In order to reduce the energy consumption and processing time of laser cladding process, this paper proposes a laser cladding process parameter optimization method based on energy efficiency. Taking specific energy consumption as the measure of laser cladding energy efficiency, the laser cladding energy consumption model is established by analyzing the energy consumption characteristics of each subsystem of laser cladding. On this basis, the specific energy consumption between the total energy consumption of the system and the forming volume is calculated, and a multi-objective optimization model of laser cladding process parameters with specific energy consumption and processing time as optimization targets is constructed and solved by NSGA-Ⅱ algorithm. The accuracy of the specific energy consumption model is verified by laser cladding test on annular thin-walled parts, and it is proved that the optimized process parameters can effectively improve the energy efficiency of laser cladding and reduce processing time.

Laser cladding is a high-precision and high-efficiency material surface modification technology, which is widely used in major national engineering fields such as national defense, aerospace, petrochemical, and automobile manufacturing. There is a close relationship between the energy efficiency of laser cladding and the energy consumption in the corresponding processing process. Dong Mengmeng et al. constructed a model that can predict the energy consumption of laser cladding process by analyzing the energy consumption characteristics of each subsystem of laser cladding. Jiang et al. established an energy consumption model for laser cladding remanufacturing process and optimized the model using the improved NSGA-Ⅱ algorithm, providing a basis for reducing the energy consumption and cost of laser cladding remanufacturing process. Duan Chengmao et al. established a low-carbon optimization model for laser welding based on processing efficiency and carbon effect as the optimization target, and proposed a low-carbon optimization method for laser welding process system. Jiang Xingyu et al. established a carbon emission model for laser additive manufacturing process and optimized related process parameters with carbon emission, powder utilization rate and cladding quality as the optimization targets. Panda et al. calculated energy consumption based on sintering area, established a laser additive manufacturing energy consumption optimization model that comprehensively considered energy consumption and production efficiency, and pointed out that the thickness of the cladding layer is an important parameter affecting the energy consumption of laser additive manufacturing. Streiten- berger et al. proposed a two-stage optimization method based on factor analysis and normal plane intersection method. By optimizing the weld geometry, the multi-objective optimization stochastic problem with the minimization of laser welding energy consumption and material consumption as the optimization target was solved.

It should be pointed out that most of the existing related studies are based on the total energy consumption of the laser cladding system for process optimization, and there are few reports on the optimization of laser cladding process parameters from the perspective of energy efficiency. In view of this, this paper, based on the construction of the energy efficiency model of laser cladding processing of mechanical parts, establishes a multi-objective optimization model for laser cladding processing with specific energy consumption and processing time as the optimization targets, and uses the NSGA-Ⅱ algorithm to solve and obtain the best process parameter combination, and verifies the accuracy of the constructed model with the help of multi-layer and multi-pass laser cladding tests on annular thin-walled parts.

1 Energy efficiency modeling of laser cladding processing

1.1 Energy efficiency index function of laser cladding processing
The energy efficiency of laser cladding processing refers to the degree to which energy and resources are effectively utilized during the laser cladding processing of mechanical parts. Its evaluation indicators include specific energy consumption, energy utilization rate, cladding efficiency, production speed, material utilization rate, process quality and defect rate, etc. The specific energy consumption of laser cladding refers to the energy consumption generated by the unit volume of forming during the cladding process. The smaller the specific energy consumption, the higher the energy efficiency of laser cladding. Existing studies have shown that specific energy consumption can accurately and intuitively characterize processing energy efficiency. Therefore, this paper uses specific energy consumption as an indicator to characterize the energy efficiency of laser cladding processing. Its calculation formula is: see formula (1) in the figure. In the formula, elc is the specific energy consumption of laser cladding, EA is the total energy consumption of the laser cladding process, and Vlc is the laser cladding forming volume.

1.2 Analysis of energy consumption characteristics of laser cladding
The operation process of laser cladding equipment can be divided into equipment startup, standby, water cooling equipment operation, cladding processing, equipment cooling, standby and shutdown stages. The equipment subsystems in each stage operate in coordination and the power changes are relatively stable. Taking the multi-layer and multi-pass laser cladding process as an example, the real-time power curve of the laser cladding equipment obtained by the power monitoring device is shown in Figure 1. Among them, the inter-layer stop light refers to the stage of waiting for the cladding layer to cool down.

