Tool wear prediction method based on part geometric features and cutting process parameters

Abstract

The present application relates to a kind of tool wear prediction methods based on part geometric feature and cutting process parameters, first complex profile parts are split into simple part geometric feature, for different part geometric feature, establish about part feature parameters, tool parameters and cutting process parameters Tool wear prediction regression model;According to the parameter range set up orthogonal experiment, measure the tool wear under different parameters, then each group tool wear, parameter logarithm is taken and is converted into the form of matrix, input into MATLAB software, the coefficient and exponential part of tool prediction regression model parameter are solved, the coefficient and exponential part logarithm is taken anti-function, obtain the tool wear prediction regression model under different characteristics;Finally, the significance test and residual analysis are carried out on tool wear prediction regression model, and the final tool wear prediction model is obtained.The parameters referred to in the present application are more, the application range is wider, and the tool wear prediction accuracy is higher.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of machining tool wear, and particularly relates to a tool wear prediction method based on part geometric features and cutting process parameters. BACKGROUND

[0002] With the continuous development of high-speed cutting technology, machining of mechanical parts requires high efficiency, high reliability, high precision and specialization of tools. Tool wear is inevitable in the cutting process, and it is very important to effectively predict tool wear under different conditions. Tool wear prediction can not only improve tool efficiency and reduce use cost, but also avoid the occurrence of insufficient workpiece machining precision, poor machining surface quality and serious workpiece scrap due to tool damage in cutting machining.

[0003] Tool wear is not only related to the materials of tool-work, but also related to cutting process parameters, tool parameters, part geometric features (i.e. the shape of the machined part) and the like. The current traditional tool prediction model has great limitations, and it is used to predict tool wear for a single material or several parameters. The prediction model has a small application range, involves few factors, and has low accuracy. Therefore, it is of great significance to establish a tool prediction model under the coupling of different characteristic parameters and cutting process parameters for different part geometric features. SUMMARY

[0004] In view of the problems of the current tool wear prediction model, such as less involved factors, small application range and low accuracy, the present application provides a tool wear prediction method based on part geometric features and cutting process parameters.

[0005] To solve the above problems, the present application adopts the following technical scheme:

[0006] The tool wear prediction method based on part geometric features and cutting process parameters comprises the following steps:

[0007] Step 1: split the complex part into part geometric features, the part geometric features are face features, hole features or groove features, and establish a tool wear prediction regression model for different features as follows:

[0008] (1) for face features:

[0009]

[0010] Wherein:

[0011] H is the tool wear amount;

[0012] K, i, w, x, o, p, q, s and t are constants;

[0013] n is the spindle speed, f z is the feed per tooth, a e is the radial depth of cut, D is the diameter of the milling cutter, γ0 is the milling cutter rake angle, α0 is the milling cutter relief angle, β0 is the milling cutter helix angle, and l is the milling cutter travel;

[0014] (2) for hole features:

[0015]

[0016] wherein:

[0017] j is a constant;

[0018] a p is the axial depth of cut;

[0019] (3) for groove features:

[0020]

[0021] Step two: according to the different feature parameter range, establish the corresponding orthogonal experiment table;

[0022] Step three: according to the orthogonal experiment table, the orthogonal experiment is carried out, and the tool wear under each group of parameters is obtained;

[0023] Step four: taking logarithm of both sides of the formula of the tool wear prediction regression model corresponding to the feature, establishing a multiple linear regression equation, and converting the multiple linear regression equation into a matrix form;

[0024] Step five: through the MATLAB software, an M file is established, the logarithmic values of each tool wear and the corresponding parameters obtained by the orthogonal experiment are input into the MATLAB software, the regress function is used for multiple linear regression calculation, the value of the regression coefficient is obtained, the logarithmic inverse function calculation of the value of the regression coefficient is carried out, each constant in the tool wear prediction regression model formula corresponding to the feature is obtained, and then the tool wear prediction regression model corresponding to the feature is obtained;

[0025] Step six: the significance test and residual analysis are carried out on the tool wear prediction regression model, and the final tool wear prediction model is obtained.

