of the samples. Infrared thermography and high-speed imaging techniques are used to observe the temperature response of the samples, the state of the plume, and the dynamic behavior of the paint layer to determine the cleaning mechanism of the paint layer. This study is expected to provide a basic reference for improving the efficiency of laser paint removal and the quality of respraying and reducing thermal damage to CFRP substrates.

laser power, scanning speed, repetition frequency , and overlap rate. Five levels are designed under each group of parameters to form an L25 orthogonal matrix. Then, a laser cleaning experiment is conducted to obtain 25 sets of samples ranging from No.1 to No.25. After the cleaning procedure is completed, the macroscopic and microscopic morphologies of the cleaned samples are observed. At the same time, the paint cleaning depth and sample surface roughness are measured via a laser confocal microscope. Finally, the obtained experimental data are analyzed using the analysis of the variance (ANOVA) and signal to noise ratio (S/N) methods. In addition, an infrared thermographic camera is used to record the temperature response of the experimental samples during the cleaning process, and a high-speed camera is used to capture the dynamic behavior of the samples.

A signal-to-noise ratio analysis is performed on the cleaning depth, surface roughness, and cleaning temperature using the expected large, large, and small characteristics, respectively. The analysis results (Table 4) indicate that for the cleaning depth, the influencing factors are ranked from high to low by weight, namely, lap rate, laser power, scanning speed, and repetition frequency. For surface roughness and cleaning temperature, the influencing factors are ranked from high to low by weight, namely, lap rate, scanning speed, laser power, and repetition frequency. The ANOVA results (Table 5) indicate that for cleaning depth, roughness, and cleaning temperature, the critical probability (P') values of the overlap rate, scanning speed, and laser power are all less than 0.05. Therefore, at a 95% confidence level, the overlap rate, scanning speed, and laser power have statistically significant effects on cleaning depth, roughness, and cleaning temperature. In contrast, the contribution rate of repetition frequency is relatively low, with a P' value greater than 0.05, making it a less important process parameter. The detection results (Fig. 8) by the infrared thermal imager indicate that the laser cleaning process causes two high-temperature areas. The first is where the laser acts on the substrate. The minimum cleaning temperature in this area is 244 ℃, and the maximum cleaning temperature is 590.4 ℃. The other high-temperature region is the high-temperature plume region above the sample. The high-speed camera monitoring results (Fig. 11) indicate that the paint layer undergoes drastic changes due to the action of the laser, the most obvious being the generation of bright plasma and the formation of a plume perpendicular to the sample. A large number of turbid particles are observed inside the plume.

This study focuses on the influence of process parameters on the laser cleaning of paint layer on the CFRP aircraft skin. For the cleaning depth, surface roughness, and cleaning temperature, the overlap rate is the most significant influencing parameter, with contribution rates of 50.51%, 59.07%, and 69.09%, respectively. A lower overlap rate is not conducive to the uniform removal of paint, and an increase in the overlap rate will significantly increase the temperature of the substrate. Laser power and scanning speed also have a significant influence on cleaning depth, surface roughness, and cleaning temperature, whereas repetition frequency has no significant effect. The removal of paint is mainly based on the thermal erosion mechanism. During the cleaning process, the surface temperature of the paint layer rapidly increases to the decomposition temperature and the paint transforms into small particles and gases, forming a high-temperature plume above the sample. The above results will provide a reference for improving laser paint removal efficiency and respraying quality and reducing substrate thermal damage.