According to the energy consumption composition of the laser cladding equipment and the operation of the equipment subsystems at different stages, the total energy consumption EA of the laser cladding process is mainly composed of five types of energy consumption, namely, the laser system energy consumption El, the water cooling system energy consumption Eh, the laser cladding special machine tool energy consumption Em, the powder feeder and protective gas system energy consumption Epg and the auxiliary system energy consumption Ea, that is: See formula (2) in the figure.

1) Laser system energy consumption
The power of the laser system changes continuously during the laser cladding process. Based on this, the laser system energy consumption El is divided into cladding process energy consumption Etc, interlayer stop light energy consumption Ed and standby energy consumption Els, that is: see formula (3) in the figure.
In the formula, Pin is the laser input power, tlc is the metal powder melting time, that is, the cladding time, Pls is the laser standby power, td is the interlayer stop light time, and tls is the laser standby time, that is, the cladding preparation time. Among them, the cladding time tlc is determined by the ratio of the total laser cladding length Slc and the laser scanning speed vs. The cladding paths of mechanical parts of different shapes are different, and the corresponding calculation methods of the total cladding length will also be different. Taking common flat parts and rotating parts as examples, the calculation formulas for the laser cladding processing time of the two are: see formula (4) in the figure.
In the calculation formula of laser cladding processing time for flat parts, m is the number of cladding layers of flat parts, n is the number of cladding layers of flat parts, and L is the length of flat parts to be clad. In the calculation formula of laser cladding processing time for rotating parts, mi is the number of laser cladding layers of the i-th layer (i=1,2,3,…,n) of the rotating part, D is the diameter of the rotating part, and Hlc is the height of a single layer of the cladding layer of the rotating part. The change of the diameter of the rotating part during the processing process should be considered when calculating the laser cladding processing time for rotating parts. In addition, the number of cladding layers n of mechanical parts is related to the surface height Hw of the part to be processed, the cladding layer height Hlc and the Z-axis lifting amount ΔZ, and the number of cladding layers of mechanical parts is related to the total cladding width W and the width of a single cladding layer Wlc. Since the spot diameter lsd of the laser used in this study is a fixed value, the width of a single cladding layer of a mechanical part Wlc=lsd. The calculation formulas for the number of cladding layers n and the number of cladding paths m of a mechanical part are as follows: See formulas (5) and (6) in the figure.
In formula (6), μ is the overlap rate. Because the number of cladding layers and the number of paths of a mechanical part in the actual processing are both integers, the number of cladding layers n and the number of paths of a mechanical part calculated based on formulas (5) to (6) should be rounded up.

2) Energy consumption of water cooling system
When high-power equipment such as lasers undergo energy conversion (such as electrical energy into light energy), part of the energy in the system will be directly converted into heat energy due to factors such as heat loss, which will seriously affect the quality of laser cladding processing and even damage the equipment. Therefore, a cooling system is required to cool equipment such as lasers. In this study, the energy consumption of the water cooling system Eh is divided into two parts according to the working state of the water cooling system: the working energy consumption Ehc of the water cooling system and the standby energy consumption Ehs, that is: see formula (7) in the figure.
In the formula, Phc is the working power of the water cooling system, Phs is the standby power of the water cooling system, thc and ths are the working time and standby time of the water cooling system, respectively. The calculation formulas of the two are: see formula (8) and (9) in the figure.
In formula (8), Plc is the laser power, ρ is the cooling water density, cp is the specific heat capacity of the cooling water, vh is the cooling water flow rate, and ΔT is the cooling water temperature difference.

3) Energy consumption of laser cladding machine tool
The laser cladding equipment machine tool used in this study is mainly composed of a rotating mechanism and a translation mechanism. The rotating mechanism is driven by a servo motor to control the rotation of the workpiece to complete the laser cladding, and the translation mechanism is driven by a stepper motor to move the workpiece to the starting position of the cladding. The energy consumption Em of the special machine tool is calculated as follows: See formula (10) in the figure.
In the formula, Pmr is the working power of the rotating mechanism, Pmt is the working power of the translation mechanism, and tmt is the working time of the translation mechanism, that is, the time for the workpiece to return to the starting position of the cladding after each layer of cladding is completed. There is: (11). In the formula, vmt is the moving speed of the translation mechanism.

