Water-jet guided laser processing progress visual identification method and system

Abstract

The invention provides a water-jet guided laser processing progress visual identification method, which comprises the following steps that a simulation model is established, a water-jet guided laser processing process is simulated, simulation results at different processing times are obtained, and the simulation results comprise an included angle formed by a hole wall and a processing depth; images in the water-jet guided laser machining process are collected, the shape contour of a spark area is extracted, the angle and the machining depth of laser sparks are calculated, and a visual result is obtained; and comparing a visualization result with a simulation result at the same processing time, and verifying the correctness of visualization identification. According to the method, the relation that the spark included angle and the depth of the machining process change along with time in the machining process is obtained, the machining progress is automatically recognized by observing the spark angle, then the machining efficiency and the finished product quality are improved, comparison verification is conducted through a simulation model, and therefore the machining progress recognition accuracy is greatly improved.

Description

Technical Field

[0001] The present invention relates to the field of water-guided laser technology, and in particular to a method and system for visualizing depth measurement of water-guided laser processing. Background Art

[0002] Water-guided laser processing technology, also known as laser microjet technology or micro-water jet laser technology, is an advanced laser processing method. The basic principle of water-guided laser processing technology is to use water as the guiding medium and cooling medium of the laser beam; when the laser beam is shot into the micro-water jet at a suitable angle, due to the difference in optical characteristics between water and air, the laser beam can be totally reflected at the interface between water and air, thereby being transmitted along the path of the water jet; this transmission method not only improves the focusing ability and energy density of the laser beam, but also effectively reduces the heat-affected zone during the processing through the cooling effect of the water flow, thereby avoiding deformation of the material or microcracks due to high temperature.

[0003] The water-conducting laser processing device and processing system with the publication number CN111098043A include a beam adjusting device, a focusing lens and a coupling water cavity; the beam adjusting device is used to adjust the position and angle of the laser; the focusing lens is used to focus the laser adjusted by the beam adjusting device and then inject it into the coupling water cavity; the coupling water cavity is used to transmit the focused laser along the outgoing water column at the water outlet of the coupling water cavity, so as to use the laser in the outgoing water column to process the workpiece; the focus of the laser in the outgoing water column deviates from the axis of the outgoing water column. A water-conducting laser processing system including the above device is also disclosed; the beam adjusting device is used to realize that the focus position of the laser in the outgoing water column deviates from the axis of the outgoing water column and is fixedly set or moves in a circle along the edge of the outgoing water column, so that the power distribution of the laser in the outgoing water column is uniform.

[0004] The existing water-conducting laser processing device cannot automatically identify the processing progress by observing the spark angle during laser processing, thereby reducing the processing efficiency and the quality of the finished product. Summary of the invention

[0005] In view of this, the present invention proposes a water-guided laser processing visualized depth measurement method and system, which can automatically identify the processing progress by observing the spark angle, thereby improving the processing efficiency and the quality of the finished product.

[0006] The technical solution of the present invention is implemented as follows: In a first aspect, the present invention provides a method for visually identifying the progress of water-conducting laser processing, the method comprising the following steps:

[0007] S1, establishing a simulation model, simulating the water-conducting laser processing process, and obtaining simulation results under different processing times, wherein the simulation results include the angle formed by the hole wall and the processing depth;

[0008] S2, collects images during the water-conducting laser processing, extracts the shape contour of the spark area, and calculates the angle of the laser spark and the processing depth to obtain a visual result;

[0009] S3, compare the visualization results with the simulation results of the same processing time to verify the correctness of the visualization recognition.

[0010] On the basis of the above technical solution, preferably, step S1 includes the following sub-steps:

[0011] S11, constructing a 3D geometric model of water-guided laser processing through modeling tools;

[0012] S12, meshing the three-dimensional geometric model using a meshing tool body;

[0013] S13, setting material parameters and laser processing parameters and selecting a Gaussian heat source model;

[0014] S14, setting boundary conditions;

[0015] S15, calculating and obtaining the laser heat flux density according to the material parameters and the laser processing parameters, and setting a corresponding convection heat transfer coefficient based on the cooling of the workpiece by the water jet fluid;

[0016] S16, solving and calculating the temperature field to obtain the temperature field distribution result;

[0017] S17, judging whether the temperature of each unit in the temperature field distribution result is greater than the melting point of the material, removing the units that have reached the melting point of the material; screening and judging the units that have not reached the melting point of the material;

[0018] S18, judging whether the processing is finished according to the simulation results and the preset processing standard, if the processing end condition is not met, returning to step S17 to re-judge until the processing end condition is met.

[0019] On the basis of the above technical solution, preferably, the boundary conditions include workpiece temperature initial conditions, heat flux density boundary conditions and convection heat transfer initial conditions, wherein:

[0020] The expression of the initial condition of the workpiece temperature is:

[0021] T=T *

[0022] Where T* is the initial temperature of the workpiece at a specific time point;

[0023] The expression of the heat flux boundary condition is:

[0024] {q} T {η}=-q *

[0025] Where, {q} T is the heat flux density vector, {η} is the unit vector of the outer normal of the boundary surface, and -q* is the specific heat flux density value on the boundary, which is a constant;

[0026] The expression of the initial condition of convective heat transfer is:

[0027] {q} T {η}=h f (T s -T B )

[0028] In the formula, h f is the convective heat transfer coefficient; T s is the wall temperature; T B is the fluid temperature.

[0029] On the basis of the above technical solution, preferably, in step S15, the laser heat flux density is calculated according to the material parameters and the laser processing parameters, and the expression is:

[0030] I(t)=I0·α

[0031] I0(t)=0.865P0(t) / (πd 2 / 4)

[0032] Where α is the material absorption coefficient; P0 is the output power of the laser; d is the spot diameter, and I is the heat flux density.

