An online identification method for non-uniform wear of end mills

Abstract

The present invention proposes an online identification method for non-uniform wear of end mills. Specifically, it includes: establishing a milling force mechanism model considering tool wear, and designing a time domain simulation of milling force under non-uniform wear conditions based on this model. The milling force F is obtained from the simulation experiment. x 、F y and F tot Feature extraction was performed, and a total of 48 eigenvalues were obtained. The feature sensitivity was analyzed using the correlation analysis method to obtain features that can reflect the state of non-uniform wear. The results were verified through physical experiments, and the experimental results showed that the feature is accurate in identifying non-uniform wear. Beneficial effect: The present invention provides a more accurate identification method for uneven wear of milling cutters. This method can effectively solve the problem of insufficient data samples caused by the difficulty in obtaining monitoring data and non-uniform wear labels, and is convenient and effective.

Description

Technical Field

[0001] The invention relates to an online identification method for an end mill, in particular to an online identification method for uneven wear of an end mill, and belongs to the technical field of numerical control machining. Background Art

[0002] Online monitoring of tool wear status is an important part of intelligent milling processing. Milling processing is widely used in high-end manufacturing industries such as aerospace, medical molds, etc. due to its good processing efficiency, processing accuracy and applicability to complex-shaped parts. Tool manufacturing errors or changes in the axial cutting depth during milling will cause non-uniform wear of the cutting edge of high-value integral end mills along the axial direction. The wear state of the tool itself is a key factor in determining processing quality and processing efficiency. Tools in a non-uniform wear state will reduce the quality of the processed surface, reduce processing efficiency, and even lead to serious losses such as scrapping of the workpiece. Online monitoring of tool wear status has become one of the focuses of attention in industry and academia. The wear value of the tool can be obtained through industrial cameras, but there is still a lack of effective online identification methods for the non-uniform wear state of the tool.

[0003] Milling force is one of the most important process parameters in the milling process. Milling force information can reflect milling status information such as tool wear, tool runout, and milling parameters. Online monitoring of tool wear using real-time milling force signals is currently a major research topic. Secondly, scholars have conducted related research using sensors such as ultrasonic, acoustic emission, and lasers.

[0004] In addition, the wear status of the side edge of the end mill is often observed through a camera. When the end mill has a large cutting depth, since the side edge is a three-dimensional curved surface, all wear information of the end mill cannot be captured within the field of view of a camera, and the overall wear status of the milling cutter cannot be obtained. On the other hand, after the end mill has been used for a period of time, different wear values will be generated at different positions along the cutting edge due to changes in working conditions during use or manufacturing errors. This results in non-uniform wear on the cutting edge, which in turn results in different monitoring signal characteristics under the same wear label VB conditions, which brings troubles to the real-time monitoring of the tool wear status.

[0005] Existing Chinese patent CN 112705766 A describes a method for monitoring the uneven wear of cutting tools. This solution uses mathematical calculations, utilizing the force exerted on the workpiece as basic data. By comparing the left and right side data of the peak of the waveform generated by the original force, the difference between the two peaks is determined. A large difference indicates that the cutting tool is in an uneven wear state. When encountering complex waveforms, the combination of mathematical calculations and waveform comparisons is used to make judgments, and the results obtained are not accurate. In addition, this solution requires calculating the milling force of a new cutting tool for each milling cutter of different specifications, which increases monitoring costs. Summary of the Invention

[0006] Purpose of the invention: In response to the deficiencies in the prior art, the present invention provides an online detection method for non-uniform wear of end mills. This method can more effectively identify the wear state of the milling cutter and quickly determine whether the wear state is in a non-uniform wear state. When the milling cutter has a large cutting depth, this method can more quickly identify the wear state of the milling cutter and determine whether the milling cutter is in a non-uniform wear state. For workpieces with relatively high machining precision, the cutter can be replaced in time, avoiding the machining deviation caused by tool problems affecting product quality.

[0007] Technical solution: An online detection method for uneven wear of end mills, comprising the following steps:

[0008] S1. The cutting edge of the end mill is a spatial spiral curve. The tool is mathematically described to obtain the instantaneous cutting thickness;

[0009] S2. Use the micro-element cutting force instantaneous mechanical force model to establish a milling force model considering wear. The resultant milling force is expressed as the matrix Mat dFx The algebraic sum of all elements in ;

[0010] S3. Design a milling force simulation test for tool non-uniform wear and analyze the milling force signal characteristics that can reflect tool non-uniform wear;

[0011] S4, extracting time domain and frequency domain features of the milling force signal obtained from the simulation experiment of S3, and using eigenvalues to quantitatively describe the signal information;

[0012] S5. Analyze the correlation between the milling force signal characteristic value and tool wear, and solve the MAPE to obtain the milling force signal characteristic with the highest correlation;

[0013] S6. Collect the milling force signal during the machining process, and identify the non-uniform wear state of the tool during the machining process based on the milling force signal feature with the highest correlation obtained in S5.