[0026] The present application has the following beneficial effects:

[0027] The tool wear prediction method provided by the present application is a prediction method based on part geometric features and cutting process parameters, firstly, the complex part is divided into simple part geometric features (surface, hole, groove), different features are established about part feature parameters (length, width, diameter, depth, length, width, height), tool parameters (diameter D, rake angle γ0, relief angle α0, helix angle β0) and cutting process parameters (spindle speed n, feed per tooth fz axial cutting depth a p The tool wear prediction regression model of the present application is established by taking the logarithm of the tool wear amount and the parameters, converting them into a matrix form, inputting them into the MATLAB software, and using the regress function to obtain the coefficients and the exponential part of the tool wear prediction regression model. BRIEF DESCRIPTION OF DRAWINGS

[0028] Figure 1 The flow chart of the tool wear prediction method based on the part geometric features and the cutting process parameters according to the embodiments of the present application is shown in the figure.

[0029] Figure 2 The schematic diagram of the hard alloy milling AL2024 material in the embodiments of the present application is shown in the figure.

[0030] Figure 3 The residual analysis diagram of the tool wear prediction regression model in the embodiments of the present application is shown in the figure. DETAILED DESCRIPTION

[0031] The technical solutions and advantages of the present application are described in detail below in combination with the figures and specific embodiments.

[0032] The present application provides a tool wear prediction method based on the part geometric features and the cutting process parameters to solve the problems of the current tool wear prediction model. The method first simplifies the complex-shaped parts (such as box-shaped parts, shell-shaped parts, disc-shaped parts, and special-shaped parts) into simple part geometric features (plane, hole, and groove), and then establishes a tool wear prediction regression model. The experimental parameters and the experimental results (tool wear amount) are taken as the logarithm, converted into a matrix, input into the MATLAB software, and the regression coefficient of the regression model is obtained by using the regress function. The logarithmic inverse function of the regression coefficient is taken to obtain the coefficient of the tool wear prediction regression model. Finally, the value of the statistical variable stats is calculated to determine the significance of the tool wear prediction formula, and the final tool wear amount prediction model is obtained.

[0033] As Figure 1As shown, the embodiment of the present application provides a tool wear prediction method based on part geometric features and cutting process parameters, which specifically comprises the following steps:

[0034] Step one: convert the complex profile part into a simple part geometric feature, the part geometric feature is one of a surface feature (plane, the involved parameters are length and width), a hole feature (the involved parameters are hole depth and diameter), and a groove feature (the involved parameters are length, width, and height); for different part geometric features, establish a tool wear prediction regression model as follows:

[0035] (1) for the surface feature:

[0036]

[0037] Wherein:

[0038] H is the tool wear amount;

[0039] K, i, w, x, o, p, q, and s are constants;

[0040] n is the spindle speed, f z is the feed per tooth, a e is the radial cutting depth (milling width), D is the milling cutter diameter, γ0 is the milling cutter rake angle, α0 is the milling cutter relief angle, β0 is the milling cutter helix angle, and l is the milling cutter stroke;

[0041] (2) for the hole feature:

[0042]

[0043] Wherein:

[0044] H is the tool wear amount;

[0045] K, i, w, j, o, p, q, and s are constants;

[0046] n is the spindle speed, f z is the feed per tooth, a p is the axial cutting depth, D is the milling cutter diameter, γ0 is the milling cutter rake angle, α0 is the milling cutter relief angle, and β0 is the milling cutter helix angle;

[0047] (3) for the groove feature:

[0048]

[0049] Wherein:

[0050] H is the tool wear amount;

[0051] K, i, j, w, x, o, p, q, s, and t are constants;

[0052] n is the spindle speed, fz feed per tooth, a p axial depth of cut, a e radial depth of cut (milling width), D is the diameter of the milling cutter, γ0is the milling cutter rake angle, a0is the milling cutter relief angle, β0is the milling cutter helix angle, and l is the milling cutter travel;

[0053] Step two: according to different part geometric features, set the parameter range, and establish the orthogonal experiment table corresponding to the part geometric features.

[0054] Step three: according to the orthogonal experiment table established in step two, carry out orthogonal experiment, and obtain the tool wear under each group of parameters.