[0033] On the basis of the above technical solution, preferably, in step S15, a corresponding convection heat transfer coefficient is set based on the cooling of the workpiece by the water jet fluid, which includes the following sub-steps:

[0034] The water jet fluid flows over the workpiece surface and undergoes convective heat exchange with it. The expression is:

[0035] Q h =h f (T-T0)

[0036] In the formula, Q h is the heat flux density of convective heat transfer; T is the current temperature; T0 is the temperature at the previous moment; h f is the convective heat transfer coefficient;

[0037] The expression of convective heat transfer coefficient is:

[0038] h f = Nuλ / R

[0039] Where Nu is the Nusselt number; λ is the thermal conductivity of the fluid, λ = 0.6; R is the characteristic length;

[0040] Based on the simulation of water-guided laser processing of flat plates, the empirical formula for Nusselt number is:

[0041] Nu=0.745Re 1 / 2 Pr 1 / 3

[0042] Where Pr is the Prandtl number, Re is the Reynolds number,

[0043] The expression of Prandtl number Pr is:

[0044]

[0045] In the formula, c pw is the specific heat of water; μ w is the hydrodynamic viscosity; k w is the thermal conductivity of water;

[0046] The Reynolds number Re expression is:

[0047]

[0048] Where ρ is the density of the fluid; y is the velocity of the fluid flowing through the cross section; l is the characteristic length of the fluid flowing through the cross section; γ is the kinematic viscosity coefficient of the fluid.

[0049] On the basis of the above technical solution, preferably, the temperature field is solved and calculated in step S16 to obtain the temperature field distribution result, which includes the following sub-steps:

[0050] The finite element equation of the transient temperature field is established as follows:

[0051] [C]{T}+[K]{T}={Q}

[0052] Where, [C] is the non-steady-state temperature matrix; {T} is the temperature vector; [K] is the temperature field stiffness matrix; {Q} is the thermal load vector;

[0053] The difference calculation of numerical calculation is transformed into the solution of steady-state temperature field, and the expression is:

[0054]

[0055] Convert to the expression:

[0056]

[0057] Where θ is the transient comprehensive coefficient; Δt is the time difference from the previous moment to the current moment, Δt = t n+1 -t n ; {T n} is the node at t n The temperature value matrix at the moment; For the node at t n The matrix of derivative values ​​of the temperature at time instant.

[0058] On the basis of the above technical solution, preferably, the step S2 of collecting images during the water-conducting laser processing, extracting the shape contour of the spark area, and calculating the angle and processing depth of the laser spark to obtain a visualization result includes the following sub-steps:

[0059] S21, collecting images during the water-conducting laser processing, and gray-processing the images to obtain gray-scale images;

[0060] S22, performing Gaussian blurring on the grayscale image, and performing binarization processing on the image using threshold segmentation of the maximum inter-class variance method to obtain a binarized image;

[0061] S23, extracting features of the shape contour of the laser spark area in the binary image, and performing polygonal approximation processing to obtain a regular shape contour pattern;

[0062] S24, extracting the coordinates of each vertex of the regular shape contour pattern, and calculating the angle and processing depth of the laser spark.

[0063] On the basis of the above technical scheme, preferably, step S23 also includes screening all different contours generated, selecting the target contour, and retaining the shape contour of the laser spark area, wherein the pixel points are traversed in turn, if the pixel points in 8 consecutive areas are all white, then the pixel point is the internal point of the target shape and it is replaced by black; otherwise, the white color remains unchanged; all contours in the image are extracted, and all extracted contours are screened by area screening method, and the contour with the largest area is the shape contour of the spark area, and the shape contour of the spark area is polygonally approximated, and the number of vertices of the new contour is set to 3, so that the regular shape contour pattern is approximated to be a triangle.

[0064] On the basis of the above technical solution, preferably, in step S24, the coordinates of each vertex of the regular shape contour pattern are extracted, and the angle and processing depth of the laser spark are calculated, wherein the image after the target contour pattern is extracted is processed by BGR to HSV, the target triangular contour is extracted, and the contour line image is obtained; the contour line image is processed using Shi-Tomasi corner points to obtain the coordinates of the three points of the target triangular contour, and the angle of the laser spark is calculated using the slope angle formula for the coordinates of the three points, and the expression is:

[0065]

[0066] Where k1 and k2 are the slopes of the two sides of the required angle respectively; A is the angle of the laser spark;

[0067] The processing depth is calculated based on the angle of the laser spark, and the expression is:

[0068] H·[tanB+tan(AB)]=D

[0069] Where H is the processing depth, D is the water jet diameter, A is the angle of the laser spark, and B is the angle between the left contour edge and the incident direction of the water-guided laser.

[0070] In a second aspect, the present invention further provides a water-conducting laser processing progress visualization recognition system, which is implemented by the water-conducting laser processing progress visualization recognition method as described above, and the system comprises:

[0071] A simulation module is used to establish a simulation model, simulate the water-conducting laser processing process, and obtain simulation results under different processing times, wherein the simulation results include the slope of the processed hole wall and the processing depth;

[0072] Visualization module, used to collect images during water-conducting laser processing, extract the shape contour of the spark area, and calculate the angle of the laser spark and the processing depth to obtain visualization results;

[0073] The comparison and verification module is used to compare the visualization results with the simulation results of the same processing time to verify the correctness of the visualization recognition.