[0014] The S1 is specifically:

[0015] S1.1. Describe the tool mathematically and establish the tool Cartesian coordinate system X T Y T Z T O T ;

[0016] Among them, Z T The axis coincides with the tool axis and its positive direction points to the tool handle. T -Y T The plane coincides with the plane formed by the tool tip, Y TThe axis coincides with the tool tip, and the blades are numbered i=1, 2, 3, ..., N in sequence. The numbering order is counterclockwise rotation around the positive direction of the Z axis, and the blade with number i=1 coincides with the Y T The axes intersect, and N is the number of tool edges;

[0017] The blade is discretized along its axis, and the discrete unit is marked as P Eij ;

[0018] Among them, i represents the blade, j represents the blade along Z T Discrete unit number in the positive direction of the axis, P E11 Indicates the j=1 discrete point corresponding to the i=1 blade, the maximum number of discrete units M;

[0019] S1.2. Establish the workpiece Cartesian coordinate system X W Y W Z W O W ;

[0020] Among them, X W The positive direction of the axis points to the tool feed direction, Z W The positive direction of the axis points to the direction of the tool. During the machining process, it is assumed that the workpiece does not move and only the tool moves, that is, the workpiece coordinate system is a fixed coordinate system. Let Y T Axis and Y W When the axis direction is consistent, it is time t0. Then at any time t, the discrete point P on the blade Eij The corresponding angle Φ i,j for:

[0021]

[0022] Among them, Φ i,j Point P Eij By Y W Axis starts to rotate around Z W The direction angle of the axis in the clockwise direction is in the range of [0°, 360°], ω is the tool rotation angular velocity (rad / s), β is the tool helix angle, D is the tool diameter, and dz is the discrete spacing of the tool edge along its axis.

[0023] Click P Eij The corresponding instantaneous cutting thickness is expressed as:

[0024]

[0025] Among them, f t is the feed rate per tooth, in mm / tooth, Φ out is the cut-out angle, its value is π, the cut-in angle Φ in It can be expressed as:

[0026] Φ in=π-acos((0.5Da e ) / 0.5D)

[0027] where a e is the cutting width;

[0028] Based on the above method, the instantaneous cutting thickness corresponding to all blades at each moment can be obtained and stored in the matrix Mat hw In the formula, it is expressed as:

[0029]

[0030] The S2 is specifically:

[0031] Using the instantaneous mechanical force model of micro-element cutting force, the milling force corresponding to the micro-element of the cutting edge at a certain time t is expressed as:

[0032] dF t =K tc h wij dz=k tc f t sin(Φ i,j )dz

[0033] dF r =K rc h wij dz=K rc f t sin(Φ i,j )dz

[0034] Among them, dF t is the blade microelement tangential force, and its direction coincides with the direction of instantaneous cutting speed; dF r is the radial force of the blade, its direction is perpendicular to dF t And points to the workpiece, K tc and K rc are the tangential force coefficient and radial force coefficient respectively;

[0035] Furthermore, the tangential force and radial force are transformed into the force F along the feed direction in the workpiece coordinate system through coordinate transformation. x and the force F perpendicular to the feed direction y , the transformation formula is:

[0036] dF x =dF r sin(Φ i,j )+dF t cos(Φ i,j )

[0037] dF y =dF r cos(Φ i,j)-dF t sin(Φ i,j )

[0038] Considering the helix angle of the milling cutter and multi-edge cutting, the time domain signal expression of the milling force is established:

[0039]

[0040]

[0041] in,

[0042] The milling force modeling considering wear, that is, the milling force coefficient is used to represent wear, and the milling force coefficient is expressed in matrix form:

[0043]

[0044] The infinitesimal force matrix can be obtained:

[0045] Mat dFx =Mat Krc Mat hw ·dz·sin(Φ i,j )+mat Ktc Mat hw ·dz·cos(Φ i,j )

[0046] The resultant milling force can be expressed as the matrix Mat dFx The algebraic sum of all elements in .

[0047] The S3 is specifically:

[0048] Set discrete points along the tool axis with the tool tip as the starting point. Divide the discrete points into at least three parts, marked as: CE-1, CE-2, CE-3, and design at least 11 groups of experiments.

[0049] The S5 is specifically:

[0050] Feature extraction was performed on the above 11 groups of tests to obtain the corresponding characteristic values of the 11 groups of tests. The mean absolute percentage error (MAPE) was used to analyze the influence of tool wear on the force characteristics, as follows:

[0051] Solve the MAPE of the TS2 and TS3 experiments relative to the TS1 experiment, which are expressed as: MAPE2 = (TS2-TS1) / TS1, MAPE3 = (TS3-TS1) / TS1;

[0052] Solve the MAPE of groups TS4 to TS11 relative to group TS3, which are expressed as: MAPE4 = (TS4-TS3) / TS3, MAPE5 = (TS5-TS3) / TS3, ..., MAPE11 = (TS11-TS3) / TS3;

[0053] Solve for the average change value, which is expressed as: MAPEa_1_3=(MAPE2+MAPE3) / 2,

[0054] MAPEa_3_11=(MAPE4+MAPE5+…,+MAPE11) / 8.

[0055] The S6 is specifically:

[0056] Step 1: Installation of dynamometer and workpiece and tool setting;

[0057] Step 2: Use the workpiece to perform cutting processing to cause tool wear;

[0058] Step 3: Use the dynamometer to collect the milling force signal, which is recorded as the T1 group signal;

[0059] Step 4: Extract features from the collected signals;

[0060] Step 5: Use the workpiece to perform cutting to increase the wear of the large wear area;

[0061] Step 6: Use the dynamometer to collect the milling force signal, which is recorded as the T2 group signal;

[0062] Step 7: Extract features from the collected signals;

[0063] Step 8: Analyze the absolute change MAPE value of the skewness feature of the two groups of signals T1 and T2. When the skewness feature MAPE value is significantly higher than other features, the wear state of the selected end mill is non-uniform wear. When the skewness feature MAPE value does not change significantly relative to other features, the wear state of the selected end mill is uniform wear.

[0064] Step 9: When the tool is in a uniform wear state, repeat steps 5, 6, 7, and 8, analyze the absolute change in skewness (MAPE) value, and determine the tool wear state again.

