Robot spraying quality analysis method based on improved double-beta paint film distribution model

Abstract

The invention relates to a robot spraying quality analysis method based on an improved double-beta paint film distribution model. The method comprises the following steps: creating a double-beta paint film distribution model based on a single-beta paint film distribution model; the spraying speed is introduced into the double-beta paint film distribution model, and a dynamic film thickness distribution model is obtained; the spraying height and the environment humidity are introduced; introducing spraying width pressure, atomization pressure and oil injection pressure into the film thickness distribution model under the dynamic condition of introducing spraying height and environment humidity to obtain a dynamic film thickness distribution model under multiple process parameters; using a least square method to fit and obtain undetermined parameters in the final dynamic film thickness distribution model under the multi-process parameters, and outputting a film thickness distribution equation when the precision and generalization performance of the final dynamic film thickness distribution model under the multi-process parameters meet preset requirements; otherwise, adjusting the parameter value of the fitting algorithm, and solving the corresponding parameter again. And performing quality analysis and prediction on the average paint film thickness and the paint film thickness standard deviation of the robot spraying process through the dynamic film thickness distribution model under multiple process parameters.

Description

Technical Field

[0001] The invention belongs to the technical field of robot spraying quality analysis, and in particular relates to a robot spraying quality analysis method based on an improved double-β paint film distribution model. Background Art

[0002] Gas turbine is a kind of efficient thermal engine. It is widely used in various military and civilian fields such as electrical production and aerospace due to its advantages such as high operating power, low pollution, fast start-up, high energy utilization rate and good reliability. Turbine blades are one of the most important core components in the design and manufacture of gas turbines. The surface quality of turbine blades directly affects the aerodynamic shape and conversion efficiency of the blades, and thus affects the operating efficiency and service life of the entire gas turbine. Due to the harsh working environment such as high temperature, high pressure and high speed, it is very easy to cause oxidation, thermal corrosion and thermal shock on the surface of gas turbine blades. Therefore, it is necessary to improve the blade quality and performance of gas turbine blades by spraying paint. Paint spraying of gas blades can not only significantly reduce the interference of the external environment on the blade quality and improve the service life of the blades, but also reduce surface friction and airflow resistance, improve aerodynamic shape and improve conversion efficiency by optimizing the design of surface coating. However, the working environment and work intensity of gas turbine blade spraying are harsh and high. Some coatings also contain toxic and harmful substances such as chromium trioxide. Traditional manual spraying operations not only cause great waste of labor costs and unstable product quality, but also pose great risks to the occupational health of operators. Therefore, it is necessary to upgrade the technology of gas turbine blade spraying process, use spraying robots to replace manual spraying operations, and build an automated gas turbine blade spraying production line. Compared with manual spraying operations, robot spraying has the advantages of improving production efficiency, reducing production costs, improving coating quality, saving spraying fuel, and reducing health risks.

[0003] The core task of using spray robots to build automated production lines is to refer to production experience, analyze and summarize the spraying process, summarize the influencing rules of the quality of the paint film sprayed on the robot's gas turbine blades, and then combine the robot's motion planning to achieve high-quality processing of gas turbine blades. How to construct a gas turbine blade quality analysis and evaluation model by analyzing and modeling the coupling effects of spraying process parameters, spraying device parameters, environmental parameters and other factors on the paint film quality of gas turbine blades is one of the pain points in the development of industry technology. At present, the main research direction of gas turbine blade quality analysis is to analyze the influence of different process parameters on the quality of gas turbine blades, construct a coating thickness distribution model for gas turbine blades under different process parameters, and then construct a mapping model through statistical or machine learning methods to achieve predictive analysis of paint film thickness.

[0004] In the prior art, some researchers have developed an off-line programming system that can complete film thickness prediction. The system can determine the spraying strategy and path by giving corresponding parameters. Through actual spraying experiment comparison, the model predicts a small deviation between the thickness and the actual thickness

[11] . Other researchers have conducted multiple sets of plane spraying experiments for spray gun speed and spray gun distance to study the film thickness distribution, and used the β model to simulate the film thickness distribution under different conditions. Some researchers have conducted experimental analysis on the influence of different spraying angles on coating distribution and the influence of spray gun speed on temperature and residual stress when the process parameters such as spray flow rate and spray gun distance remain unchanged. The experimental results show that adjusting the appropriate spraying angle and variable speed spraying can improve the spraying quality. Other researchers have used Fluent software to build a spraying simulation model, and based on the model, orthogonal simulations of different process parameters were performed. The elliptical double β model and genetic algorithm were used to fit the experimental data to obtain the final film thickness distribution formula. Some salt-baked researchers have used Ansys software to simulate the spray flow based on the DPM and TAB models, and analyzed the influence of molding air pressure on the air flow field and coating thickness distribution. Others have proposed a coating thickness distribution model based on electrostatic rotary cup uniform spraying, and analyzed the influence of characteristic points as various coating process parameters change through a multi-level, multi-factor orthogonal experimental group.

[0005] However, the above research on spraying process parameters and spray paint film thickness still has some research breakpoints. First, the current analysis of spraying process parameters is not comprehensive enough. Most of the analysis focuses on the four input parameters of spraying speed, spraying height, atomization pressure and injection pressure, while ignoring the effects of factors such as injection pressure based on paint characteristics and ambient humidity on paint film quality. Second, current research basically stays on single-pass spraying, while actual blade spraying operations often use multi-pass spraying. Therefore, the overlap distance between spraying trajectories also affects the overall paint film quality on the blade surface. Third, the research on paint film quality only stays on the prediction of paint film thickness, and has not formed a complete and standardized paint film quality analysis method, nor has it proposed clear paint film quality assessment standards.