4) Energy consumption of powder feeder and protective gas system
The powder feeder and protective gas system provide metal powder and protective gas required to prevent powder oxidation during the laser cladding process of mechanical parts. The powder feeder of the laser cladding equipment used in this study is a gravity powder feeder, and its working power is a fixed value. The shielding gas system is responsible for blowing powder into the molten pool and preventing the powder in the molten pool from high-temperature oxidation. Therefore, the shielding gas system and the powder feeder operate synchronously, and their working time is equal to the laser cladding time tlc. The energy consumption Epg of the powder feeder and the protector system is calculated as follows: (12).
Where Ppg is the working power of the powder feeder and the shielding gas system.
5) Auxiliary system energy consumption
The laser cladding auxiliary system includes lighting, integrated control, and air conditioning for heat dissipation of the laser control cabinet. Its power Pa is a fixed value, and its working time is the same as the time consumed by the entire laser cladding process of mechanical parts (tlc +tls +td). The auxiliary system energy consumption Ea is calculated as follows: (13).
In summary, during the laser cladding process of mechanical parts, although the operating status of each subsystem of the cladding equipment changes greatly, the system power changes at different stages are relatively stable, the power jump time is short, and the energy consumption is less. Therefore, this study only considers the energy consumption of each subsystem in the working and standby stages when establishing the laser cladding energy consumption model, and the corresponding total energy consumption EA of the laser cladding process can be expressed as: (14).
The specific energy consumption elc is used to characterize the energy efficiency of the laser cladding process, and its calculation formula is: (15).

2 Laser cladding parameter optimization based on energy efficiency

2.1 Problem description
The multi-objective optimization problem of laser cladding processing of mechanical parts based on energy efficiency can be described as: multi-layer and multi-pass laser cladding processing of mechanical parts requires the lowest specific energy consumption and the shortest processing time under the premise of meeting the processing size requirements. This study uses three process parameters that have a greater impact on the specific energy consumption and processing time of laser cladding processing of mechanical parts, such as laser power Plc, scanning speed vs and powder feeding amount Sp, as optimization variables, and multi-objective optimization is carried out with the goal of minimizing specific energy consumption and shortest processing time. The constructed optimization model is shown in Figure 2.

2.2 Objective function and constraints
1) Objective function The specific energy consumption elc and processing time tA of laser cladding processing of mechanical parts are taken as optimization targets, and the latter can be expressed as: (16).
2) Constraints
The smooth implementation of the laser cladding processing process of mechanical parts requires the satisfaction of multiple constraints such as laser input power, laser scanning speed, powder feeding amount and processing allowance.
① Laser power constraint The change of laser power will affect the surface heat flux density of mechanical parts. When the laser power is too small, the cladding powder cannot be completely melted, and excessive power will cause high-temperature deformation of mechanical parts. Therefore, according to the equipment parameters used and the actual processing requirements, the selected laser power Plc cannot be lower than the minimum laser power Plc-min, nor higher than the maximum laser power Plc-max, that is: (17).
② Laser scanning speed constraint The laser scanning speed directly affects the forming quality of the cladding layer of mechanical parts. Too slow laser scanning speed will cause the dilution rate of the cladding layer of mechanical parts to be too high, thereby reducing the hardness of the cladding layer. However, if the laser scanning speed is too fast, the cladding layer of the mechanical parts will not be able to form a metallurgical bond with the substrate, which will have an adverse effect on the bonding strength of the parts. Therefore, the selected laser scanning speed vs cannot be lower than the minimum laser scanning speed vs-min, nor can it exceed the maximum laser scanning speed vs-max, that is: (18).
③ Powder feeding amount constraint In the actual processing process, the size of the powder feeding amount directly affects the quality and performance of the laser cladding processing of mechanical parts. Too small a powder feeding amount can easily lead to pores and inclusions in the cladding layer of mechanical parts, while too large a powder feeding amount will make it difficult for the powder to fully melt, thereby reducing the quality of the laser cladding processing of mechanical parts. Therefore, the powder feeding amount Sp in the laser cladding processing of mechanical parts cannot be lower than the minimum limit value Sp-min, nor can it be higher than the maximum limit value Sp-max, that is: (19).
④ Processing allowance constraint The surface of the mechanical parts after laser cladding processing is wavy and rough, and cannot be used directly. Therefore, during the laser cladding process, a certain processing allowance h must be reserved for the mechanical parts to facilitate subsequent processing. The machining allowance of laser cladding of mechanical parts must meet the actual processing requirements. In order to prevent the machining allowance of laser cladding of mechanical parts from being too small and unable to be processed subsequently, the machining allowance of cladding should not be lower than the minimum machining size hmin of the parts. At the same time, in order to save costs and reduce subsequent processing steps, the machining allowance of laser cladding of mechanical parts should not exceed its maximum machining size hmax, that is: (20).
The multi-objective optimization model of laser cladding of mechanical parts under the above four constraints is constructed as follows: (21).