[0065] The step 1 is specifically as follows:

[0066] Install a three-axis dynamometer, camera, and workpiece inside the machine tool;

[0067] The dynamometer is installed on the side wall near the machine tool spindle, and a clamp is used to achieve a tight connection between the dynamometer and the machine tool. The workpiece is installed in the center of the machine tool, and a clamp is used to achieve a tight connection between the workpiece and the machine tool. The image acquisition device is installed on the side wall near the machine tool spindle, and a clamp is used to achieve a tight connection between the acquisition device and the machine tool. The tool is installed on the machine tool spindle, and the machine tool spindle is controlled to reach a position near the workpiece for tool setting. In the handwheel mode, the machine tool spindle is controlled to move the tool to the vicinity of the three intersecting and perpendicular end faces of the workpiece to complete the workpiece tool setting.

[0068] Establish the workpiece coordinate system, and again use the handwheel mode to control the machine tool spindle to move the tool near the dynamometer system to complete the tool setting work of the dynamometer system;

[0069] Establish the dynamometer system coordinate system. Control the machine tool spindle in handwheel mode to move the tool near the image acquisition device. Debug the vision system and establish the image acquisition system coordinate system.

[0070] The steps 1 to 9 as the uniform wear state experimental method are specifically as follows:

[0071] Further, the step 2 is: obtaining the initial uniform wear state of the tool, recorded as State 01, setting milling parameters, using a camera to record the wear image of the uniform wear state State 01, and marking the wear value VB;

[0072] Further, the step 3 is: obtaining a milling force signal of the uniform wear state State 01 after machining, which is recorded as T01;

[0073] The further step 5 is: continue processing, obtain the uniform wear state of the tool again, record it as State 02, set the milling parameters, increase the total cutting length, use a camera to record the wear image of the uniform wear state State 02, and mark the wear value VB;

[0074] Further, the step 6 is: obtaining a milling force signal of the uniform wear state State 02 after machining, which is recorded as T02;

[0075] The further step 8 is: the milling force signals F of the T01 group and the T02 group have been obtained. x 、F y and F tot Perform feature extraction and obtain eigenvalues; solve the MAPE value based on the eigenvalues of the two sets of signals extracted, and solve the formula:

[0076] MAPE=|(T02-T01) / T01|.

[0077] The steps 1 to 9 as the experimental method for non-uniform wear state are specifically as follows:

[0078] Replace the new tool, define the cutting edge, and specify a one-dimensional coordinate system along the tool axis: the tool tip is the origin and points to the tool handle; divide the cutting edge into two equal parts, A and B;

[0079] Further, the step 2 is: obtaining a uniform wear state of the tool, recorded as State 001, setting milling parameters, with the tool cutting edge involved in cutting as segment A, using a camera to record a wear image of the uniform wear state State 001, and marking the wear value VB;

[0080] Further, the step 3 is: obtaining a milling force signal corresponding to the uniform wear state State 001, which is recorded as T001;

[0081] Further, step 5 is: obtaining a uniform non-tool wear state, recorded as State 002, setting milling parameters, the tool cutting edge involved in cutting is segment B, the total cutting length is reduced, using a camera to record the wear image of the uniform wear state State 002, and marking the wear value VB;

[0082] Further, the step 6 is: obtaining a milling force signal corresponding to the non-uniform wear state State 002, which is recorded as T002;

[0083] The further step 8 is: the milling force signals F of the T001 group and the T002 group have been obtained. x 、F y and F tot Perform feature extraction and obtain eigenvalues; solve the MAPE value based on the eigenvalues of the two sets of signals extracted, and solve the formula:

[0084] MAPE=|(T002-T001) / T001|.

[0085] Beneficial effects: The present invention aims at the wear state of the milling cutter and proposes an online identification method for the non-uniform wear of the end mill. On the one hand, this method can more effectively identify the wear state of the milling cutter and quickly determine whether the wear state is in a non-uniform wear state. On the other hand, when the milling cutter has a large cutting depth, due to the limitations of the camera's field of view, a camera's field of view cannot capture all the wear information of the end mill, and it is impossible to obtain the full picture of the wear of the milling cutter. This method can more quickly identify the wear state of the milling cutter and determine whether the milling cutter is in a non-uniform wear state. For workpieces with relatively high processing precision, the cutter can be replaced in time, avoiding the processing deviation caused by tool problems affecting product quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0086] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0087] Figure 1 It is the mathematical geometry model of the blade of the present invention.

[0088] Figure 2 This is the "tool-workpiece" geometric relationship for milling processing in the present invention.

[0089] Figure 3 This is the milling force model of the present invention.

[0090] Figure 4 It is designed for the simulation experiment of the present invention.

[0091] Figure 5 This is the simulation experiment design of the present invention.

[0092] Figure 6 Extract features for simulation data of the present invention.

[0093] Figure 7 It is the MAPE value of the simulation data feature of the present invention.

[0094] Figure 8 This is the characteristic result of the simulation data of the present invention.

[0095] Figure 9 This is the experimental site of the present invention.

[0096] Figure 10 This is the sensitivity analysis of the uniform wear characteristics of the experiment in this invention.

[0097] Figure 11 This is the non-uniform wear process of the present invention experiment.

[0098] Figure 12 The wear image and measured value of the tool edge 3 of the present invention.

[0099] Figure 13 This is the time domain signal diagram and time domain signal period of the non-uniform wear experiment T001 of the present invention.

[0100] Figure 14 This is the time domain signal diagram and time domain signal period of the non-uniform wear experiment T002 of the present invention.

[0101] Figure 15 This is the sensitivity analysis of the experimental non-uniform wear characteristics of the present invention.

[0102] Figure 16 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0103] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0104] In the description of the present invention, it should be understood that the terms "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore should not be understood as limiting the present invention.

[0105] In the present invention, unless otherwise expressly specified or limited, a first feature being "above" or "below" a second feature may include the first and second features being in direct contact, or may include the first and second features being in contact not directly but through another feature between them. Furthermore, a first feature being "above," "above," and "above" a second feature may include the first feature being directly above or obliquely above the second feature, or may simply mean that the first feature is higher in level than the second feature. A first feature being "below," "below," and "below" a second feature may include the first feature being directly below or obliquely below the second feature, or may simply mean that the first feature is lower in level than the second feature.