[0006] Therefore, the present invention aims at the problem that there are still some research breakpoints in the research on spraying process parameters and spray paint film thickness, and proposes a single-pass paint film distribution model for gas turbine blades, and a prediction method for the average film thickness and film thickness difference of gas turbine blades based on the model, thereby forming a clear quality assessment standard and a standardized paint film quality analysis method for gas turbine blades, so as to solve the technical problems that the analysis of spraying process parameters is not comprehensive enough, the research basically stays on the research of single-pass spraying, and does not involve the multi-pass spraying method commonly used in blade spraying operations, and no complete and standardized paint film quality analysis method has been formed. Summary of the invention

[0007] The purpose of the present invention is to provide a robot spraying quality analysis method based on an improved double-β paint film distribution model, and a method for predicting the average film thickness and film thickness difference of gas turbine blades based on the model, so as to form a clear quality assessment standard and a standardized paint film quality analysis method for gas turbine blades, so as to solve the technical problems that the analysis of spraying process parameters is not comprehensive enough, the research basically stays on the research of single-pass spraying, and does not involve the multi-pass spraying method commonly used in blade spraying operations, and no complete and standardized paint film quality analysis method has been formed.

[0008] In order to solve the above technical problems, the technical solution adopted by the present invention is as follows:

[0009] The robot spraying quality analysis method based on the improved double-β paint film distribution model includes the following steps:

[0010] S1: Create a single β paint film distribution model;

[0011] S2: Create a double β paint film distribution model based on the single β paint film distribution model;

[0012] S3: introducing the spraying speed into the double-β paint film distribution model to obtain a dynamic film thickness distribution model;

[0013] S4: introducing the spraying height and the ambient humidity into the film thickness distribution model under the dynamic state;

[0014] S5: Introducing the dynamic film thickness distribution model under the spraying height and ambient humidity, introducing the spray width pressure, atomization pressure, and injection pressure to obtain the final dynamic film thickness distribution model under multiple process parameters;

[0015] S6: using the least square method to fit and obtain the undetermined parameters in the final dynamic film thickness distribution model under multiple process parameters, and judging whether the accuracy and generalization performance of the final dynamic film thickness distribution model under multiple process parameters meet the preset requirements, if so, outputting the film thickness distribution equation based on the fitting parameters at this time, and ending; if not, executing step S7;

[0016] S7: Adjust the fitting algorithm parameter values, re-calculate the corresponding parameters, and again determine whether the accuracy and generalization performance of the dynamic film thickness distribution model under the final multi-process parameters meet the preset requirements, until the accuracy and generalization performance of the dynamic film thickness distribution model under the final multi-process parameters meet the preset requirements.

[0017] Preferably, the expression of the single β paint film distribution model in step S1 is:

[0018]

[0019] Among them, β represents the shaping parameter, which determines the distribution shape of the model film thickness, q maxIt represents the maximum film thickness at the center point, q(x) represents the film thickness at a distance x from the center point, and r represents the spray radius of the spray gun.

[0020] Preferably, the specific process of creating the double β paint film distribution model based on the single β paint film distribution model in step S2 is:

[0021] S21: Based on the expression of the single β paint film distribution model, the film thickness distribution formula at the specified point A(x,y) is obtained:

[0022]

[0023] Among them, q max is the maximum film thickness growth rate value at the center point O of the paint elliptical distribution area image, and β1 is the shaping parameter value of the cross section in this direction;

[0024] S22: Calculate the film thickness growth rate at the x and y positions in the y direction. The calculation formula is as follows:

[0025]

[0026] This formula is the double-β paint film distribution model formula;

[0027] Among them, a and b represent the major axis and minor axis values ​​of the paint elliptical distribution area image respectively, q(x,y) represents the film thickness growth rate value at the x and y positions, and it is assumed that the x direction and the y direction have the same film thickness distribution morphology, that is, the β values ​​of each section in the same direction are the same, β1 and β2 are the film thickness shaping parameter values ​​in different directions, and their sizes are determined by the shaping air pressure process parameter value.

[0028] Preferably, in step S3, the specific process of introducing the spraying speed into the double-β paint film distribution model to obtain the dynamic film thickness distribution model is as follows:

[0029] S31: Calculate the distance swept by the spray gun. The specific formula is:

[0030]

[0031] S32: Calculate the time the spray gun sweeps through. The specific formula is as follows:

[0032]

[0033] S33: Calculate the accumulated coating thickness at any point O (x0, y0), that is, the integral of the time that the spray gun sprays over the point O:

[0034]

[0035] S34: Change the integration limit to obtain:

[0036]

[0037] S35: Let β1 = β2 = 2, and substitute After that, bring in The coating profile thickness distribution is obtained when the spray gun sprays at a uniform speed along the workpiece surface.