2.3 Optimization model solution

The multi-objective optimization model of laser cladding of mechanical parts constructed in this study is optimized with the help of NSGA-Ⅱ algorithm. The solution process (see Figure 3) is as follows: ① Initialize the population. Randomly generate a group of individuals corresponding to a group of process parameter combinations of laser power, scanning speed and powder feeding amount. ② Evaluate fitness. Based on the specific energy consumption model and related experiments, calculate the specific energy consumption and processing time corresponding to the randomly generated individuals, and assign the calculated values ​​of the two to the two objective functions of the individuals respectively. ③ Non-dominated sorting. All individuals are non-dominated and sorted into multiple different fronts. Front 1 contains the optimal solution, front 2 contains the suboptimal solution, and so on. ④ Crowding calculation. Calculate the crowding distance of individuals in each front. ⑤ Select the next generation. Use the selection operator to determine the individuals of the next generation. ⑥ Crossover mutation. Perform crossover mutation on the individuals of the next generation to generate new individuals. In this step, the process parameter values ​​are modified probabilistically and candidate solutions are generated. ⑦ Update the population. The newly generated population is merged with the original population to form a new population. ⑧ Iteration. Repeat steps ②~⑦ until the maximum number of iterations is reached.

3 Case analysis

3.1 Experimental conditions
The BS-ODE6000 semiconductor laser surface treatment equipment (see Figure (4)) is used to perform laser cladding processing on annular thin-walled parts, and the real-time power of the laser cladding equipment is collected with the help of power monitoring equipment. The material of the sample is ZG32MnMo cast steel, with an outer diameter of 165 mm, a height of 3 mm and a width of 20 mm to be processed. The cladding powder used is an iron-based alloy, and the cooling water is deionized water or pure water. According to the actual processing state and equipment performance, the fixed parameters of the laser cladding processing equipment for mechanical parts are determined as shown in Table 1.

3.2 Related parameter acquisition test

3.2.1 Laser system power
The laser cladding equipment laser used in the test is a semiconductor laser. With the help of power monitoring equipment, different laser input powers and their corresponding laser powers are measured. The results are shown in Table 2.
The data in Table 2 are fitted, and the functional relationship between the laser input power and the laser power can be obtained: (22).

3.2.2 Special machine tool power
The special machine tool for laser cladding equipment includes a rotating mechanism and a three-dimensional translation mechanism. The rotating mechanism is driven by a servo motor, and its power is determined by the torque and the spindle speed. The speed of the machine tool spindle in this experiment is the laser scanning speed, which has a small range of variation, so the servo motor working power does not change much. The servo motor working power is measured multiple times using power detection equipment and the average value is calculated, and the rotation mechanism working power Pr is 157 W. The three-dimensional translation mechanism is driven by a stepper motor, and its working power is mainly determined by the spindle movement speed. In this experiment, the horizontal movement speed vtm of the special machine tool is 10 mm/s, and the three-dimensional translation mechanism working power Pt is measured by the power monitoring equipment to be 80 W.

3.2.3 Cladding layer height and interlayer stop beam
Under different laser power, scanning speed and powder feeding conditions, the sample was subjected to a single-layer single-pass laser cladding processing test. Afterwards, the sample was cut along the normal direction of the cladding layer to measure the cladding layer height Hlc. Figure 5 shows a cross-sectional photo of the sample cut. The measurement results of the sample cladding layer height under different parameter conditions are shown in Table 3.
The least squares method was used to fit the laser cladding process parameters in Table 3 with the single-layer cladding layer height data of the corresponding sample, and the calculation formula for the sample cladding layer height was obtained as follows: (23).
Preliminary research results of this research group show that as the number of sample cladding layers n increases, a certain amount of time needs to be waited after each layer of cladding is completed to ensure that the cladding area cools down, thereby avoiding a decrease in the hardness of the cladding layer. Among them, the waiting time when the laser beam temporarily stops is the interlayer stop beam time. Combined with the materials used in this experiment and the actual processing requirements, the optimal interlayer stop beam time is determined to be: (24).