[0074] The water-guided laser processing visualization depth measurement method of the present invention has the following beneficial effects compared with the prior art:

[0075] (1) By collecting images of the water-conducting laser processing process in real time and performing image processing and data analysis, the relationship between the spark angle and the depth of the processing process over time is obtained, so that the processing progress can be automatically identified by observing the spark angle, thereby improving the processing efficiency and the quality of the finished product. The simulation model is used for comparison and verification, which greatly improves the accuracy of processing progress identification, ensures the consistency and reliability of the processing results, and reduces the processing errors caused by recognition errors;

[0076] (2) By setting the birth and death units, the units heated to a temperature above the melting point of the material can be removed in the thermal analysis module to achieve the purpose of material reduction, and the change in processing depth during the processing process can be observed by the change of material reduction over time;

[0077] (3) By using the ADPL command stream to create a function for the laser heat source of water-conducting laser processing, the required parameter values ​​and loading conditions can be changed quickly and easily;

[0078] (4) A grayscale image is obtained by grayscale processing the image, and the grayscale image is Gaussian blurred. The image is binarized using the threshold segmentation of the maximum inter-class variance method to obtain a binary image. The shape contour of the laser spark area in the binary image is then feature extracted and polygonal approximation is performed to obtain a regular shape contour pattern, which provides support for the subsequent vertex extraction of the triangular contour and improves the recognition accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0080] Figure 1 It is a flow chart of the method for visually identifying the progress of water-conducting laser processing of the present invention;

[0081] Figure 2 It is a simulation flow chart of the method for visually identifying the progress of water-conducting laser processing of the present invention;

[0082] Figure 3 It is a diagram of the total reflection phenomenon of laser in water beam in the method for visualizing and identifying the progress of water-guided laser processing of the present invention;

[0083] Figure 4 It is a simplified state diagram of the total reflection phenomenon of laser in water beam in the method for visually identifying the progress of water-guided laser processing of the present invention;

[0084] Figure 5 It is an ADPL function input interface diagram of the method for visually identifying the progress of water-conducting laser processing of the present invention;

[0085] Figure 6 A command flow cross-sectional diagram is generated for the ADPL of the method for visualizing and identifying the progress of water-conducting laser processing of the present invention;

[0086] Figure 7 A Gaussian heat source simulation diagram generated by the ADPL command stream of the method for visually identifying the progress of water-conducting laser processing of the present invention;

[0087] Figure 8 It is a diagram of simulation results of the water-conducting laser processing process at different times according to the method for visually identifying the progress of water-conducting laser processing of the present invention;

[0088] Fig. 9 The image is processed by grayscale in the method for visualizing and recognizing the progress of water-conducting laser processing according to the present invention;

[0089] Fig.10 The image is a binary processed image of the method for visualizing and recognizing the progress of water-conducting laser processing of the present invention;

[0090] Fig.11 A regular shape contour pattern of the water-conducting laser processing progress visualization identification method of the present invention;

[0091] Fig.12 It is a geometric model diagram after the regular shape contour pattern is extracted in the method for visualizing and identifying the progress of water-conducting laser processing of the present invention;

[0092] Fig.13 It is a specific processing depth simulation diagram of the water-conducting laser processing progress visualization identification method of the present invention. DETAILED DESCRIPTION

[0093] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0094] like Figure 3 As shown, the main technical principle of water-guided laser processing in this embodiment is the principle of optics, which is a composite processing technology that cuts the workpiece to be processed by guiding and concentrating the laser beam through a water jet; the focus of this principle is that the refractive index of water and air is different, which directly leads to the phenomenon that when the laser beam irradiates the interface between water and air at a certain angle, the incident angle is less than the critical angle of total reflection. In this way, the laser will not be transmitted during the propagation of the jet, but will continuously perform total reflection so that the energy of the laser is always confined in the water beam and propagates along the direction of the water beam, and will not diverge to cause the loss of laser energy and cause the processing material failure.

[0095] The process of water-guided laser can be preliminarily explained: the laser generated by the laser source is focused by a convex lens and enters the coupled water chamber. In this process, the distance between the focusing lens and the nozzle hole is adjusted so that the focus of the laser is at the center of the upper surface of the nozzle hole. In this way, the laser can enter the stable water jet and use total reflection to concentrate the laser energy to complete the processing work.

[0096] like Figure 4As shown in the figure, during the water-conducting laser processing, pulsed laser / quasi-continuous laser is emitted and passes through the water jet. After the material begins to absorb heat, the absorbed laser energy generates a plasma layer to separate the water and the material. After the laser pulse is completed and the plasma phenomenon ends, the water jet cools the material and completely takes away the heat through convection. It can be found that for the processing point, the instantaneous state of the water flow is very unstable due to the influence of the plasma phenomenon, and is basically dominated by irregular turbulence under impact, which also makes it difficult to analyze the fluid state for the processing point.

[0097] like Figure 1 As shown, a method for visually identifying the progress of water-conducting laser processing of the present invention comprises the following steps:

[0098] S1, establish a simulation model, simulate the water-conducting laser processing process, and obtain simulation results under different processing times. The simulation results include the angle formed by the hole wall and the processing depth.

[0099] like Figure 2 As shown, step S1 includes the following sub-steps:

[0100] S11, construct a three-dimensional geometric model of water-guided laser processing through modeling tools.

[0101] It should be noted that, in this embodiment, the three-dimensional geometric model of water-guided laser processing can be constructed through SOLODWORKS or CAD.

[0102] S12, meshing the three-dimensional geometric model using meshing.

[0103] It should be noted that in this embodiment, the three-dimensional geometric model is imported into the ANSYS simulation software and divided based on the built-in meshing tool. The hexahedral mesh is used for division, and each unit is a cube with a mesh size of 0.04 mm, which is convenient for subsequent finite element analysis.

[0104] S13, setting material parameters and laser processing parameters and selecting a Gaussian heat source model;

[0105] Among them, the Gaussian heat source model is closely related to the Gaussian curve. For the Gaussian curve, if the surface temperature of a circle is the same, then its heat source model is equivalent to the same temperature in a circle with the same diameter. The Gaussian heat source model is qualitatively understood as the temperature in the circle with the same diameter is high in the center and low at the edge. The quantitative understanding is that in the circle with the same diameter, the temperature is distributed according to the Gaussian curve; the formula of the Gaussian moving heat source is:

[0106]

[0107] Where Q is the Gaussian heat source density; P is the laser energy power, in W / m2 ; R is the radius of the laser light source; r is the distance between the laser processing point and the coordinate origin.

[0108] In this embodiment, since there is no need to apply a moving heat source, there is no need to consider the laser moving speed during calculation, and only the processing position needs to be considered.