[0038] Preferably, the specific process of introducing the spraying height and the ambient humidity into the film thickness distribution model under the dynamic state in step S4 is as follows:

[0039] S41: Set the fog cone angle α in the Y-axis direction y Along the direction of the spray gun trajectory, the spray cone angle α in the X-axis direction x Perpendicular to the direction of movement of the spray gun, the long and short axes of the ellipse and the spray cone angle are obtained. The specific formula is:

[0040]

[0041] S42: Establish the relationship between the spraying distance d and the major axis and minor axis a, b of the elliptical spray spot:

[0042]

[0043] S43: Establish the relationship between the paint application rate ε and d:

[0044]

[0045] Among them, k ε is the paint rate coefficient, h is the spraying distance;

[0046] S44: When other spraying parameters are constant, the spraying distance does not affect the total amount of paint sprayed by the spray gun per unit time. The total amount of paint at any time is fixed, and the relationship is obtained:

[0047]

[0048] Among them, S(h1) and S(h2) are the areas of the coating along the longitudinal section obtained by single-pass spraying on the flat plate at spraying distances h1 and h2, respectively. The calculation formula is:

[0049]

[0050] Then we get:

[0051]

[0052] Preferably, the specific process of introducing the spray width pressure, atomization pressure, and fuel injection pressure in step S5 is as follows:

[0053] S51: Establish atomization pressure p w , spray pressure pf , injection pressure p i The relationship between them is as follows:

[0054]

[0055] S52: Establish injection pressure p i With the maximum coating thickness d max The relationship is:

[0056]

[0057] S53: Establish a multi-parameter coating growth model to obtain the maximum coating thickness within the elliptical spray spot:

[0058]

[0059] S54: Calculate the coating thickness of the spray gun spraying at a uniform speed along the workpiece surface:

[0060]

[0061] Among them, k x , k y , a1, b1, a2, b2 are unknown coefficients, K m is the coating growth rate coefficient, and K' is the generalization coefficient, which are obtained by fitting the experimental data.

[0062] Preferably, it also includes establishing an expression for the paint film thickness of two adjacent spraying operations in a single spraying section under a multi-pass spraying process:

[0063]

[0064] Among them, q1(x, v) represents the expression of film thickness distribution on the first path, q2(x, v) represents the expression of film thickness distribution on the second path. In multi-pass spraying operation, that is, the film thickness distribution in single-pass spraying operation obeys q1(x, v), and the overlapping area obeys q2(x, v).

[0065] Preferably, the method further includes establishing a paint film average film thickness prediction model under a single cross section:

[0066]

[0067] Where d is the spray track spacing, n is the number of spray tracks, x is the corresponding point of the spray section, and T(x) is the film thickness value at the corresponding point.

[0068] Preferably, the method further includes establishing a standard deviation model of the paint film thickness under a single cross section:

[0069]

[0070] Where N is the number of data points, x i is the cross-sectional point corresponding to N sampling data points, T(x i ) is a single x i The film thickness value at the location.

[0071] The beneficial effects of the present invention include:

[0072] The robot spraying quality analysis method based on the improved double β paint film distribution model provided by the present invention creates a double β paint film distribution model based on the single β paint film distribution model; introduces the spraying speed in the double β paint film distribution model to obtain the dynamic film thickness distribution model; introduces the spraying height and ambient humidity; introduces the spray width pressure, atomization pressure, and injection pressure into the dynamic film thickness distribution model under multiple process parameters; uses the least squares method to fit and obtain the undetermined parameters in the final dynamic film thickness distribution model under multiple process parameters, and the accuracy and generalization performance of the final dynamic film thickness distribution model under multiple process parameters meet the preset requirements to output the film thickness distribution equation; otherwise, adjust the fitting algorithm parameter value and re-obtain the corresponding parameters. The dynamic film thickness distribution model under multiple process parameters is used to realize the quality analysis and prediction of the average film thickness and the standard deviation of the paint film thickness of the robot spraying process.

[0073] First, a single β paint film distribution model is created, and the applicable film thickness distribution model is obtained by adjusting the β value according to different processes. Since the simulated spray width range of the single β distribution model formula is an ideal conical distribution, in actual spraying operations, additional gas channels are set on both sides of the automatic spray gun to adjust the spray shape in order to facilitate adjustment and control of the coating shape, resulting in the final workpiece surface spray width profile being elliptical rather than circular. Therefore, considering the influence of the spray width pressure on the shape of the mist cone in actual spraying and air spraying, the elliptical double β paint film distribution model is used for modeling and analysis, which is more in line with the elliptical spray plate shape generated during the actual spraying operation, and more in line with the actual production operation, so as to achieve modeling analysis and accurate prediction of the law of process parameter changes.

[0074] Secondly, since the process parameters will affect the quality of paint film during the robot's spraying process, the spraying speed, spraying height, spray width pressure, atomization pressure, and injection pressure are set as the input parameters of the improved double-β paint film distribution model in combination with the robot spraying operation experiment of gas turbine blades. Then, the influence of environmental factors is introduced into the spraying height to ensure that the model is more in line with the actual spraying effect.

[0075] Thirdly, since a single-pass spraying operation cannot meet the process requirements of paint film spraying, multiple-pass spraying operations are required to ensure that the paint film quality on the blade surface meets the standards. During single-pass spraying, the paint film thickness distribution on the spraying path presents a normal distribution with thicker in the middle and thinner on both sides. When multiple-pass spraying operations are performed at the same time, there will be overlap in the spraying range, and the paint film thickness in the overlapping area will increase due to repeated spraying. Therefore, the present invention combines the spraying process requirements and analyzes and optimizes the overlap distance during multiple-pass spraying operations to ensure the consistency of the paint film thickness between the single-track spraying area and the overlap spraying operation area, thereby ensuring the film thickness and paint film uniformity of the paint film cross section of the gas turbine blade.

