Carbon steel is widely used in various modern industrial fields because of its low price and good performance, but its surface easily oxidizes and rusts. Thus, it requires periodic cleaning and maintenance. As a green, efficient, and non-contact advanced cleaning technology, laser cleaning has attracted much attention from academia and industry. At the same time, image processing, as one of the commonly used laser cleaning quality evaluation methods, has attracted much attention because of its high efficiency. In order to explore the application of laser cleaning to the rust layer on the surface of carbon steel, an orthogonal test conducted using an infrared nanosecond laser is used to determine the influence of process parameters such as the laser power, scanning speed, scanning time, and filling spacing on the rust removal effect. An evaluation standard for the carbon steel derusting effect based on the gray mean feature is established through the grayscale image processing of an area before and after laser cleaning, along with the range and variance analyses of the gray mean. The goal is to use the image recognition method to quickly and accurately determine the surface state of carbon steel, and provide quantitative evaluation indicators for the subsequent laser cleaning process.

In this study, a laser derusting experiment is performed on corroded carbon steel using the Taguchi method to explore the application of the laser cleaning process to a metal surface and the influence of the main process parameters of the laser cleaning on the cleaning quality. Then, a grayscale image based on the experimental results is obtained, and the corresponding gray mean value is used as a quality characteristic to evaluate the rust removal effect. The optimal process parameter combination is determined based on the results of range and variance analyses. Finally, the optimal combination of process parameters is verified by experiments, and a laser cleaning process for carbon steel with a better cleaning effect is obtained.

This study presents the influence of process parameters such as the laser power, scanning speed, scanning time, and filling spacing on the derusting effect based on the results from an orthogonal test conducted using an infrared nanosecond laser. If the laser power and scanning time are too large, respectively, the substrate will be oxidized and discolored. However, if the laser power and scanning time are too small, respectively, a better cleaning effect cannot be obtained. If the scanning speed and line spacing are too large, the spot overlap is sparse and the cleaning effect is poor. If the scanning speed and line spacing are too small, the spot overlap is close and the cleaning is excessive. In the case of rust removal, the surface color of the substrate will change as a result of the influence of the process parameters selected (Fig. 5). An evaluation standard for the carbon steel derusting effect based on the gray mean feature is established (Table 4) through the grayscale processing of the area before and after laser cleaning, along with range and variance analyses of the gray mean. The range analysis provides the parameters for the laser derusting using the Taguchi method, with results closer to those for an unrusted surface (Fig. 8). The variance analysis shows that the scanning speed has the greatest influence on the cleaning effect, followed by the scanning time and laser power (Table 6). The best laser rust removal process parameters (Table 7) are obtained after analyzing the range and variance analysis results for the gray mean. Finally, the obtained results are verified by experiments and gray mean measurements, and the maximum gray mean value of the experiment is obtained (Fig. 10). The surface state of carbon steel can be quickly and accurately determined using the image recognition method, which provides a quantitative evaluation index for the subsequent laser cleaning process.

The experimental results with process parameters that include the laser power, scanning speed, scanning time, and line spacing show that excessive laser power leads to the excessive cleaning and oxidation of the surface substrate, which makes the substrate lose its primary color luster and decreases the cleaning efficiency. If the scanning speed is too fast, the spot overlap is sparse and the cleaning effect is poor. If the scanning speed is too slow, the spot overlap is close and the cleaning is excessive. Too many scanning times will cause the substrate to be oxidized and discolored, and too few scanning times will limit the cleaning effect. The line spacing will affect the surface color of the substrate. If the line spacing is too small, the spots are closely overlapped, and the surface substrate is excessively oxidized to a dark gray or yellow color after corrosion removal. Moreover, the cleaning effect is poor if the line spacing is too large. In the evaluation of the derusting effect using the grayscale mean, the parameters of the laser derusting effect in Taguchi method are obtained based on the range analysis, with the results close to those for an unrusted surface. A variance analysis shows that the scanning speed has the greatest influence on the cleaning effect, followed by the scanning time and laser power. Based on the results of the range analysis and variance analysis of the grayscale mean, the optimal laser derusting process parameters are the laser power of 20 W, scanning speed of 1700 mm/s, scanning time of 26, and line spacing of 0.09 mm. Experimental results and grayscale mean measurements verify these values, with a maximum grayscale mean of 192.0431 obtained in the experiment. Thus, the optimal derusting result is reported.