[0109] S14, setting boundary conditions;

[0110] In this embodiment, three boundary conditions are considered at the same time: the boundary conditions include the workpiece temperature initial condition, the heat flux density boundary condition and the convection heat transfer initial condition, wherein:

[0111] The expression of the initial condition of the workpiece temperature is:

[0112] T=T *

[0113] Where T is the current temperature and T* is the initial temperature of the workpiece at a specific time point;

[0114] The expression of the heat flux boundary condition is:

[0115] {q} T {η}=-q *

[0116] Where, {q} T is the heat flux density vector, {η} is the unit vector of the outer normal of the boundary surface, and -q* is the specific heat flux density value on the boundary, which is a constant;

[0117] The expression of the initial condition of convective heat transfer is:

[0118] {q} T {η}=h f (T s -T B )

[0119] In the formula, h f is the convective heat transfer coefficient; T s is the wall temperature; T B is the fluid temperature.

[0120] S15, calculating and obtaining the laser heat flux density according to the material parameters and the laser processing parameters, and setting a corresponding convection heat transfer coefficient based on the cooling of the workpiece by the water jet fluid.

[0121] Among them, when simulating the laser processing process, the laser heat source generally uses an ordinary body heat source. However, for water-guided laser processing, according to the principle of water jet guiding the laser, the laser undergoes multiple total reflections in the water beam, so the energy can be nearly uniformly distributed in the water beam cross section; and because the depth generated by the laser is very shallow relative to the etching depth when it acts, the laser energy used is regarded as a surface load for simulation in the present invention.

[0122] At the same time, it is also assumed to simplify the calculation of heat source energy, that is, the thermal parameters of the material will not change with the change of temperature, and the calculation formula of laser heat flux density can be obtained:

[0123] I(t)=I0·α

[0124] I0(t)=0.865P0(t) / (πd 2 / 4)

[0125] Where α is the material absorption coefficient; P0 is the output power of the laser; d is the spot diameter, and I is the heat flux density.

[0126] In addition, unlike traditional laser processing, the surface of water-conducting laser processed parts has a strong cooling effect brought by fluid flushing. The heat exchange that occurs when fluid flows over the solid surface is called convection heat transfer. The basic formula for convection heat transfer is the Newton cooling formula, that is:

[0127] Q h =h f (T-T0)

[0128] In the formula, Q h is the heat flux density of convective heat transfer; T is the current temperature; T0 is the temperature at the previous moment; h f is the convective heat transfer coefficient;

[0129] According to Gauss's moving heat source formula, since the side wall of the workpiece is very thin and far away from the heating point, the edge can be assumed to be in an adiabatic state; in addition to the side conditions, the air convection and energy loss expressions on the upper and bottom surfaces of the workpiece are: q out =Q h .

[0130] Because convective heat transfer is the result of the combined effect of two basic heat transfer modes: heat conduction and convection. Therefore, the key to convective heat transfer calculation is to obtain the heat transfer coefficient between the fluid and the solid surface. That is, the calculation formula of the convective heat transfer coefficient is:

[0131] h f = Nuλ / R

[0132] Where Nu is the Nusselt number; λ is the thermal conductivity of the fluid, λ = 0.6; R is the characteristic length;

[0133] Based on the simulation of water-guided laser processing of flat plates, the empirical formula for the Nusselt number is:

[0134] Nu=0.745Re 1 / 2 Pr 1 / 3

[0135] Where Pr is the Prandtl number and Re is the Reynolds number.

[0136] Among them, the Reynolds number, as an important parameter in fluid mechanics, is a dimensionless number that can be used to represent the flow of fluid. It is the ratio of the inertial force on a moving object to the viscosity or friction force. The Reynolds number is usually used to distinguish whether the flow of a fluid is laminar or turbulent. At the same time, the Reynolds number can also be used to determine the magnitude of the resistance encountered by an object in the flow. The smaller the Reynolds number, the more significant the influence of viscosity, and the larger the Reynolds number, the more significant the influence of inertia. The value of the Reynolds number requires observation of the fluid state, so as to select a suitable Reynolds number for calculation.

[0137] The expression of Prandtl number Pr is:

[0138]

[0139] In the formula, c pw is the specific heat of water; μ w is the hydrodynamic viscosity; k w is the thermal conductivity of water;

[0140] The expression of Reynolds number Re is:

[0141]

[0142] Where ρ is the density of the fluid; v is the velocity of the fluid flowing through the cross section; l is the characteristic length of the fluid flowing through the cross section; γ is the kinematic viscosity coefficient of the fluid.

[0143] Since the points of action of laser heat are basically concentrated in the stagnation zone, the convective heat transfer of water outside the zone is less, so a smaller convective heat transfer coefficient can be taken for the units outside the stagnation zone.

[0144] At the same time, during the water-conducting laser processing, a plasma generation process will occur due to the concentration of laser energy. This process is similar to an explosion, that is, it will produce a significant shock wave effect. For the same shock wave effect, the liquid impact in the water-conducting laser will be much greater than the impact caused by the gas pressure, and the duration is relatively shorter. It is precisely because of this feature that the melt can be quickly and effectively discharged during water-conducting laser processing, thereby solving the problem of the recast layer caused by the accumulation of melt under ordinary laser processing. At the same time, the rapid removal of the melt by the high-speed water flow also prevents its influence on the processing process. There is no need to consider the factors of processing the liquid molten layer, which simplifies the simulation of the processing process.

[0145] In addition, another major advantage of water-guided laser is that the high-speed water flow can bring about a rapid cooling effect for the workpiece, which will reduce the diffusion range of the temperature field during processing. This will directly greatly reduce the heat-affected zone of the workpiece caused by the high-temperature processing of the laser, thereby reducing the residual thermal stress after processing. This process is similar to the quenching effect of hot processing. Because of the above advantages of water-guided laser processing, there is no need to analyze thermal stress during the simulation process.