[0076] Finally, although in the actual spraying process, the spraying process parameters will fluctuate due to the change of the spray gun or robot state, which will lead to certain fluctuations in the paint film quality in the paint film sections at different positions during the single-pass spraying process, the average film thickness prediction model and the paint film thickness standard deviation model under a single cross section can achieve reasonable prediction and accurate analysis of the paint film quality under different paint film process parameters during the robot spraying operation of the gas turbine blade. Based on the double-β paint film distribution model, the present invention constructs a complete paint film quality analysis method for the robot spraying operation of the gas turbine blade. Based on this method, it not only provides standard and reliable evaluation indicators and analysis schemes for the iteration of the gas turbine blade process, but also significantly improves the paint film process analysis method and increases the iteration speed of the paint film process. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] Figure 1 It is a schematic flow chart of the robot spraying quality analysis method based on the improved double-β paint film distribution model of the present invention.

[0078] Figure 2 This is a schematic diagram showing different film thickness distribution forms according to different β value distribution models of the present invention.

[0079] Figure 3 It is a schematic diagram of the dynamic spraying process of the spray gun of the present invention.

[0080] Figure 4 Schematic diagram of the relationship between the major axis and minor axis of the elliptical spray spot and the spraying height of the present invention.

[0081] Figure 5 Schematic diagram of the overlapping distance of adjacent tracks of the present invention.

[0082] Figure 6 Schematic diagram of film thickness distribution with overlapping distance. DETAILED DESCRIPTION

[0083] The following is combined with Figure 1 to Figure 6 The present invention is further described in detail:

[0084] Example 1

[0085] See attached Figure 1 As shown, the robot spraying quality analysis method based on the improved double-β paint film distribution model includes the following steps:

[0086] S1: Create a single β paint film distribution model;

[0087] S2: Create a double β paint film distribution model based on the single β paint film distribution model. By creating a single β paint film distribution model and adjusting the β value according to different processes, a suitable film thickness distribution model is obtained. Since the simulated spray width range of the single β distribution model formula is an ideal conical distribution, in actual spraying operations, in order to facilitate adjustment and control of the coating shape, additional gas channels are set on both sides of the automatic spray gun to adjust the spray shape, resulting in the final workpiece surface spray width profile being elliptical rather than circular. Therefore, taking into account the influence of the spray width pressure on the shape of the mist cone in air spraying in actual spraying, the use of the elliptical double β paint film distribution model for modeling and analysis is more in line with the elliptical spray plate shape generated during the actual spraying operation, and is more in line with the actual production operation conditions, and the modeling analysis and accurate prediction of the law of process parameter changes can be achieved.

[0088] S3: introducing the spraying speed into the double-β paint film distribution model to obtain a dynamic film thickness distribution model;

[0089] S4: introducing the spraying height and the ambient humidity into the film thickness distribution model under the dynamic state;

[0090] S5: Introducing the dynamic film thickness distribution model under spraying height and ambient humidity, the spray width pressure, atomization pressure and fuel injection pressure are introduced to obtain the final dynamic film thickness distribution model under multiple process parameters; Since the process parameters will affect the film quality of the paint during the robot's spraying process, the spraying speed, spraying height, spray width pressure, atomization pressure and fuel injection pressure are set as the input parameters of the improved double-β paint film distribution model in combination with the robot spraying operation experiment of the gas turbine blades, and then the influence of environmental factors is introduced in the spraying height to ensure that the model is more in line with the actual spraying effect.

[0091] S6: using the least square method to fit and obtain the undetermined parameters in the final dynamic film thickness distribution model under multiple process parameters, and judging whether the accuracy and generalization performance of the final dynamic film thickness distribution model under multiple process parameters meet the preset requirements, if so, outputting the film thickness distribution equation based on the fitting parameters at this time, and ending; if not, executing step S7;

[0092] S7: Adjust the fitting algorithm parameter values, re-calculate the corresponding parameters, and again determine whether the accuracy and generalization performance of the dynamic film thickness distribution model under the final multi-process parameters meet the preset requirements, until the accuracy and generalization performance of the dynamic film thickness distribution model under the final multi-process parameters meet the preset requirements.

[0093] In this embodiment, the expression of the single β paint film distribution model in step S1 is:

[0094]

[0095] Among them, β represents the shaping parameter, which determines the distribution shape of the model film thickness, q max represents the maximum film thickness at the center point, q(x) represents the film thickness at a distance x from the center point, and r represents the spray radius of the spray gun. Figure 2 As shown in the figure, from the outside to the inside, they are the film thickness distribution forms when β is 1.5, 2, 4, and 8. It can be seen that with different β values, the distribution model presents different film thickness distribution forms. When β is 1.5, the film thickness profile is semi-elliptical, when β is 2, the film thickness profile becomes a parabola, and when β is larger, the profile shape becomes a Gaussian distribution. It can be seen that the β value affects the sharpness and flatness of this distribution. As the β value increases, this trend also increases. Therefore, in practical applications, the applicable film thickness distribution model can be obtained by adjusting the β value according to different processes.