[0106] like Figures 1 to 16 As shown, an online detection method for uneven wear of an end mill comprises the following steps:

[0107] S1. The cutting edge of the end mill is a spatial spiral curve. The tool is mathematically described to obtain the instantaneous cutting thickness;

[0108] S2. Use the micro-element cutting force instantaneous mechanical force model to establish a milling force model considering wear. The resultant milling force is expressed as the matrix Mat dFx The algebraic sum of all elements in ;

[0109] S3. Design a milling force simulation test for tool non-uniform wear and analyze the milling force signal characteristics that can reflect tool non-uniform wear;

[0110] S4, extracting time domain and frequency domain features of the milling force signal obtained from the simulation experiment of S3, and using eigenvalues to quantitatively describe the signal information;

[0111] S5. Analyze the correlation between the milling force signal characteristic value and tool wear, and solve the MAPE to obtain the milling force signal characteristic with the highest correlation;

[0112] S6. Collect the milling force signal during the machining process, and identify the non-uniform wear state of the tool during the machining process based on the milling force signal feature with the highest correlation obtained in S5.

[0113] The S1 is specifically:

[0114] S1.1. Describe the tool mathematically and establish the tool Cartesian coordinate system X T Y T Z T O T ;

[0115] Among them, Z T The axis coincides with the tool axis and its positive direction points to the tool handle. T -Y T The plane coincides with the plane formed by the tool tip, Y T The axis coincides with the tool tip, and the blades are numbered i=1, 2, 3, ..., N in sequence. The numbering order is counterclockwise rotation around the positive direction of the Z axis, and the blade with number i=1 coincides with the Y T The axes intersect, and N is the number of tool edges;

[0116] The blade is discretized along its axis, and the discrete unit is marked as P Eij ;

[0117] Among them, i represents the blade, j represents the blade along Z T Discrete unit number in the positive direction of the axis, P E11 Indicates the j = 1 discrete point corresponding to the i = 1 blade, the maximum number of discrete units M, see Figure 1 ;

[0118] S1.2. Establish the workpiece Cartesian coordinate system X W Y W Z W O W ;

[0119] Among them, X W The positive direction of the axis points to the tool feed direction, Z W The positive direction of the axis points to the direction of the tool. During the machining process, it is assumed that the workpiece does not move and only the tool moves, that is, the workpiece coordinate system is a fixed coordinate system. Let Y T Axis and Y W When the axis direction is consistent, it is time t0. Then at any time t, the discrete point P on the blade Eij The corresponding angle Φ i,j for:

[0120]

[0121] Among them, Φ i,j Point P Eij By Y W Axis starts around Z W The direction angle of the axis in the clockwise direction is in the range of [0°, 360°], ω is the tool rotation angular velocity (rad / s), β is the tool helix angle, D is the tool diameter, and dz is the discrete spacing of the blade along its axis. Figure 2 ;

[0122] Click P Eij The corresponding instantaneous cutting thickness is expressed as:

[0123]

[0124] Among them, f t is the feed rate per tooth, in mm / tooth, Φ out is the cut-out angle, its value is π, the cut-in angle Φ in It can be expressed as:

[0125] Φ in =π-acos((0.5Da e ) / 0.5D)

[0126] where a e is the cutting width;

[0127] Based on the above method, the instantaneous cutting thickness corresponding to all blades at each moment can be obtained and stored in the matrix Mat hw In the expression:

[0128]

[0129] The S2 is specifically:

[0130] Using the instantaneous mechanical force model of micro-element cutting force, the milling force corresponding to the micro-element of the cutting edge at a certain time t is expressed as:

[0131] dF t =K tc h wij dz=K tc f t sin(Φ i,j )dz

[0132] dF r =K rc h wij dz=K rc f t sin(Φ i,j )dz

[0133] Among them, dF t is the blade microelement tangential force, and its direction coincides with the direction of instantaneous cutting speed; dF r is the radial force of the blade, its direction is perpendicular to dF t And point to the workpiece, see Figure 3 , K tc and K rc are the tangential force coefficient and radial force coefficient respectively;

[0134] The tangential force and radial force are converted into the force F along the feed direction in the workpiece coordinate system through coordinate transformation x and the force F perpendicular to the feed direction y , the transformation formula is:

[0135] dF x =dF r sin(Φ i,j )+dF t cos(Φ i,j )

[0136] dF y =dF r cos(Φ i,j )-dF t sin(Φ i,j )

[0137] Considering the helix angle of the milling cutter and multi-edge cutting, the time domain signal expression of the milling force is established:

[0138]

[0139]

[0140] in,

[0141] The milling force modeling considering wear, that is, the milling force coefficient is used to represent wear, and the milling force coefficient is expressed in matrix form:

[0142]

[0143] The infinitesimal force matrix can be obtained:

[0144] Mat dFx =Mat Krc Mat hw ·dz·sin(Φ i,j )+Mat Ktc Mat hw ·dz·cos(Φ i,j )

[0145] The resultant milling force can be expressed as the matrix Mat dFxThe algebraic sum of all elements in .

[0146] The S3 is specifically:

[0147] The tool is set along the axis direction, with the tool tip as the starting point, and the discrete points are divided into three parts, marked as: CE-1, CE-2, CE-3, and 11 groups of experiments are designed. Figure 4 and Figure 5 .