[0146] Similar to the advantages brought by the high-speed water flow mentioned above, it can also quickly discharge the plasma layer generated during the processing, which greatly reduces the impact of this energy on the transmission of laser energy, so there is no need to consider the thermal impact caused by this phenomenon during the analysis process.

[0147] S16, solving and calculating the temperature field to obtain the temperature field distribution result.

[0148] The temperature field in the water-conducting laser processing process is essentially a transient temperature field of a continuous medium, and it needs to be solved numerically, that is, by establishing finite element equations and initial and boundary conditions, and using numerical difference methods to transform it into a solution for a steady-state temperature field.

[0149] It includes the following sub-steps:

[0150] The finite element equation of the transient temperature field is established as follows:

[0151] [C]{T}+[K]{T}={Q}

[0152] Where, [C] is the non-steady-state temperature matrix; {T} is the temperature vector; [K] is the temperature field stiffness matrix; {Q} is the thermal load vector;

[0153] The difference calculation of numerical calculation is transformed into the solution of steady-state temperature field, and the expression is:

[0154]

[0155] Convert to the expression:

[0156]

[0157] Where θ is the transient comprehensive coefficient; Δt is the time difference from the previous moment to the current moment, Δt = t n+1 -t n ; {T n} is the node at t n The temperature value matrix at the moment; For the node at t n The matrix of derivative values ​​of the temperature at time instant.

[0158] The implicit Euler backward difference method is used in the calculation process to ensure the accuracy and stability of the results.

[0159] Considering that thermal transient analysis is used to complete this simulation, the number of loading steps for transient analysis is very large, and certain overload and out-of-calculation errors will occur during calculation and analysis. Therefore, the ADPL command stream is used to create a function for the laser heat source of water-conducting laser processing, which can quickly and easily change the required parameter values ​​and loading conditions.

[0160] like Figure 5 As shown in the figure, in the Mechanical ADPL interface, use Functions in the Parameters column to create a function. During the input process, define the parameter variables according to the Gaussian moving heat source formula and the convection heat transfer formula to achieve the function of being able to change the parameters at any time to obtain a general conclusion under different data.

[0161] like Figure 6 As shown, the original coordinate system is selected as the reference coordinate system of the process, and the processing starts at t=0; and this function is called for self-running. The laser heat flux calculation formula can complete the calculation of the laser power density parameters, and can also determine the Reynolds number and spot radius of convective heat transfer; after entering the parameters, the program will get the results according to the function by saving and reading the function file, and get the ADPL command flow to complete the mobile heat source.

[0162] like Figure 7 As shown in the figure, through this command flow, a Command module can be created in the Workbench. By entering the command flow here, the required Gaussian heat source can be applied to the model.

[0163] like Figure 8 As shown in Figure 2, according to the simulation results, it can be obtained that as the processing time changes, the processing depth gradually becomes deeper and the slope of the processed hole wall gradually decreases.

[0164] S17, judging whether the temperature of each unit in the temperature field distribution result is greater than the melting point of the material, removing the units that have reached the melting point of the material; screening and judging the units that have not reached the melting point of the material;

[0165] S18, judging whether the processing is finished according to the simulation results and the preset processing standard, if the processing end condition is not met, returning to step S17 to re-judge until the processing end condition is met.

[0166] For the simulation model, the temperature field of water-conducting laser processing needs to be simulated by the finite element software Ansys. Unlike ordinary laser processing, water-conducting laser processing has a strong cooling effect of water jet, and in the simulation process of this method, the frequency of the laser is set relatively small for the sake of process clarity; at the same time, in order to remove the thermal effects reflected by the units that have reached the melting temperature, the technology of birth and death units is used; the purpose of the birth and death unit technology is not to delete the specified unit and make it disappear permanently, but to multiply its characteristic matrix by a very small factor to return its unit load and specific heat and other thermal parameters to zero, and its mass and energy will not be involved in the model solution results. However, there will also be a problem that the position of the unit to be killed cannot be located when using the birth and death unit technology because the processing process is a nonlinear transient process, so it is necessary to determine the birth and death state of the unit through the temperature calculation value, that is, to screen the required units according to the temperature value of the unit. In the ADPL command stream, the ETABLE command and the ESEL command can be used to determine the relevant data of the selected unit, select the unit that meets the killing condition, and then use the command EKILL to kill the selected unit. Since there is phase change in the laser processing process, it is necessary to repeatedly apply boundary conditions through loop statements to respond to the time-varying characteristics of boundary conditions during the heat transfer process.

[0167] By setting the birth and death units, the units heated to above the melting point of the material can be removed in the thermal analysis module to achieve the purpose of material reduction. The change in processing depth during the processing can be observed by the change of material reduction over time.

[0168] Among them, the Coanda effect, also known as the wall attachment effect, is a specific phenomenon in which the fluid, when flowing through an object, does not follow the original ideal route to leave the object, but instead tends to flow along the protruding surface of the object, and causes the surrounding fluid to escape into this jet. As the direction of fluid movement changes, a region with lower pressure is generated around it, which is called the Coanda effect. The reason for this phenomenon can be based on Prandtl's boundary layer concept. Since there is a viscous boundary layer on the surface of the object, and the viscous fluid will generate vorticity on the wall that can diffuse and convect into the flow field, it is precisely because of this vorticity that in viscous flow, the fluid and the surface of the object through which it flows can produce viscous friction while slowing down the flow rate of the fluid, and the fluid is adsorbed on the wall and flows, thus causing the Coanda effect.

[0169] During the laser processing, the sparks formed are the result of the impact of water flow and the splashing of molten material. Due to this effect, the angle of the sparks formed during water-conducting laser processing will change to a certain extent.

[0170] S2, collects images during the water-conducting laser processing, extracts the shape contour of the spark area, and calculates the angle of the laser spark and the processing depth to obtain the visualization results.

[0171] Step S2 includes the following sub-steps:

[0172] like Fig. 9 As shown, S21, collecting images during the water-conducting laser processing, and performing grayscale processing on the images to obtain grayscale images.