[0096] Since the gas turbine blade spray gun generally uses air spraying for spraying operations, in actual air spraying production, in order to make the paint mist form a more uniform paint film on the surface of the workpiece, it is necessary to draw compressed air on both sides of the spray gun mouth to flatten the paint mist cone, forming a fan-shaped body with a certain thickness on the plane perpendicular to the axis of the spray gun to form an elliptical paint film area, which is quite different from the circular paint film area formed by the right cone approximated by the β model. In comprehensive consideration of the influence of the spray pressure on the shape of the mist cone in air spraying, the elliptical double β curve model, i.e., the double β paint film distribution model, is used for modeling and analysis, which is more in line with the elliptical spray plate shape generated during the actual spraying operation.

[0097] The specific process of creating a double β paint film distribution model based on the single β paint film distribution model in step S2 is:

[0098] S21: Based on the expression of the single β paint film distribution model, the film thickness distribution formula at the specified point A(x,y) is obtained:

[0099]

[0100] Among them, q max is the maximum film thickness growth rate value at the center point O of the paint elliptical distribution area image, and β1 is the shaping parameter value of the cross section in this direction;

[0101] S22: Calculate the film thickness growth rate at the x and y positions in the y direction. The calculation formula is as follows:

[0102]

[0103] This formula is the double-β paint film distribution model formula.

[0104] Among them, a and b represent the major axis and minor axis values ​​of the paint elliptical distribution area image, respectively, q(x,y) represents the film thickness growth rate value at the x and y positions, and it is assumed that the x direction and the y direction have the same film thickness distribution morphology, that is, the β values ​​of each section with the same direction are the same, β1 and β2 are the film thickness shaping parameter values ​​in different directions, and their sizes are determined by the shaping air pressure process parameter value. It can be seen that the single β distribution model formula simulates the spray width range as an ideal conical distribution. However, in actual spraying operations, in order to facilitate adjustment and control of the coating shape, additional gas channels will be set on both sides of the automatic spray gun to adjust the spray shape, resulting in the final workpiece surface spray width profile being elliptical rather than circular. Therefore, the double β distribution model formula is more in line with actual production operations. The present invention derives the film thickness distribution formula for spraying operations based on the double β paint film distribution model.

[0105] Example 2

[0106] When using the traditional double beta distribution model to predict the distribution of paint film thickness, the model lacks process parameter items, so it cannot reflect the actual effect of process parameter changes on the distribution of paint film thickness, and only reflects a high prediction accuracy under specific process parameters. As a result, when facing complex process conditions, the application scope and prediction accuracy of the model cannot meet the requirements. Therefore, in order to improve the applicability and generalization ability of the model, it is necessary to introduce process parameters into the traditional double beta curve model, construct an improved double beta curve distribution prediction model based on process parameters, and realize modeling analysis and accurate prediction of the law of process parameter changes.

[0107] Therefore, in the actual spraying process, the spray gun does not need to be stationary, but needs to be moved to coordinate with the spraying process. And the best paint film effect is when the spray gun moves at a constant speed with the short axis direction of the spray spot as the spraying direction. Figure 3 shown.

[0108] On the basis of Example 1, in step S3, the spraying speed is introduced into the double-β paint film distribution model, and the specific process of obtaining the dynamic film thickness distribution model is as follows:

[0109] S31: Calculate the distance swept by the spray gun. The specific formula is:

[0110]

[0111] S32: Calculate the time the spray gun sweeps through. The specific formula is as follows:

[0112]

[0113] S33: Calculate the accumulated coating thickness at any point O (x0, y0), that is, the integral of the time that the spray gun sprays over the point O:

[0114]

[0115] S34: Change the integration limit to obtain:

[0116]

[0117] S35: Let β1 = β2 = 2, and substitute After that, bring in The coating profile thickness distribution is obtained when the spray gun sprays at a uniform speed along the workpiece surface.

[0118] During the spraying process, the paint is atomized by the spray gun and sprayed out in a cone-shaped mist, forming an elliptical spray spot on the surface of the object. Figure 4 As shown, since the major axis and minor axis of the elliptical spray spot are directly related to the spraying height, the concept of fog cone angle is set to describe the changing relationship between the major axis and minor axis and the spraying height.

[0119] The specific process of introducing the spraying height and the ambient humidity into the film thickness distribution model under the dynamic state in step S4 is as follows:

[0120] S41: Set the fog cone angle α in the Y-axis direction y Along the direction of the spray gun trajectory, the spray cone angle α in the X-axis direction x Perpendicular to the direction of movement of the spray gun, the long and short axes of the ellipse and the spray cone angle are obtained. The specific formula is:

[0121]

[0122] S42: The spraying distance d is a parameter directly controlled by the motion trajectory of the end of the robot arm and has nothing to do with the spray gun settings. When the spray cone angle is constant, the major axis and minor axis a, b of the elliptical spray spot become larger and larger with the increase of the spraying distance. The relationship between the spraying distance d and the major axis and minor axis a, b of the elliptical spray spot is established:

[0123]

[0124] S43: The spraying distance and the humidity of the spraying environment will affect the paint application rate ε on the surface. The increase in the spraying distance will result in some paint not being able to be sprayed on the workpiece surface, that is, the lower the paint application rate. Within a certain spraying distance range, the relationship between the paint application rate ε and d is established as:

[0125]

[0126] Among them, k ε is the paint rate coefficient, h is the spraying distance;

[0127] S44: When other spraying parameters are constant, the spraying distance does not affect the total amount of paint sprayed by the spray gun per unit time. The total amount of paint at any time is fixed, and the relationship is obtained:

[0128]

[0129] Among them, S(h1) and S(h2) are the areas of the coating along the longitudinal section obtained by single-pass spraying on the flat plate at spraying distances h1 and h2, respectively. The calculation formula is:

[0130]

[0131] Then we get:

[0132]

[0133] In the actual spraying process, by adjusting the atomization pressure p w and spray pressure p f Change the fan shape and atomization effect of the paint. The cone angle in the X-axis direction changes with the atomization pressure p w and spray pressure p f The fog cone angle in the Y-axis direction increases with the increase of the atomization pressure p w The spray pressure p f The injection pressure has almost no effect on the spray cone angle. The injection pressure only determines the content of droplets in the spray cone. When the injection pressure is p i When α increases, x and α y The specific process of introducing the spray width pressure, atomization pressure and injection pressure in step S5 is as follows:

[0134] S51: Establish atomization pressure p w , spray pressure p f , injection pressure p i The relationship between them is as follows:

[0135]

[0136] S52: Although the injection pressure p i It does not affect the spray cone angle, but the injection pressure p i Directly affects the amount of paint deposited on the surface of the object, p iAs the value increases, the coating deposition thickness increases. Assuming that the coating cross section still obeys the same β distribution, the injection pressure p is established by fitting the experimental data. i With the maximum coating thickness d max The relationship is:

[0137]

[0138] S53: Establish a multi-parameter coating growth model to obtain the maximum coating thickness within the elliptical spray spot:

[0139]

[0140] S54: Calculate the coating thickness of the spray gun spraying at a uniform speed along the workpiece surface:

[0141]

[0142] Among them, k x , k y , a1, b1, a2, b2 are unknown coefficients, K m is the coating growth rate coefficient, and K' is the generalization coefficient, which are obtained by fitting the experimental data.

[0143] Based on the five process parameters of spraying speed, spraying height, spray width pressure, atomization pressure, injection pressure and ambient humidity, a multi-process parameter coating thickness accumulation model for a single-pass spraying trajectory is established for the robot spraying operation requirements of gas turbine blades, and the coefficients of the formula are fitted and calculated in combination with experiments.

[0144] Example 3

[0145] Due to the paint film quality requirements and spray gun structure limitations of the robot spraying operation of the gas turbine blade, a single-pass spraying operation cannot meet the process requirements of the paint film spraying, so multiple-pass spraying operations are required to ensure that the paint film quality on the blade surface meets the standards. During single-pass spraying, the paint film thickness distribution on the spraying path presents a normal distribution with thicker in the middle and thinner on both sides. When multiple-pass spraying operations are performed at the same time, there will be overlaps in the spraying range, and the paint film thickness in the overlapping area increases due to repeated spraying. Therefore, it is necessary to analyze and optimize the overlap distance during multiple-pass spraying operations in combination with the spraying process requirements to ensure the consistency of the paint film thickness between the single-track spraying area and the overlapped spraying operation area, thereby ensuring the film thickness and uniformity of the paint film cross section of the gas turbine blade. During multiple-pass spraying operations, all single-pass spraying paint film thickness distribution models are set to be the same, and the paint film overlap distance is the same, so the paint film thickness analysis under the entire cross section can be simplified to the film thickness analysis of two adjacent paint film tracks. Therefore, taking two adjacent spraying operations as an example, the spraying overlap is analyzed. The overlap distance of adjacent tracks can be found in Figure 5 As shown, the overlap distance and film thickness distribution refer to Figure 6 shown.

[0146] On the basis of Example 1 or Example 2, based on the requirements of the spraying operation, the present invention sets parameters for the overlap problem of the robot spraying process of the gas turbine blade. First, the moving center Y axis of the elliptical spraying trajectory of the first spraying operation is set to coincide with the upper edge of the paint film cross section, that is, the origin of the single-pass paint film thickness distribution model is located at the upper edge of the paint film cross section. Second, the area with a middle distance d within the elliptical spraying range is set as the overlap area. The gas turbine blade is sprayed once in the single-pass spraying area and is repeatedly sprayed in the overlap area. The paint film thickness of the two adjacent spraying operations under a single spraying cross section under the multi-pass spraying process is expressed as:

[0147]

[0148] Among them, q1(x, v) represents the expression of film thickness distribution on the first path, q2(x, v) represents the expression of film thickness distribution on the second path. In multi-pass spraying operation, that is, the film thickness distribution in single-pass spraying operation obeys q1(x, v), and the overlapping area obeys q2(x, v). The film thickness distribution under a single cross section presents a segmented cyclic distribution, and the film thickness in the single-track spraying area and the overlapping area as well as the uniformity of the film thickness in different areas can be optimized by changing the track spacing d.

[0149] In this embodiment, a paint film average thickness prediction model under a single cross section is also established:

[0150]

[0151] Where d is the spray track spacing, n is the number of spray tracks, x is the corresponding point of the spray section, and T(x) is the film thickness value at the corresponding point.

[0152] For the robot spraying operation of gas turbine blades, the core of the paint film quality analysis is whether the paint film thickness meets the standard and whether the paint film uniformity meets the requirements. And due to the coupling influence of the spraying process parameter setting and the overlap of the spraying trajectory, the paint film distribution of the paint film cross section during the spraying operation will have a greater impact on the paint film quality. Therefore, it is necessary to set a paint film sampling section in the processing direction of the gas turbine blade, and ensure the paint film quality on the blade surface by ensuring the paint film thickness and paint film uniformity on the section. Based on the single-pass paint film thickness distribution model and the paint film thickness expression of two adjacent spraying operations under a single spraying section, the present invention constructs a paint film average film thickness prediction model and a paint film thickness standard deviation model under a single cross section of the gas turbine blade with a limited sampling number, thereby realizing approximate analysis and prediction of the gas turbine blade cross section.