[0148] The S4 is specifically:

[0149] The time domain and frequency domain features of the milling force signal obtained from the simulation data are extracted, and the characteristic value is used to quantitatively describe the signal information. 16 features are selected to respectively describe F x / F y / F tot Perform feature extraction and finally obtain 48 eigenvalues, see Figure 6 The specific features and their order are: 1. Maximum value, 2. Minimum value, 3. Peak-to-peak value, 4. Average value, 5. Absolute mean, 6. Standard deviation, 7. Root mean square, 8. Skewness, 9. Kurtosis, 10. Form factor, 11. Crest factor, 12. Pulse factor, 13. Margin factor, 14. Center of gravity frequency, 15. Root mean square frequency, 16. Frequency standard deviation.

[0150] The S5 is specifically:

[0151] Feature extraction was performed on the above 11 groups of tests to obtain the corresponding characteristic values of the 11 groups of tests. The mean absolute percentage error (MAPE) was used to analyze the influence of tool wear on the force characteristics. The larger the MAPE value, the higher the correlation between the characteristics and the uneven wear state, as follows:

[0152] Solve the MAPE of the TS2 and TS3 experiments relative to the TS1 experiment, which are expressed as: MAPE2 = (TS2-TS1) / TS1, MAPE3 = (TS3-TS1) / TS1;

[0153] Solve the MAPE of groups TS4 to TS11 relative to group TS3, which are expressed as: MAPE4 = (TS4-TS3) / TS3, MAPE5 = (TS5-TS3) / TS3, ..., MAPE11 = (TS11-TS3) / TS3;

[0154] Solve for the average change value, which is expressed as: MAPEa_1_3=(MAPE2+MAPE3) / 2,

[0155] MAPEa_3_11=(MAPE4+MAPE5+…,+MAPE11) / 8.

[0156] For simulation results, see Figure 7 and Figure 8 The results show that the deflection characteristic is very sensitive to uneven wear conditions but not to uniform wear conditions. Specifically, changes in deflection indicate uneven wear of the cutting tool and do not change VB. Its value remains unchanged, but when the tool is subjected to uneven wear, its value changes significantly, especially in the force signal perpendicular to the tool feed direction, where the change is most pronounced.

[0157] "Skewness" is a statistical feature of digital signals. There is no need to calculate the tool milling force model. It only needs to monitor the obtained milling force signal data and then extract statistical features of the data to identify the non-uniform wear state in real time and efficiently.

[0158] The S6 is specifically:

[0159] Step 1: Installation of dynamometer and workpiece and tool setting;

[0160] Step 2: Use the workpiece to perform cutting processing to cause tool wear;

[0161] Step 3: Use the dynamometer to collect the milling force signal, which is recorded as the T1 group signal;

[0162] Step 4: Extract features from the collected signals;

[0163] Step 5: Use the workpiece to perform cutting to increase the wear of the large wear area;

[0164] Step 6: Use the dynamometer to collect the milling force signal, which is recorded as the T2 group signal;

[0165] Step 7: Extract features from the collected signals;

[0166] Step 8: Analyze the absolute change MAPE value of the skewness feature of the two groups of signals T1 and T2. When the skewness feature MAPE value is significantly higher than other features, the wear state of the selected end mill is non-uniform wear. When the skewness feature MAPE value does not change significantly relative to other features, the wear state of the selected end mill is uniform wear.

[0167] Step 9: When the tool is in a uniform wear state, repeat steps 5, 6, 7, and 8 to analyze the absolute change in the deflection (MAPE) value and judge the tool wear state again. When the tool is judged to be in a non-uniform wear state, that is, the tool does not meet the requirements for continuing to perform finishing, a suitable tool needs to be replaced.

[0168] The step 1 is specifically as follows:

[0169] Install a three-axis dynamometer, camera, and workpiece inside the machine tool;

[0170] The dynamometer is installed on the side wall near the machine tool spindle, and a clamp is used to achieve a tight connection between the dynamometer and the machine tool. The workpiece is installed in the center of the machine tool, and a clamp is used to achieve a tight connection between the workpiece and the machine tool. The image acquisition device is installed on the side wall near the machine tool spindle, and a clamp is used to achieve a tight connection between the acquisition device and the machine tool. The tool is installed on the machine tool spindle, and the machine tool spindle is controlled to reach a position near the workpiece for tool setting. In the handwheel mode, the machine tool spindle is controlled to move the tool to the vicinity of the three intersecting and perpendicular end faces of the workpiece to complete the workpiece tool setting.

[0171] Establish the workpiece coordinate system, and again use the handwheel mode to control the machine tool spindle to move the tool near the dynamometer system to complete the tool setting work of the dynamometer system;

[0172] Establish the dynamometer system coordinate system. Control the machine tool spindle in handwheel mode to move the tool near the image acquisition device. Debug the vision system and establish the image acquisition system coordinate system.

[0173] The three-axis dynamometer is responsible for collecting cutting force signal data, and the camera is responsible for recording wear image data. The role of the workpiece is to cause tool wear. The experimental equipment is as shown in Table 1:

[0174] Table 1 Test equipment

[0175]

[0176]

[0177] The steps 1 to 9 as the uniform wear state experimental method are specifically as follows:

[0178] Further, step 2 is as follows: obtaining the initial uniform wear state of the tool, recorded as State 01, setting the milling parameters as follows: cutting depth ap = 1.6 mm, cutting width ae = 1.6 mm, spindle speed n = 6000 r / min, feed speed f = 720 mm / min, sampling frequency fs = 50 kHz, workpiece size 300 mm × 160 mm × 50 mm, total cutting length 160 × 15 = 2400 mm, using a camera to record the wear image of the uniform wear state State 01, and marking the wear value VB;

[0179] Further, the step 3 is: obtaining the milling force signal of the uniform wear state State 01 after processing, recorded as T01; the milling parameters are: cutting depth ap = 1.6 mm, cutting width ae = 1.6 mm, spindle speed n = 6000 r / min, feed speed f = 720 mm / min, sampling frequency fs = 50 kHz.