[0055] Step four: take the logarithm of both sides of the formula of the tool wear prediction regression model corresponding to the selected part geometric features (surface feature or hole feature or groove feature), establish a multiple linear regression equation, and convert the multiple linear regression equation into a matrix form.

[0056] Taking the hole feature as an example, taking the logarithm of both sides of formula (2) obtains:

[0057] lgH = lgK + i·lgn + w·lgf z + j·lga p + o·lgD + p·lgγ0+ q·lgα0+ s·lgβ0 (4)

[0058] Let y = lgH, a0= lgK, a1 = i, a2 = w, a3 = j, a4 = o, a5 = p, a6 = q, a7 = s, x1 = lgn, x2 = lgf z , x3 = lga p , x4 = lgD, x5 = lgγ0, x6 = lgα0, x7 = lgβ0;

[0059] Then formula (4) can be rewritten as formula (5) showing the linear relationship between the dependent variable and the independent variable:

[0060] Y = a0+ a1·x1+ a2·x2+ a3·x3+ a4·x4+ a5·x5+ a6·x6+ a7·x7 (5)

[0061] Use y i to represent the experimental results, and the independent variables are x i1 , x i2 , x i3 , x i4 , x i5 , x i6 , x i7 , and the experimental error is ε i , where the number of experiments i = 1 ~ 18. The multiple linear regression equation is as follows:

[0062]

[0063] Equation (6) is expressed by matrix as follows:

[0064] Y = Xλ + ε (7)

[0065] wherein:

[0066]

[0067]

[0068]

[0069]

[0070] The least square method is used to estimate the parameter λ, and the regression coefficients b0, b1, b2, b3, b4, b5, b6, b7 are the least square estimates of the parameters λ0, λ1, λ2, λ3, λ4, λ5, λ6, λ7 (including the influence of ε), and the regression equation can be expressed as:

[0071]

[0072] wherein is a statistical variable, wherein:

[0073]

[0074] Step five: an M file is established by MATLAB software, each group of tool wear values obtained by orthogonal experiment and the logarithmic values of the parameters corresponding to the tool wear are input into the MATLAB software, the regress function is used for multiple linear regression calculation, the values of the regression coefficients b0, b1, b2, b3, b4, b5, b6, b7 in matrix (9) are obtained, and then the logarithmic inverse functions of the regression coefficients b0, b1, b2, b3, b4, b5, b6, b7 are taken, the values of each constant K, i, w, j, o, p, q, s in the tool wear prediction regression model formula corresponding to the hole feature are calculated, so that the tool wear prediction regression model based on the cutting process parameters and the tool parameters under the hole feature is obtained. The calculation process of the tool wear prediction regression model for the surface feature and the groove feature is similar to that of the hole feature, and details are not described here.

[0075] Step six: the tool wear prediction regression model obtained in step five is subjected to significance test, and whether the value of the statistical variable stats satisfies the goodness of fit R 2Whether close to 1, P < alpha = 0.05, while residual analysis, determine the accuracy of the tool wear prediction regression model, to pass the significance test and the model after residual analysis is the most final tool wear prediction model, for the tool wear prediction.

[0076] The tool wear prediction method based on part geometric features and cutting process parameters provided by the application firstly splits the complex profile part into simple part geometric features (surface, hole, groove), establishes the tool wear prediction regression model about geometric part feature parameters (surface: length, width, hole: diameter, depth, groove: length, width, height), tool parameters (diameter D, rake angle gamma0, relief angle alpha0, helix angle beta0) and cutting process parameters (spindle speed n, feed per tooth f z , axial depth of cut a p , etc.) for different part geometric features. According to the set parameter range, the orthogonal experiment is established, the tool wear under different parameters is measured, each group of tool wear and parameters are taken logarithm and converted into matrix form, input into MATLAB software, the coefficients and exponential parts of the tool wear prediction regression model are calculated by using the regress function, the coefficients and exponential parts are taken logarithm inverse function to obtain the tool wear prediction regression model under different features. Finally, the significance test and residual analysis are performed on the tool wear prediction regression model to obtain the final tool wear prediction model. The tool wear prediction model in the application involves many machining parameters, covers the main elements affecting tool wear in milling, is beneficial to obtain the tool wear under the interaction of multiple factors, has a wide application range (the method is suitable for different machining materials, different machining parameters and different machining shapes), and has higher tool wear prediction accuracy.