[0173] It should be noted that in this embodiment, the color image is first converted into a grayscale image by using image grayscale conversion. For a color image, the pixel value of each pixel is determined by the three components of RGB, and the value of each component can vary between 0-255, that is, there are 256 choices, which will result in 16 million (256*256*256) possibilities for the pixel value of a pixel determined by the three components; but for a grayscale image, the three component values ​​of RGB of each pixel are the same, so for a grayscale image, each pixel has only 256 possibilities, and such an order of magnitude difference can greatly reduce the amount of calculation in image processing.

[0174] like Fig.10 As shown, S22, Gaussian blur is performed on the grayscale image, and the image is binarized using threshold segmentation of the maximum inter-class variance method to obtain a binary image.

[0175] It should be noted that, in this embodiment, after grayscale processing, image binarization is used to process the image in order to convert the grayscale image into a black and white image with an obvious black and white effect. This operation will greatly reduce the amount of data in the image, thereby being able to meet the needs of contour extraction, and this processing method will not affect the overall and local features of the image, that is, it can still reflect the characteristics of the image and analyze it.

[0176] Binarization is to select a certain threshold to divide the pixels of 256 brightness levels into two parts: pixels with grayscale greater than or equal to the threshold and pixels with grayscale less than the threshold. The former has its grayscale value set to 255, while the latter has its grayscale value set to 0 to represent the background or exceptional object area. The following advantages of binarization can be obtained: remove noise in the image to make the image clearer; enhance contrast to facilitate distinction; reduce the amount of data and speed up image processing.

[0177] A pre-processing method for grayscale images using Gaussian blur, which is a low-pass filter that uses convolution calculations between an image and a normal distribution to reduce image noise and reduce the level of detail.

[0178]

[0179] Where r is the blur radius and σ is the standard deviation of the normal distribution.

[0180] Using this formula, the convolution matrix composed of pixels with non-zero distribution is transformed with the original image, so that each pixel value is equal to the weighted average of the surrounding adjacent pixel values, and the original pixel value occupies the largest weighted weight. The weight of adjacent pixels decreases with the distance, thereby better preserving the edge effect and reducing oscillation.

[0181] After Gaussian blurring, the image is processed using the maximum inter-class variance method. In the binarization process, the difference between the pixels on both sides of the threshold needs to be as large as possible. The maximum inter-class variance method defines two variances for this purpose, the intra-class variance and the inter-class variance. The former represents the weighted sum of the variances of the regions on both sides of the threshold, while the latter represents the weighted variance of the mean of the data on both sides of the threshold from the total mean.

[0182] Within-class variance:

[0183] s w 2 =w0s0 2 +w1s1 2

[0184] Between-class variance:

[0185] s b 2 =w0(m0-mc ) 2 +w1(m1-m c ) 2

[0186] In the formula, w0 and w1 are the weights of the two regions respectively; s0 2 , s1 2 are the variances of the two regions respectively; m c is the mean, and m c =w0m0+w1m1.

[0187] Among them, the closer the internal data of the two regions divided by the threshold are, the better the binarization effect is, that is, the smaller the intra-class variance is, the better; and the inter-class variance is opposite to the intra-class variance. The larger the inter-class variance is, the greater the difference between the foreground and the background is, and the better the binarization effect is.

[0188] The maximum between-class variance method expression is:

[0189]

[0190] Where i is the threshold segmentation point; p n is the weight of the nth pixel in the entire image.

[0191] Among them, the sum of the intra-class variance and the inter-class variance is a constant, that is, maximizing the inter-class variance is equivalent to minimizing the intra-class variance. Threshold segmentation using the inter-class variance maximum method will minimize the probability of misclassification, thereby improving the accuracy of the algorithm.

[0192] The selection of thresholds needs to meet the following conditions: the required part of the image should be displayed as clearly as possible; the unnecessary part should be as black as possible; if there is an extra part of the image, the display area should be as small as possible. Based on the above conditions, after continuously changing the selection and testing of the threshold, the threshold was finally locked in the range of 190-210, so that the pixels that need to be retained can be extracted from many colors.

[0193] After the binarization is successfully performed, the shape of the fire part is displayed. In order to calculate its angle, its outline needs to be extracted. However, for the binarized image, all irrelevant elements cannot be completely filtered out, so all the different outlines generated need to be screened and the target outline needs to be selected.

[0194] like Fig.11 As shown, S23, feature extraction is performed on the shape contour of the laser spark area in the binary image, and polygonal approximation processing is performed to obtain a regular shape contour pattern.

[0195] Step S23 also includes screening all the different contours generated, selecting the target contour, and retaining the shape contour of the laser spark area, wherein the pixel points are traversed in turn, if the pixel points in 8 consecutive areas are all white, then the pixel point is the internal point of the target shape and is replaced by black; otherwise, the white color remains unchanged; all contours in the image are extracted, and all extracted contours are screened using the area screening method. The contour with the largest area is the shape contour of the spark area, and the shape contour of the spark area is polygonally approximated, and the number of vertices of the new contour is set to 3, so that the regular shape contour pattern is approximated to be a triangle.

[0196] S24, extracting the coordinates of each vertex of the regular shape contour pattern, and calculating the angle and processing depth of the laser spark.

[0197] Among them, in step S24, the image after the target contour pattern is extracted is converted from BGR to HSV, the target triangle contour is extracted, and the contour line image is obtained; the contour line image is processed using the Shi-Tomasi corner point to obtain the coordinates of the three points of the target triangle contour, and the angle of the laser spark is calculated using the slope angle formula. The expression is:

[0198]

[0199] Where k1 and k2 are the slopes of the two sides of the required angle respectively; A is the angle of the laser spark;

[0200] The processing depth is calculated based on the angle of the laser spark, and the expression is:

[0201] H·[tanB+tan(AB)]=D

[0202] Where H is the processing depth, D is the water jet diameter, A is the angle of the laser spark, and B is the angle between the left contour edge and the incident direction of the water-guided laser.