[0153] Establish the standard deviation model of paint film thickness under a single cross section:

[0154]

[0155] Where N is the number of data points, x i is the cross-sectional point corresponding to N sampling data points, T(x i ) is a single x i The film thickness value at the position. Since the formula of the single-layer paint film thickness distribution model is relatively complex, the integral is not easy to solve directly. However, in theory, the average film thickness can be calculated by numerical integration. By discretizing the range of coating thickness, and then multiplying the thickness of each discrete point by the corresponding probability density value to sum it up, the average film thickness can be approximately calculated.

[0156] Although in the actual spraying operation, the changes in the state of the spray gun or robot will cause the spraying process parameters to fluctuate during the spraying operation, which will in turn cause certain fluctuations in the paint film quality in the paint film sections at different positions during a single-pass spraying process, but based on the paint film average film thickness prediction model and the paint film thickness standard deviation model under a single cross-section, the present invention can achieve reasonable prediction and accurate analysis of the paint film quality under different paint film process parameters during the robot spraying operation of gas turbine blades. Based on all the above mathematical models, the present invention constructs a complete paint film quality analysis method for the robot spraying operation of gas turbine blades. This method not only provides standard and reliable evaluation indicators and analysis schemes for the iteration of gas turbine blade processes, but also significantly improves the paint film process analysis method and increases the speed of paint film process iteration.

[0157] In summary, the robot spraying quality analysis method based on the improved double β paint film distribution model provided by the present invention creates a single β paint film distribution model, and obtains a suitable film thickness distribution model by adjusting the β value according to different processes. Since the single β distribution model formula simulates a spray width range of an ideal conical distribution, in order to facilitate adjustment and control of the coating shape in actual spraying operations, additional gas channels are set on both sides of the automatic spray gun to adjust the spray shape, resulting in the final workpiece surface spray width profile being elliptical rather than circular. Therefore, taking into account the influence of the spray width pressure on the shape of the mist cone in air spraying during actual spraying, the use of the elliptical double β paint film distribution model for modeling and analysis is more in line with the elliptical spray plate shape generated during the actual spraying operation, and is more in line with the actual production operation conditions, thereby achieving modeling analysis and accurate prediction of the law of process parameter changes. Since the process parameters will affect the quality of paint film during the robot's spraying process, the spraying speed, spraying height, spray width pressure, atomization pressure and injection pressure are set as the input parameters of the improved double-β paint film distribution model in combination with the robot spraying operation experiment of the gas turbine blade. Then, the influence of environmental factors is introduced in the spraying height to ensure that the model is more in line with the actual spraying effect.

[0158] Since a single-pass spraying operation cannot meet the process requirements of paint film spraying, multiple-pass spraying operations are required to ensure that the paint film quality on the blade surface meets the standards. When single-pass spraying is performed, the paint film thickness distribution on the spraying path presents a normal distribution with thicker in the middle and thinner on both sides. When multiple-pass spraying operations are performed at the same time, there will be overlap in the spraying range, and the paint film thickness in the overlapping area will increase due to repeated spraying. Therefore, the present invention combines the spraying process requirements and analyzes and optimizes the overlap distance during multiple-pass spraying operations to ensure the consistency of the paint film thickness in the single-track spraying area and the overlap spraying operation area, thereby ensuring the film thickness and paint film uniformity of the paint film cross section of the gas turbine blade. Although in the actual spraying process, the spraying process parameters fluctuate due to the change of the spray gun or robot state, which in turn causes certain fluctuations in the paint film quality in the paint film sections at different positions during a single spraying process, the present invention realizes the reasonable prediction and accurate analysis of the paint film quality under different paint film process parameters during the robot spraying operation of the gas turbine blade based on the average film thickness prediction model and the paint film thickness standard deviation model under a single cross section. Based on the double-β paint film distribution model, the present invention constructs a complete paint film quality analysis method for the robot spraying operation of gas turbine blades. Based on this method, it not only provides standard and reliable evaluation indicators and analysis schemes for the process iteration of gas turbine blades, but also significantly improves the paint film process analysis method and increases the speed of paint film process iteration.

Claims

1. A robot spraying quality analysis method based on an improved double-β paint film distribution model, characterized in that: The following steps are involved: S1: Create a single β paint film distribution model; S2: Create a double β paint film distribution model based on the single β paint film distribution model; S3: introducing the spraying speed into the double-β paint film distribution model to obtain a dynamic film thickness distribution model; S4: introducing the spraying height and the ambient humidity into the film thickness distribution model under the dynamic state; S5: Introducing the dynamic film thickness distribution model under the spraying height and ambient humidity, introducing the spray width pressure, atomization pressure, and injection pressure to obtain the final dynamic film thickness distribution model under multiple process parameters; S6: using the least square method to fit and obtain the undetermined parameters in the final dynamic film thickness distribution model under multiple process parameters, and judging whether the accuracy and generalization performance of the final dynamic film thickness distribution model under multiple process parameters meet the preset requirements, if so, outputting the film thickness distribution equation based on the fitting parameters at this time, and ending; If not, execute step S7; S7: Adjust the fitting algorithm parameter values, re-calculate the corresponding parameters, and again determine whether the accuracy and generalization performance of the dynamic film thickness distribution model under the final multi-process parameters meet the preset requirements, until the accuracy and generalization performance of the dynamic film thickness distribution model under the final multi-process parameters meet the preset requirements.