[0180] Further, step 5 is as follows: continue machining, obtain the uniform wear state of the tool again, record it as State 02, set the milling parameters as follows: cutting depth ap = 1.6 mm, cutting width ae = 1.6 mm, spindle speed n = 6000 rpm, feed speed f = 720 mm / min, sampling frequency fs = 50 kHz, workpiece size 300 mm × 160 mm × 50 mm, total cutting length 160 × 75 = 12000 mm, use a camera to record the wear image of the uniform wear state State 02, and mark the wear value VB;

[0181] Further, the step 6 is: obtaining the milling force signal of the uniform wear state State 02 after processing, recorded as T02; the milling parameters are: cutting depth ap = 1.6 mm, cutting width ae = 1.6 mm, spindle speed n = 6000 r / min, feed speed f = 720 mm / min, sampling frequency fs = 50 kHz.

[0182] The further step 8 is: the milling force signals F of the T01 group and the T02 group have been obtained. x 、F y and F tot Perform feature extraction and finally obtain 48 eigenvalues;

[0183] According to the eigenvalues of the two sets of signals extracted, the MAPE value is solved and the formula is solved:

[0184] MAPE=|(T02-T01) / T01|

[0185] The specific features and their order are: 1. Maximum value, 2. Minimum value, 3. Peak-to-peak value, 4. Average value, 5. Absolute mean, 6. Standard deviation, 7. Root mean square, 8. Skewness, 9. Kurtosis, 10. Shape factor, 11. Peak factor, 12. Pulse factor, 13. Margin factor, 14. Center of gravity frequency, 15. Root mean square frequency, 16. Frequency standard deviation. Three groups of F x 、F y and F tot The force signal has 48 features in total. Table 2 shows the obtained MAPE values.

[0186] Table 2 Characteristic formula

[0187]

[0188]

[0189] Table 3 MAPE of T01 and T02 groups under uniform wear

[0190]

[0191]

[0192] From Table 3 and Figure 10 It can be seen that F x and F y The skewness eigenvalue does not change significantly, and thus F tot (F x and F y The skewness characteristics are also not obvious, the MAPE value of the absolute change of skewness is less than 1, and the milling cutter is in a state of uniform wear.

[0193] Table 4 Milling cutter wear values VB for groups T01 and T02 under uniform wear

[0194]

[0195] Table 4 shows that the camera-measured wear values for the T01 milling cutter fluctuate between 0.040 mm and 0.049 mm, while for the T02 milling cutter, the VB fluctuates between 0.080 mm and 0.089 mm. The milling cutters are experiencing uniform wear, and the VB value is increasing. This confirms the correctness of the conclusion.

[0196] The steps 1 to 9 as the experimental method for non-uniform wear state are specifically as follows:

[0197] Replace the new tool, define the cutting edge, and specify a one-dimensional coordinate system along the tool axis: the tool tip is the origin and points to the tool handle; divide the cutting edge into two sections, A and B; section A corresponds to the 0-3mm part of the tool edge, and section B corresponds to the 3-6mm part of the tool edge.

[0198] Further, step 2 is as follows: obtaining a uniform wear state of the tool, recorded as State 001, setting milling parameters as follows: cutting depth ap = 3 mm, cutting width ae = 1 mm, spindle speed n = 6000 r / min, feed speed f = 720 mm / min, sampling frequency fs = 50 kHz, tool length involved in cutting is 0-3 mm, workpiece size is 300 mm × 160 mm × 50 mm, and total cutting length is 160 × 320 = 51200 mm, using a camera to record a wear image of the uniform wear state State 001 and marking the wear value VB;

[0199] Further, the step 3 is: obtaining the milling force signal corresponding to the uniform wear state State 001, recorded as T001; the milling parameters are ap=6mm, ae=1mm, n=6000r / min, f=720mm / min, the tool involved in cutting is 0~6mm, and the sampling frequency fs=50kHz.

[0200] Further, step 5 is as follows: obtaining a uniform non-tool wear state, denoted as State 002, setting milling parameters as follows: cutting depth ap = 3 mm, cutting width ae = 1 mm, spindle speed n = 6000 rpm, feed rate f = 720 mm / min, sampling frequency fs = 50 kHz, tool diameter of 3 to 6 mm, workpiece size of 300 mm × 160 mm × 50 mm, and total cutting length of 100 × 200 = 20,000 mm, using a camera to record a wear image of the uniform wear state State 002 and marking the wear value VB;

[0201] Further, step 6 is: obtaining the milling force signal corresponding to the uneven wear state State 002, recorded as T002; the milling parameters are ap=6mm, ae=1mm, n=6000r / min, f=720mm / min, the tool involved in cutting is 0~6mm, and the sampling frequency fs=50kHz.

[0202] The further step 8 is: respectively performing the operations on the acquired T001 group and T002 group signals F x 、F y and F tot Perform feature extraction and finally obtain 48 eigenvalues;

[0203] According to the eigenvalues of the two sets of signals extracted, the MAPE value is solved and the formula is solved:

[0204] MAPE=|(T002-T001) / T001|

[0205] The specific features and their order are: 1. Maximum value, 2. Minimum value, 3. Peak-to-peak value, 4. Average value, 5. Absolute mean, 6. Standard deviation, 7. Root mean square, 8. Skewness, 9. Kurtosis, 10. Shape factor, 11. Peak factor, 12. Pulse factor, 13. Margin factor, 14. Center of gravity frequency, 15. Root mean square frequency, 16. Frequency standard deviation. Three groups of F x 、F y and F tot The force signal has 48 features in total. Table 5 shows the MAPE values of the signals of group T001 and group T002.

[0206] Table 5 MAPE of T001 and T002 groups under uneven wear

[0207]

[0208]

[0209] From Table 5 and Figure 15 It can be seen that F yThe skewness characteristic value changes significantly, the MAPE value of the absolute change of skewness is greater than 1, and the absolute change rate is much higher than other characteristics. The milling cutter is in a state of non-uniform wear.