[0077] The technical scheme of the application is described in detail below by taking the milling of AL2024 material by hard alloy (WC) as an example.

[0078] Step one: simplify the complex profile part into surface, hole and groove features. Take the hole feature as an example, the tool wear prediction regression model is established as follows:

[0079]

[0080] Wherein:

[0081] H is the tool wear;

[0082] K, i, w, j, o, p, q and s are constants;

[0083] n is the spindle speed, f z is the feed per tooth, a p is the axial depth of cut, D is the diameter of the milling cutter, gamma0 is the rake angle of the milling cutter, alpha0 is the relief angle of the milling cutter, and beta0 is the helix angle of the milling cutter.

[0084] Step two: set the parameter range for hole features:

[0085] Spindle speed n (r / min): 4000-6000; feed per tooth f z (mm / r): 0.1-0.3; axial depth of cut a p (mm): 1-3; milling cutter diameter D (mm): 8-12; milling cutter rake angle γ0 (°): 5-15; milling cutter relief angle α0 (°): 10-18; milling cutter helix angle β0 (°): 30-40;

[0086] An orthogonal experiment table for milling AL2024 material with WC cutter was established, as shown in Table 1.

[0087] Step three: using DEFORM finite element software, a WC milling AL2024 model was established, as shown in Figure 2 , an orthogonal experiment was conducted, and the maximum wear of WC cutter under different parameters H was obtained, as shown in Table 1, a total of 18 experiments were conducted.

[0088] Table 1 Orthogonal simulation experiment data

[0089] Experiment No. n f z ]]> a p ]]> D [gamma]0 [alpha]0 [Alpha0] Experiment Result H (μm) Experiment 1 4000 0.1 1 8 5 10 30 0.0397 Experiment 2 4000 0.2 2 10 10 14 35 0.0722 Experiment 3 4000 0.3 3 12 15 18 40 0.0932 Experiment 4 5000 0.1 1 10 10 18 40 0.0562 Experiment 5 5000 0.2 2 12 15 10 30 0.0709 Experiment 6 5000 0.3 3 8 5 14 35 0.0692 Experiment 7 6000 0.1 2 8 15 14 40 0.0996 Experiment 8 6000 0.2 3 10 5 18 30 0.122 Experiment 9 6000 0.3 1 12 10 10 35 0.0322 Experiment 10 4000 0.1 3 12 10 14 30 0.131 Experiment 11 4000 0.2 1 8 15 18 35 0.0290 Experiment 12 4000 0.3 2 10 5 10 40 0.0499 Experiment 13 5000 0.1 2 12 5 18 35 0.0967 Experiment 14 5000 0.2 3 8 10 10 40 0.112 Experiment 15 5000 0.3 1 10 15 14 30 0.0195 Experiment 16 6000 0.1 3 10 15 10 35 0.213 Experiment 17 6000 0.2 1 12 5 14 40 0.0502 Experiment 18 6000 0.3 2 8 10 18 30 0.0459

[0090] Step four: taking the logarithm of the experimental results H and milling parameters in the table, then Y and X in the matrix converted by multiple linear regression are respectively:

[0091]

[0092]

[0093] Step five: through MATLAB software, M file is established, and the logarithmic values of each cutter wear and each parameter in Table 1 are input into the program for multiple linear regression calculation.

[0094] According to the results obtained by MATLAB, a0=-5.4286, a1=0.5317, a2=-0.6103, a3=1.0644, a4=0.5000, a5=-0.0012, a6=-0.0870, a7=0.7523.