2. The robot spraying quality analysis method based on the improved double-β paint film distribution model according to claim 1 is characterized in that: The expression of the single β paint film distribution model in step S1 is: Among them, β represents the shaping parameter, which determines the distribution shape of the model film thickness, q max It represents the maximum film thickness at the center point, q(x) represents the film thickness at a distance x from the center point, and r represents the spray radius of the spray gun.

3. The robot spraying quality analysis method based on the improved double-β paint film distribution model according to claim 1 is characterized in that: The specific process of creating a double β paint film distribution model based on the single β paint film distribution model in step S2 is: S21: Based on the expression of the single β paint film distribution model, the film thickness distribution formula at the specified point A(x,y) is obtained: Among them, q max is the maximum film thickness growth rate value at the center point O of the paint elliptical distribution area image, and β1 is the shaping parameter value of the cross section in this direction; S22: Calculate the film thickness growth rate at the x and y positions in the y direction. The calculation formula is as follows: This formula is the double-β paint film distribution model formula; Among them, a and b represent the major axis and minor axis values ​​of the paint elliptical distribution area image respectively, q(x,y) represents the film thickness growth rate value at the x and y positions, and it is assumed that the x direction and the y direction have the same film thickness distribution morphology, that is, the β values ​​of each section in the same direction are the same, β1 and β2 are the film thickness shaping parameter values ​​in different directions, and their sizes are determined by the shaping air pressure process parameter value.

4. The robot spraying quality analysis method based on the improved double-β paint film distribution model according to claim 1 is characterized in that: In step S3, the specific process of introducing the spraying speed into the double-β paint film distribution model and obtaining the dynamic film thickness distribution model is as follows: S31: Calculate the distance swept by the spray gun. The specific formula is: S32: Calculate the time the spray gun sweeps through. The specific formula is as follows: S33: Calculate the accumulated coating thickness at any point O (x0, y0), that is, the integral of the time that the spray gun sprays over the point O: S34: Change the integration limit to obtain: S35: Let β1 = β2 = 2, and substitute After that, bring in The coating profile thickness distribution is obtained when the spray gun sprays at a uniform speed along the workpiece surface.

5. The robot spraying quality analysis method based on the improved double-β paint film distribution model according to claim 1 is characterized in that: The specific process of introducing the spraying height and the ambient humidity into the film thickness distribution model under the dynamic state in step S4 is as follows: S41: Set the fog cone angle α in the Y-axis direction y Along the direction of the spray gun trajectory, the spray cone angle α in the X-axis direction x Perpendicular to the direction of movement of the spray gun, the long and short axes of the ellipse and the spray cone angle are obtained. The specific formula is: S42: Establish the relationship between the spraying distance d and the major axis and minor axis a, b of the elliptical spray spot: S43: Establish the relationship between the paint application rate v and d: Among them, k ε is the paint rate coefficient, h is the spraying distance; S44: When other spraying parameters are constant, the spraying distance does not affect the total amount of paint sprayed by the spray gun per unit time. The total amount of paint at any time is fixed, and the relationship is obtained: Among them, S(h1) and S(h2) are the areas of the coating along the longitudinal section obtained by single-pass spraying on the flat plate at spraying distances h1 and h2, respectively. The calculation formula is: Then we get:

6. The robot spraying quality analysis method based on the improved double-β paint film distribution model according to claim 1 is characterized in that: The specific process of introducing the spray width pressure, atomization pressure, and fuel injection pressure in step S5 is as follows: S51: Establish atomization pressure p w , spray pressure p f , injection pressure p i The relationship between them is as follows: S52: Establish injection pressure p i With the maximum coating thickness d max The relationship is: S53: Establish a multi-parameter coating growth model to obtain the maximum coating thickness within the elliptical spray spot: S54: Calculate the coating thickness of the spray gun spraying at a uniform speed along the workpiece surface: Among them, k x , k y , a1, b1, a2, b2 are unknown coefficients, K m is the coating growth rate coefficient, and K' is the generalization coefficient, which are obtained by fitting the experimental data.

7. The robot spraying quality analysis method based on the improved double-β paint film distribution model according to claim 1 is characterized in that: It also includes the establishment of the expression for the film thickness of two adjacent spraying operations under a single spraying section under a multi-pass spraying process: Among them, q1(x, v) represents the expression of film thickness distribution on the first path, q2(x, v) represents the expression of film thickness distribution on the second path. In multi-pass spraying operation, that is, the film thickness distribution in single-pass spraying operation obeys q1(x, v), and the overlapping area obeys q2(x, v).

8. The robot spraying quality analysis method based on the improved double-β paint film distribution model according to claim 1 is characterized in that: It also includes the establishment of a model to predict the average film thickness of the paint film under a single cross section: Where d is the spray track spacing, n is the number of spray tracks, x is the corresponding point of the spray section, and T(x) is the film thickness value at the corresponding point.

9. The robot spraying quality analysis method based on the improved double-β paint film distribution model according to claim 1 is characterized in that: It also includes the establishment of a standard deviation model of the paint film thickness under a single cross section: Where N is the number of data points, x i is the cross-sectional point corresponding to N sampling data points, T(x i ) is a single x i The film thickness value at the location.

Images