The laying environment of petroleum drill pipe is always poor, and the laid soil contains oxidizing bacteria, moisture, air, etc. During service, rust often occurs at the internal and external threads of the drill pipe joint, resulting in unstable connection of internal and external threads. In order to improve the utilization rate of steel, derusting methods are often used to repair the surface of metal parts in industrial field. The traditional rust removal methods are mainly mechanical, manual and chemical methods, among which the mechanical methods are sand blasting and high-pressure water flow. Compared with traditional cleaning methods, laser cleaning has the characteristics of green environmental protection, high efficiency and non-contact. It has gradually become a common cleaning method in industry. As an advanced cleaning technology, laser cleaning focuses high-energy laser on the corrosion layer to be cleaned. After the surface corrosion layer absorbs laser energy, the thermal coupling phenomenon and plasma shock wave effect appear, and the effect of surface rust removal is realized. Therefore, the pulse laser is used to clean the corrosion on the surface of NC50 petroleum pipe thread, and the significance of the influence of different factors on the cleaning effect is systematically analyzed to provide theoretical guidance for the laser cleaning of industrial petroleum pipe thread.

The laser cleaning experiment is carried out on the petroleum pipe thread by the orthogonal experiment method, and the orthogonal data table is analyzed with the surface roughness as the evaluation index, and the significance order of the factors influencing the laser cleaning effect is obtained. The surface roughnesses and oxygen contents under different process parameters are obtained by the single factor experiment. The surface damage of cleaning samples is measured with oxygen contents, and the process parameters are optimized by taking surface roughness as the optimization objective. Based on the response surface methodology, the mathematical model is established to describe the relationship between the optimization objective and laser process parameters. By combining the mathematical model with the optimized particle swarm optimization algorithm, the optimized process parameters are obtained.

According to the orthogonal experimental data, the influence of laser power on the experimental results is significant. On this basis, single factor experiments are carried out to analyze the variation of surface roughness and oxygen content of cleaned workpiece with laser power, scanning speed, and defocusing amount. Among them, when the scanning speed is 1500 mm/s, the defocusing amount is + 1 mm and the laser power is variable, the workpiece surface roughness value (Fig. 6) and oxygen content (Fig. 7) first decrease and then increase, and both reach the lowest value when the laser power is about 500 W. With the increase of the laser power (below 500 W), the corrosion layer vaporizes obviously, the corrosion layer on the workpiece surface is removed, the surface roughness value decreases, and the oxygen content decreases. Increasing the laser power continuously causes the melt to adhere around the spot pit and increase the surface roughness. In order to expand the particle search space and improve the population diversity in the iterative process, a particle swarm optimization algorithm is proposed by using the combination of power function and learning factor as an improved operator fused into the inertia weight. The convergence speed of the curve obtained by the improved algorithm is faster than that obtained by the traditional particle swarm optimization algorithm (Fig. 16). The optimized process parameter combination is obtained by combining the improved particle swarm optimization algorithm with the mathematical model of response surface method. The particle swarm optimization algorithm predicts that the surface roughness after cleaning is about 4.73 μm. According to the optimization algorithm, the combination of process parameters is as follows: the laser power of 488 W, the defocusing amount of + 3 mm and the scanning speed of 3000 mm/s. Using this parameter combination for laser cleaning and micro-morphology detection experiments, the surface roughness of the sample is about 4.64 μm. The accuracy of the prediction of the improved particle swarm optimization algorithm is improved obviously.

In this study, the significances ranking of the factors (the laser power, defocusing amount, and scanning speed) influencing the cleaning effect of petroleum pipe threads is obtained by the orthogonal experiment. The relationship mathematical model between the workpiece surface roughness and laser process parameters is established based on the response surface method, which is combined with the optimized particle swarm optimization algorithm to obtain the optimized process parameter combination. The factors influencing the cleaning effect of petroleum pipe threads are the laser power, defocusing amount, and scanning speed according to their significances ranking from the highest to the lowest. The combination of power function and learning factor is used as the improved operator of the particle swarm optimization algorithm to dynamically adjust the inertia weight. The improved algorithm is obviously better than the traditional particle swarm optimization algorithm in convergence speed. According to the optimization algorithm, the combination of process parameters is as follows: the laser power is 488 W, the defocusing amount is + 3 mm, and the scanning speed is 3000 mm/s. The laser cleaning experiment is carried out by using the optimized process parameters. The results show that the micro-morphology of the workpiece cleaned under the combination of process parameters is relatively smooth in the molten pool, and the shape of the pit is close to the shape of the Gaussian light source. The oxygen content is removed, and the optimization effect is obvious. The accuracy of the method proposed in this paper is proved, and it can provide theoretical guidance for the optimization of process parameters of laser cleaning.