3.3 Energy efficiency model verification
The processing energy efficiency test values ​​obtained by multi-layer and multi-pass laser cladding tests on annular thin-walled samples and the corresponding theoretical values ​​obtained by using the laser cladding processing energy efficiency model established in this paper are shown in Table 4. As can be seen from Table 4, the theoretical values ​​of the laser cladding processing energy efficiency of multiple groups of samples are basically consistent with the corresponding experimental values, and the average error is small, about 3%. This result proves the accuracy of the energy efficiency model established in this paper.

3.4 Optimal parameter solution and verification
The NSGA-Ⅱ algorithm is used to solve the multi-objective optimization model of laser cladding processing of mechanical parts. The Pareto frontier coefficient is set to 0.3, the population size is 100, and the number of iterations is 200. The Pareto frontier solution set containing 27 points is shown in Figure 6.
The optimal process parameter combinations under different optimization objectives and the corresponding optimization results are listed in Table 5. Analysis of the data listed in Table 5 shows that when the minimum specific energy consumption is the main optimization goal, the smaller laser power and slower scanning speed reduce the processing energy consumption, and the larger powder feeding amount leads to an increase in the volume of laser cladding forming, which reduces the specific energy consumption of laser cladding, but the slow scanning speed also causes the laser cladding processing time to be extended. When the shortest processing time is the main optimization goal, the larger laser power and faster scanning speed increase the processing energy consumption and shorten the processing time, but the small change in the volume of laser cladding forming causes the laser cladding specific energy consumption to increase relatively. When the specific energy consumption is low and the processing time is short, the specific energy consumption is increased by 7.8% when only the minimum specific energy consumption is considered, and is reduced by 11% when only the shortest time is considered. The processing time is reduced by 25% and increased by 12% respectively. Therefore, the process parameters with the lowest specific energy consumption and the shortest time as the goals can balance the two goals, while ensuring the processing efficiency while maximizing the energy efficiency of laser cladding processing.

In order to verify the accuracy of the optimization results, the optimal process parameter combination shown in Table 5 and the empirical values ​​of laser cladding processing are selected as test parameters, and the laser cladding comparison test is carried out on the samples with the processing width of 20 mm and height of 3 mm as the target. The comparison results are shown in Table 6. It can be seen from Table 6 that the specific energy consumption and processing time of laser cladding using the optimized parameters are reduced by 5.1% and 9.5% respectively compared with the corresponding values ​​with the empirical values ​​as parameters. Based on the above analysis, the multi-objective optimization model for laser cladding processing of mechanical parts established in this paper has high accuracy and effectiveness. At the same time, in order to ensure that the quality of the cladding layer will not be affected when the optimized parameters are used for laser cladding processing, the quality of the sample cladding layer under the two groups of test parameters is tested and compared. Hardness is taken as one of the main indicators of the quality of the reaction cladding layer, and its size directly reflects the strength of the wear resistance of the sample surface. The hardness test of the sample cladding layer is carried out, and the hardness change corresponding to different depths of the sample cladding layer (i.e., the distance from the hardness test point to the top of the cladding layer) is obtained as shown in Figure 7. As shown in Figure 7, the hardness value of the cladding layer of the sample using the optimized process parameter combination fluctuates slightly within 2.4%, which shows that the use of the optimization model constructed in this paper will not reduce the quality of the cladding layer while improving the energy efficiency and processing efficiency of laser cladding processing.

4 Conclusion

1) Use specific energy consumption to characterize the energy efficiency of laser cladding processing of mechanical parts. Based on the energy consumption characteristics of laser cladding processing of mechanical parts, a cladding processing energy efficiency model is constructed. The relevant parameters are obtained and the energy efficiency model is verified by using sample laser cladding tests. The results show that the average error is 3%, which proves that the constructed energy efficiency model is accurate and reliable.

2) Based on the energy efficiency model, a multi-objective optimization model for laser cladding processing of mechanical parts is constructed and the solution results are compared and analyzed, and the optimal process parameter combination that can simultaneously meet the highest energy efficiency and the shortest processing time is obtained: Plc=3062W, vs=11.2mm/s, Sp=49.8g/min.

3) Compared with the empirical parameters of laser cladding processing, the use of the optimized process parameter combination can reduce the specific energy consumption of the sample laser cladding process by 5.1% and the processing time by 9.5%, while ensuring the quality of the cladding layer.