[0210] The wear value of the milling cutter was measured by camera. The measurement results are shown in Table 6. It can be seen that the milling cutter is in a state of non-uniform wear, which verifies the correctness of the conclusion. x The skewness characteristics of Figure 13 and Figure 14 It can be seen that the noise of the signal is relatively large, which leads to differences between the results and the measured results.

[0211] A total of 24 tool wear images were collected, and the maximum wear value of the tool flank was measured, as shown in Table 6. Table 6 shows that the wear values at different positions on the same cutting edge fluctuate significantly. When the absolute change in the fluctuating wear value is greater than 0.6, the milling cutter can be considered to be in a state of non-uniform wear.

[0212] It can be seen from Table 6 that there is a significant difference in the wear values VB of the milling cutter in the +0mm~+3mm and +3mm~+6mm sections. The wear value VB of the +0mm~+3mm section is more than twice that of the +3mm~+6mm section. It can be considered that this tool is in the stage of non-uniform wear.

[0213] Table 6 Maximum wear values of tool flank in T001 and T002 experiments

[0214]

[0215] Through comparative analysis between this method and manual measurement with a camera, the milling cutter was found to be in a state of uniform wear, which verified the correctness of the conclusion.

[0216] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.

[0217] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An online detection method for uneven wear of an end mill, characterized in that: The following steps are involved: S1. The cutting edge of the end mill is a spatial spiral curve. The tool is mathematically described to obtain the instantaneous cutting thickness; S2. Use the micro-element cutting force instantaneous mechanical force model to establish a milling force model considering wear. The resultant milling force is expressed as the matrix Mat dFx The algebraic sum of all elements in ; S3. Design a milling force simulation test for tool non-uniform wear and analyze the milling force signal characteristics that can reflect tool non-uniform wear; S4, extracting time domain and frequency domain features of the milling force signal obtained from the simulation experiment of S3, and using eigenvalues to quantitatively describe the signal information; S5. Analyze the correlation between the milling force signal characteristic value and tool wear, and solve the MAPE to obtain the milling force signal characteristic with the highest correlation; S6, collecting the milling force signal during the machining process, and identifying the non-uniform wear state of the tool during the machining process based on the most relevant milling force signal feature obtained in S5; The S1 is specifically: S1.

1. Describe the tool mathematically and establish the tool Cartesian coordinate system X T Y T Z T O T ; Among them, Z T The axis coincides with the tool axis and the positive direction points to the tool handle. T -Y T The plane coincides with the plane formed by the tool tip, Y T The axis coincides with the tool tip, and the blades are numbered i=1, 2, 3, ..., N in sequence. The numbering order is counterclockwise rotation around the positive direction of the Z axis, and the blade with number i=1 coincides with the Y T The axes intersect, and N is the number of tool edges; The blade is discretized along its axis, and the discrete unit is marked as P Eij , Among them, i represents the blade, j represents the blade along Z T Discrete unit number in the positive direction of the axis, P E11 represents the j=1th discrete point corresponding to the i=1th blade; S1.

2. Establish the workpiece Cartesian coordinate system X W Y W Z W O W ; Among them, X W The positive direction of the axis points to the tool feed direction, Z W The positive direction of the axis points to the direction of the tool. During the machining process, it is assumed that the workpiece does not move and only the tool moves, that is, the workpiece coordinate system is a fixed coordinate system. Let Y T Axis and Y W When the axis direction is consistent, it is time t0. Then at any time t, the discrete point P on the blade Eij The corresponding angle Φ i,j for: Among them, Φ i,j Point P Eij By Y W Axis starts around Z W The direction angle of the axis in the clockwise direction is in the range of [0°, 360°], ω is the tool rotation angular velocity (rad / s), β is the tool helix angle, D is the tool diameter, and dz is the discrete spacing of the tool edge along its axis. Click P Eij The corresponding instantaneous cutting thickness is expressed as: Among them, f t is the feed rate per tooth, in mm / tooth, Φ out is the cut-out angle, its value is π, the cut-in angle Φ in It can be expressed as: F in =π-acos((0.5Da e ) / 0.5D) where a e is the cutting width; Based on the above method, the instantaneous cutting thickness corresponding to all blades at each moment can be obtained and stored in the matrix Mat hw In the expression: The S2 is specifically: Using the instantaneous mechanical force model of micro-element cutting force, the milling force corresponding to the micro-element of the cutting edge at a certain time t is expressed as: dF t =K tc h wij dz=K tc f t sin(Φ i,j )dz dF r =K rc h wij dz=K rc f t sin(Φ i,j )dz Among them, dF t is the blade microelement tangential force, and its direction coincides with the direction of instantaneous cutting speed; dF r is the radial force of the blade, its direction is perpendicular to dF t And points to the workpiece, K tc and K rc are the tangential force coefficient and radial force coefficient respectively; Furthermore, the tangential force and radial force are transformed into the force F along the feed direction in the workpiece coordinate system through coordinate transformation. x and the force F perpendicular to the feed direction y , the transformation formula is: dF x =dF r sin(Φ i,j )+dF t cos(Φ i,j ) dF y =dF r cos(Φ i,j )-dF t sin(Φ i,j ) Considering the helix angle of the milling cutter and multi-edge cutting, the time domain signal expression of the milling force is established: in, The milling force modeling considering wear, that is, the milling force coefficient is used to represent wear, and the milling force coefficient is expressed in matrix form: The infinitesimal force matrix can be obtained: Mast dFx =Mat Krc ·Mast hw ·dz·sin(Φ i,j )+Mat Ktc ·Mast hw ·dz·cos(Φ i,j ) The resultant milling force can be expressed as the matrix Mat dFx The algebraic sum of all elements in ; The S3 is specifically: Set discrete points along the tool axis with the tool tip as the starting point. Divide the discrete points into at least three parts, marked as CE-1, CE-2, and CE-3, and design at least 11 groups of tests. The S5 is specifically: Feature extraction is performed on at least 11 groups of tests mentioned above to obtain characteristic values corresponding to at least 11 groups of tests. The influence of tool wear on the force characteristics is analyzed using the mean absolute percentage error (MAPE), as follows: Solve the MAPE of the TS2 and TS3 experiments relative to the TS1 experiment, which are expressed as: MAPE2 = (TS2-TS1) / TS1, MAPE3 = (TS3-TS1) / TS1; Solve the MAPE of groups TS4 to TS11 relative to group TS3, which are expressed as: MAPE4 = (TS4-TS3) / TS3, MAPE5 = (TS5-TS3) / TS3, ..., MAPE11 = (TS11-TS3) / TS3; Solve for the average change value, which is expressed as: MAPEa_1_3=(MAPE2+MAPE3) / 2, MAPEa_3_11=(MAPE4+MAPE5+…,+MAPE11) / 8; The S6 is specifically: Step 1: Installation of dynamometer and workpiece and tool setting; Step 2: Use the workpiece to perform cutting processing to cause tool wear; Step 3: Use the dynamometer to collect the milling force signal, which is recorded as the T1 group signal; Step 4: Extract features from the collected signals; Step 5: Use the workpiece to perform cutting to increase the wear of the large wear area; Step 6: Use the dynamometer to collect the milling force signal, which is recorded as the T2 group signal; Step 7: Extract features from the collected signals; Step 8: Analyze the absolute change MAPE value of the skewness feature of the two groups of signals T1 and T2. When the skewness feature MAPE value is significantly higher than other features, the wear state of the selected end mill is non-uniform wear. When the skewness feature MAPE value does not change significantly relative to other features, the wear state of the selected end mill is uniform wear. Step 9: When the tool is in a uniform wear state, repeat steps 5, 6, 7, and 8, analyze the absolute change in skewness (MAPE) value, and determine the tool wear state again.