[0203] Among them, Shi-Tomasi corner point is an improvement of Harris corner point extraction. Its principle is similar to Harris, that is, a local small window is defined and moved in various directions to observe the changes in image intensity within the small window. Only when its intensity changes significantly in all directions, the desired corner point exists in this window.

[0204] The mathematical model is based on the Taylor expansion formula:

[0205]

[0206] Among them, I(x,y) is the gray value of the image at point (x,y), I x , I yis the partial derivative function of I(x,y); w(x,y) is the window function; Δx and Δy are the displacement distances in the x and y directions respectively.

[0207] According to the above formula, we can get its matrix form:

[0208]

[0209] Using orthogonal matrix multiplication, we can get the following formula:

[0210]

[0211] Among them, λ1, λ2 are the eigenvectors of the intermediate matrix. Let this matrix be M, then its corner point response value R = detM-K(trace(M)) 2 , K is the experience value.

[0212] However, in the Shi-Tomasi corner point, since a strong corner point is obtained if the smaller of the two eigenvalues ​​is greater than the minimum threshold, and the larger uncertainty depends on the smaller eigenvalue, a better corner point can be obtained by finding the maximum value of the smallest eigenvalue, that is, a better and more complete result can be obtained.

[0213] Before that, the image after extracting the target contour pattern needs to be processed, because only three points need to be found, and the remaining points are superfluous interference items. Therefore, it is hoped that the image will only retain the triangular contour as much as possible, and the rest is all black. It is necessary to use the RGB to HSV method to extract the color, that is, first convert the image from BGR to HSV, and then determine the HSV high and low thresholds of the color to accurately extract the target triangular contour.

[0214] The HSV color conversion table is shown in Table 1

[0215] Table 1

[0216]

[0217] In this embodiment, the red color is used for the outline stroke, so the color is adjusted and selected from (0,43,46) to (10,255,255), and the color range is finally determined to obtain a processed image with only the outline. Then, the Shi-Tomasi corner point calculation can be used to extract the vertices of the triangle outline.

[0218] After obtaining the above results, we need to complete the calculation of the angle. Since the coordinates of three points are required, we need to store the three points in the form of a two-dimensional table, and then calculate the slope and angle in turn. Since the order from left to right can make the second point the vertex, there is no need to determine the vertex of the angle.

[0219] like Figure 12-13As shown in,S3, the visualization results and simulation results with the same processing time are compared to verify the correctness of the visualization recognition.

[0220] In this embodiment, the angle of the laser spark is calculated to be 53° by extracting the vertices of the triangular contour, the diameter of the water jet is 200 μm, and the angle between the left contour edge and the incident direction of the water-guided laser is 11°, and the processing depth H can be calculated to be approximately equal to 182.684 μm;

[0221] In the simulation process, since it is easier to select the processing depth than the angle, the simulation process diagram with a processing depth of about 150-180μm is selected. The angle formed by the hole wall in the simulation diagram is measured and compared with the angle calculated by visualization. After measurement, it can be found that the angle formed by the hole wall in the simulation diagram is approximately equal to the angle calculated by visualization, which can verify the basic correctness of the initial assumption, that is, during the processing process, the processing depth can be judged by the angle of the spark generated by the processing at different times; therefore, by combining different image processing methods, the angle of the spark generated by the laser can be accurately identified.

[0222] In a second aspect, the present invention further provides a water-conducting laser processing progress visualization recognition system, which is implemented by the water-conducting laser processing progress visualization recognition method as described above, and the system comprises:

[0223] A simulation module is used to establish a simulation model, simulate the water-conducting laser processing process, and obtain simulation results under different processing times, wherein the simulation results include the slope of the processed hole wall and the processing depth;

[0224] Visualization module, used to collect images during water-conducting laser processing, extract the shape contour of the spark area, and calculate the angle of the laser spark and the processing depth to obtain visualization results;

[0225] The comparison and verification module is used to compare the visualization results with the simulation results of the same processing time to verify the correctness of the visualization recognition.

[0226] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for visually identifying the progress of water-guided laser processing, characterized in that: The method comprises the following steps: S1, establishing a simulation model, simulating the water-conducting laser processing process, and obtaining simulation results under different processing times, wherein the simulation results include the angle formed by the hole wall and the processing depth; S2, collects images during the water-conducting laser processing, extracts the shape contour of the spark area, and calculates the angle of the laser spark and the processing depth to obtain a visual result; S3, compare the visualization results with the simulation results of the same processing time to verify the correctness of the visualization recognition.

2. The method for visually identifying the progress of water-conducting laser processing according to claim 1, characterized in that: Step S1 includes the following sub-steps: S11, constructing a 3D geometric model of water-guided laser processing through modeling tools; S12, meshing the three-dimensional geometric model using a meshing tool body; S13, setting material parameters and laser processing parameters and selecting a Gaussian heat source model; S14, setting boundary conditions; S15, calculating and obtaining the laser heat flux density according to the material parameters and the laser processing parameters, and setting a corresponding convection heat transfer coefficient based on the cooling of the workpiece by the water jet fluid; S16, solving and calculating the temperature field to obtain the temperature field distribution result; S17, judging whether the temperature of each unit in the temperature field distribution result is greater than the melting point of the material, removing the units that have reached the melting point of the material; screening and judging the units that have not reached the melting point of the material; S18, judging whether the processing is finished according to the simulation results and the preset processing standard, if the processing end condition is not met, returning to step S17 to re-judge until the processing end condition is met.