[0095] Therefore, the multiple regression model is:

[0096] Y=-5.4286+0.5317x1-0.6103x2+1.0644x3+0.5000x4-0.0012x5-

[0097] 0.0870x6+0.7523x7

[0098] The coefficient in the above formula is calculated by the logarithmic inverse function, and the tool wear prediction regression model corresponding to the hole characteristics is obtained as follows:

[0099]

[0100] Step six: The significance test is performed on the tool wear prediction regression model, and the results obtained by MATLAB are as follows:

[0101] The tool wear regression coefficient is:

[0102] a0=-5.4286, the confidence interval of a0 is (-7.4865, -3.3707);

[0103] a1=0.5317, the confidence interval of a1 is (0.0734, 0.9899);

[0104] a2=-0.6103, the confidence interval of a2 is (-0.7778, -0.4428);

[0105] a3=1.0644, the confidence interval of a3 is (0.8969, 1.2320);

[0106] a4=0.5000, the confidence interval of a4 is (0.0418, 0.9582);

[0107] a5=-0.0012, the confidence interval of a5 is (-0.1687, 0.1664);

[0108] a6=-0.0870, the confidence interval of a6 is (-0.4025, 0.2285);

[0109] a7=0.7523, the confidence interval of a7 is (0.1062, 1.3984).

[0110] The statistical variable stats is obtained: R 2 =0.9662, F=40.8633, P=0.0000, wherein, the goodness of fit R 2 is expressed as the square of the correlation coefficient, which is generally in the range of 0.8-1, and can be used to judge the linear correlation between the regression independent variable and the dependent variable, and the value of R 2 is 0.9662, which is close to 1, indicating that the linear correlation is strong. F represents the hypothesis test statistic, and the probability P corresponding to F is P<0.0001, which obviously satisfies P<=0.05. The results deduced by the above three statistical variables are consistent, indicating that there is a significant linear relationship between the tool wear and the experimental parameters, and the obtained linear regression model is accurate.

[0111] The residual analysis diagram is obtained by using MATLAB to analyze the coefficient, as shown in the following figure Figure 3 As can be seen from the residual analysis diagram, the residual of the data is close to zero, and the confidence interval of the residual contains zero. This shows that the regression model can better meet the original data, and it is believed that the regression model is reliable.

[0112] Therefore, the final tool wear prediction model of the cemented carbide tool machining AL2024 hole features in the set parameter range is:

[0113]

[0114] For other materials, other features, other processing methods, and other processing parameters, this method can be used for prediction.

[0115] The technical features of the above-described embodiments can be combined arbitrarily, and to make the description simple, all possible combinations of the technical features in the above-described embodiments are not described, however, as long as the combinations of the technical features do not contradict, they should be considered as the scope of the description.

[0116] The above-described embodiments only express several embodiments of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the scope of the patent. It should be pointed out that for ordinary skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are all within the scope of the present application. Therefore, the protection scope of the patent of the present application should be subject to the appended claims.

Claims

1. A tool wear prediction method based on part geometry and cutting process parameters, characterized in that, The method comprises the following steps: Step one: split the complex profile part into simple part geometric features, the part geometric features are face features, hole features or groove features, and establish tool wear prediction regression models for different part geometric features as follows: (1) for face features: Wherein: H is the tool wear amount; K, i, w, x, o, p, q, s, t are constants; n is the spindle speed, f z is the feed per tooth, a e is the radial depth of cut, D is the diameter of the milling cutter, γ0is the milling cutter rake angle, α0is the milling cutter relief angle, β0is the milling cutter helix angle, and l is the milling cutter travel; (2) for hole features: Wherein: J is a constant; a p is the axial depth of cut; (3) for groove features: Step two: set parameter ranges according to different part geometric features, and establish corresponding orthogonal experiment tables; Step three: perform orthogonal experiments according to the orthogonal experiment tables to obtain tool wear amounts under each group of parameters; Step four: take the logarithm of both sides of the tool wear prediction regression model formula corresponding to the part geometric features, establish a multiple linear regression equation, and convert the multiple linear regression equation into a matrix form; Step five: establish an M file through MATLAB software, input the logarithmic values of each tool wear amount and the corresponding parameters obtained through the orthogonal experiments into the MATLAB software, perform multiple linear regression calculation by using the regress function, obtain the value of the regression coefficient, perform logarithmic inverse function calculation on the value of the regression coefficient, obtain each constant in the tool wear prediction regression model formula corresponding to the feature, and further obtain the tool wear prediction regression model corresponding to the feature; Step six: perform significance test and residual analysis on the tool wear prediction regression model to obtain the final tool wear amount prediction model.

Images