2. The method for online detection of uneven wear of end mills according to claim 1, characterized in that: The step 1 is specifically as follows: Install a three-axis dynamometer, camera, and workpiece inside the machine tool; The dynamometer is installed on the side wall near the machine tool spindle, and a clamp is used to achieve a tight connection between the dynamometer and the machine tool. The workpiece is installed in the center of the machine tool, and a clamp is used to achieve a tight connection between the workpiece and the machine tool. The image acquisition device is installed on the side wall near the machine tool spindle, and a clamp is used to achieve a tight connection between the acquisition device and the machine tool. The tool is installed on the machine tool spindle, and the machine tool spindle is controlled to reach a position near the workpiece for tool setting. In the handwheel mode, the machine tool spindle is controlled to move the tool to the vicinity of the three intersecting and perpendicular end faces of the workpiece to complete the workpiece tool setting. Establish the workpiece coordinate system, and again use the handwheel mode to control the machine tool spindle to move the tool near the dynamometer system to complete the tool setting work of the dynamometer system; Establish the dynamometer system coordinate system. Control the machine tool spindle in handwheel mode to move the tool near the image acquisition device. Debug the vision system and establish the image acquisition system coordinate system.

3. The online detection method for uneven wear of end mills according to claim 1, characterized in that: The steps 1 to 9 as the uniform wear state experimental method are specifically as follows: Further, the step 2 is: obtaining the initial uniform wear state of the tool, recorded as State 01, setting milling parameters, using a camera to record the wear image of the uniform wear state State 01, and marking the wear value VB; Further, the step 3 is: obtaining a milling force signal of the uniform wear state State 01 after machining, which is recorded as T01; The further step 5 is: continue processing, obtain the uniform wear state of the tool again, record it as State 02, set the milling parameters, increase the total cutting length, use a camera to record the wear image of the uniform wear state State 02, and mark the wear value VB; Further, the step 6 is: obtaining a milling force signal of the uniform wear state State 02 after machining, which is recorded as T02; The further step 8 is: the milling force signals F of the T01 group and the T02 group have been obtained. x 、F y and F tot Perform feature extraction and obtain eigenvalues, where F tot F x 、F y The resultant force; According to the eigenvalues of the two sets of signals extracted, solve the MAPE value and solve the formula: MAPE=|(T02-T01) / T01|.

4. The method for online detection of uneven wear of end mills according to claim 1, characterized in that: The steps 1 to 9 as the experimental method for non-uniform wear state are specifically as follows: Replace the new tool, define the cutting edge, and specify a one-dimensional coordinate system along the tool axis: the tool tip is the origin and points to the tool handle; divide the cutting edge into two equal parts, A and B; Further, the step 2 is: obtaining a uniform wear state of the tool, recorded as State 001, setting milling parameters, with the tool cutting edge involved in cutting as segment A, using a camera to record a wear image of the uniform wear state State 001, and marking the wear value VB; Further, the step 3 is: obtaining a milling force signal corresponding to the uniform wear state State 001, which is recorded as T001; Further, step 5 is: obtaining a non-uniform wear state, recorded as State 002, setting milling parameters, the cutting edge of the tool involved in cutting is segment B, the total cutting length is reduced, using a camera to record the wear image of the uniform wear state State 002, and marking the wear value VB; Further, the step 6 is: obtaining a milling force signal corresponding to the non-uniform wear state State 002, which is recorded as T002; The further step 8 is: the milling force signals F of the T001 group and the T002 group have been obtained. x 、F y and F tot Perform feature extraction and obtain eigenvalues, where F tot F x 、F y The resultant force; According to the eigenvalues of the two sets of signals extracted, solve the MAPE value and solve the formula: MAPE=|(T002-T001) / T001|.

Images