3. The method for visually identifying the progress of water-guided laser processing according to claim 2, characterized in that: The boundary conditions include workpiece temperature initial conditions, heat flux density boundary conditions and convection heat transfer initial conditions, where: The expression of the initial condition of the workpiece temperature is: T=T * Where T* is the initial temperature of the workpiece at a specific time point; The expression of the heat flux boundary condition is: {q} T {η}=-q * Where, {q} T is the heat flux density vector, {η} is the unit vector of the outer normal of the boundary surface, and -q* is the specific heat flux density value on the boundary, which is a constant; The expression of the initial condition of convective heat transfer is: {q} T {η}=h f (T s -T B ) In the formula, h f is the convective heat transfer coefficient; T s is the wall temperature; T B is the fluid temperature.

4. The method for visually identifying the progress of water-conducting laser processing according to claim 3, characterized in that: In step S15, the laser heat flux density is calculated based on the material parameters and the laser processing parameters, and the expression is: I(t)=I0·α I0(t)=0.865P0(t) / (πd 2 / 4) Where α is the material absorption coefficient; P0 is the output power of the laser; d is the spot diameter, I is the laser heat flux density, and t is the current processing time.

5. The method for visually identifying the progress of water-conducting laser processing according to claim 4, characterized in that: In step S15, a corresponding convection heat transfer coefficient is set based on the cooling of the workpiece by the water jet fluid, which includes the following sub-steps: The water jet fluid flows over the workpiece surface and undergoes convective heat exchange with it. The expression is: Q h =h f ·(T-T0) In the formula, Q h is the heat flux density of convective heat transfer; T is the current temperature; T0 is the temperature at the previous moment; h f is the convective heat transfer coefficient; The expression of convective heat transfer coefficient is: h f =Nuλ / R Where Nu is the Nusselt number; λ is the thermal conductivity of the fluid, λ = 0.6; R is the characteristic length; Based on the simulation of water-guided laser processing of flat plates, the empirical formula for Nusselt number is: Now=0.745Re 1 / 2 Per 1 / 3 Where Pr is the Prandtl number, Re is the Reynolds number, The expression of Prandtl number Pr is: In the formula, c pw is the specific heat of water; μ w is the hydrodynamic viscosity; k w is the thermal conductivity of water; The Reynolds number Re expression is: Where ρ is the density of the fluid; y is the velocity of the fluid flowing through the cross section; l is the characteristic length of the fluid flowing through the cross section; γ is the kinematic viscosity coefficient of the fluid.

6. The method for visually identifying the progress of water-conducting laser processing according to claim 5, characterized in that: In step S16, the temperature field is solved and calculated to obtain the temperature field distribution result, which includes the following sub-steps: The finite element equation of the transient temperature field is established as follows: [C]{T}+[K]{T}={Q} Where, [C] is the non-steady-state temperature matrix; {T} is the temperature vector; [K] is the temperature field stiffness matrix; {Q} is the thermal load vector; The difference calculation of numerical calculation is transformed into the solution of steady-state temperature field, and the expression is: Convert to the expression: Where θ is the transient comprehensive coefficient; Δt is the time difference from the previous moment to the current moment, Δt = t n+1 -t n ; {T n } is the node at t n The temperature value matrix at the moment; For the node at t n The matrix of derivative values ​​of the temperature at time instant.

7. The method for visually identifying the progress of water-conducting laser processing according to claim 2, characterized in that: The step S2 of collecting images during the water-conducting laser processing, extracting the shape contour of the spark area, and calculating the angle and processing depth of the laser spark to obtain a visualization result includes the following sub-steps: S21, collecting images during the water-conducting laser processing, and gray-processing the images to obtain gray-scale images; S22, performing Gaussian blurring on the grayscale image, and performing binarization processing on the image using threshold segmentation of the maximum inter-class variance method to obtain a binarized image; S23, extracting features of the shape contour of the laser spark area in the binary image, and performing polygonal approximation processing to obtain a regular shape contour pattern; S24, extracting the coordinates of each vertex of the regular shape contour pattern, and calculating the angle and processing depth of the laser spark.

8. The method for visually identifying the progress of water-conducting laser processing according to claim 7, characterized in that: Step S23 also includes screening all the different contours generated, selecting the target contour, and retaining the shape contour of the laser spark area, wherein the pixel points are traversed in turn, if the pixel points in 8 consecutive areas are all white, then the pixel point is the internal point of the target shape and is replaced by black; otherwise, the white color remains unchanged; all contours in the image are extracted, and all extracted contours are screened using the area screening method. The contour with the largest area is the shape contour of the spark area, and the shape contour of the spark area is polygonally approximated, and the number of vertices of the new contour is set to 3, so that the regular shape contour pattern is approximated to be a triangle.

9. The method for visually identifying the progress of water-conducting laser processing according to claim 8, characterized in that: In step S24, the coordinates of each vertex of the regular shape contour pattern are extracted, and the angle and processing depth of the laser spark are calculated. The image after the target contour pattern is extracted is converted from BGR to HSV, the target triangular contour is extracted, and the contour line image is obtained; the contour line image is processed using Shi-Tomasi corner points to obtain the coordinates of the three points of the target triangular contour, and the angle of the laser spark is calculated using the slope angle formula for the coordinates of the three points. The expression is: Where k1 and k2 are the slopes of the two sides of the required angle respectively; A is the angle of the laser spark; The processing depth is calculated based on the angle of the laser spark, and the expression is: H·[tan B+tan(AB)]=D Where H is the processing depth, D is the water jet diameter, A is the angle of the laser spark, and B is the angle between the left contour edge and the incident direction of the water-guided laser.

10. A water-conducting laser processing progress visualization recognition system, implemented by the water-conducting laser processing progress visualization recognition method according to any one of claims 1 to 9, characterized in that: The system comprises: A simulation module is used to establish a simulation model, simulate the water-conducting laser processing process, and obtain simulation results under different processing times, wherein the simulation results include the slope of the processed hole wall and the processing depth; Visualization module, used to collect images during water-conducting laser processing, extract the shape contour of the spark area, and calculate the angle of the laser spark and the processing depth to obtain visualization results; The comparison and verification module is used to compare the visualization results with the simulation results of the same processing time to verify the correctness of the visualization recognition.

Images