1 Introduction Milling is a widely used cutting method. The testing and research of milling forces can provide scientific theoretical basis and reasonable processing parameters for the development and application of new materials, new tools and new processes. However, the milling force test data processing process is cumbersome, the calculation volume is large, manual data processing is easy to make mistakes, and the empirical formula obtained is not accurate. With the continuous improvement of the main frequency and memory capacity of IPCs, the development of milling force test data processing software that is suitable for on-site use and high degree of automation has become an inevitable requirement for the research on milling force testing. For this reason, we chose Visual C++ 6.0 as the programming tool, designed the experimental program using the orthogonal design principle, and developed the data processing software of milling force test based on Windows 98. The software uses multivariate regression analysis to process the test data. The coefficient and index reflecting the relationship between milling force and process parameters can be calculated in real time in the empirical formula of multi-factor milling force, and the significance test of the regression equation can be performed on the test site. And missed test. 2 Milling Force Test Data Processing Principles and Steps The establishment of regression equation is known from metal cutting principle. The empirical formula of milling force and milling amount is Fe/z=CFeaek1 d0k2 fzk3 apk4 (1) where: Fe—milling force is in the horizontal direction. Component force in the direction (N) z - number of teeth of the cutter CFe - coefficient ae of the Fe milling force formula - milling width (mm) d0 - diameter of the cutter (mm) fz - feed per tooth (mm/z) ap - Milling depth (mm) k1, k2, k3, k4 - parameters to be determined Logarithmic on both sides of equation (1), available ln(Fe/z)=lnCFe+k1lnae+k2lnd0+k3lnfz +k4lnap Let y=ln(Fe/z), k=lnCFe, z1=lnae, z2=lnd0, z3=lnfz, z4=lnap. The available equation y=k+k1z1+k2z2+k3z3+k4z4 This equation is a linear equation for y for z1, z2, z3, and z4, which can be processed using multiple linear regression analysis. If N tests are performed, the data of the i-th test is yi, z1i, z2i, z3i, z4i (i=1, 2,..., N). Let k^, k^1, k^2, k^3, and k^4 be the least-square estimates of the parameters k, k1, k2, k3, and k4, respectively, then the regression equation of the above equation is y^=k^+ k^1 z1+k^2z2+k^3z3+k^4z4 (2) Determination and coding of each factor level The range of variation of the above four factors is aemin to aemax, d0min to d0max, fzmax to fzmin, apmin to apmax, respectively. With linear transformation, different dimensionless codes can be used to replace different levels of factors to simplify calculations. Let the upper and lower levels of the factor be zjmax, zjmin; the zero level is zj0=(zjmax+zjmin)/2; the change interval is ∆j=(zjmax-zjmin)/2, then the corresponding dimensionless code is xj=(zj-zj0 )/∆j. In this way, a one-to-one relationship between the factor zj and the coded value xj is established. Therefore, the regression problem of y on z1, z2, z3, and z4 is converted into the regression problem of y on x1, x2, x3, and x4. The formula (2) is y^=b0+b1 x1+b2 x2+b3 x3+b4 x4 after the coding is finished. (3) The regression coefficient is calculated by using the least square method to determine the coefficients b0, b1, b2 in equation (3). B3, b4. First, let the formula Q=∑a(ya-y^a)2=∑a(ya-b0-b1xa1-b2xa2-b3xa3-b4xa4)2 be the smallest, and then calculate one for b0, b1, b2, b3, and b4, respectively. The derivative of the partial order and making it equal to zero, a linear equation group can be obtained, and the solution can be obtained by solving the regression coefficient. To reduce the number of trials, use the L8 (27) orthogonal table. According to the principle of orthogonality, except for the diagonal elements in the coefficient matrix of the above linear equation, all other elements should be zero. Therefore, the regression coefficient is b0=∑aya/N=B0/N
Bj=∑axajya/∑aajx2=Bj/dj Thus, regression equation (3) can be solved. Then substituting the xj value into (3) and taking the antilogarithm, the required mathematical model can be obtained. The test of the regression equation is due to the approximation algorithm used in the linear regression model, and there is an approximation error. Therefore, after the linear regression equation is obtained, it needs to be statistically tested. The software can make significant test and unrealistic test of the regression equation. 3 Software Design
Figure 1 Object Model Diagram
Object model milling force test data processing software object model shown in Figure 1. Software Function Design and Implementation The milling force test data processing software system is mainly composed of the following subsystems with specific functions: File Operation Subsystem The file operation subsystem has four functions for creating, opening, saving, and deleting files. Because the software system needs to save more data, for the convenience of management, it draws on the way of creating project files in VC++, and creates a project folder for each new test project. All the data related to this experiment are stored by type. In this folder. When it is necessary to open a previously created experimental project, simply click on the file with the same name under the test project folder in the open dialog box. This file is a virtual file and only serves as a data save path. Actual data Both are saved under the data folder in the project folder. The contents of the files stored in the data folder include the names, quantities, and processing sequence numbers of the test materials; the coefficients and indices in the empirical formulas for the various materials; the level values ​​of the various factors and the number of teeth for each tool; each material is processed Need to save the data and so on. The three functions of creating, opening, and saving files are implemented by the functions OnNew(), OnOpen(), and OnSave(), respectively. Test Conditions Subsystem Test Conditions The main functions of the subsystem are: 1 Enter the name of the test piece and the upper and lower levels of each factor. After the check is legal, the zero level will be automatically generated and saved; 2 The milling cutters corresponding to the 3 milling cutters used for the input and saving test are saved. Number of teeth; 3 Add, delete, and save the names of materials participating in the experiment at any time. The above functions are implemented by the CTestCondition class. In order to enter each factor level value in a graphical control, a subclass CtestRequire is inherited from the CStatic class. By clicking the left mouse button, the self-made grid control code is added to the message response function OnLButtonDown() of the inherited class, an edit box can be dynamically generated at the input and edit record position, and is automatically destroyed after the user input is completed. Box and save the data. The specific implementation method is to read the coordinates of the click position of the mouse in the graphic control, and use the “if... else†statement to realize different coordinates corresponding to different condition codes, and the data input in different positions can be transmitted to the corresponding data storage. In the unit, the self-made grid control can guarantee the stability and security of the software. The system adds or deletes test piece material names by using left and right migration methods. When the user clicks the "Move in > >" button, the corresponding message response function code can obtain the string in the candidate material combination box and add it to the sample material list box. When you click the "Exit <<" button, the selected item can be deleted from the specimen material list box. The entire operation is completed by clicking on the BNCLICKED message that was activated when the "Shift in> >" or "Shift out <" key was clicked. The transfer data is implemented by the string list member CStringListmaterial of the structure MaterialProperty. Trial design subsystem The trial design subsystem constitutes the main window interface of the software, distributing all function buttons, test piece material combo boxes and milling force component combo boxes, and make the subsystems work together. At the same time, it is also a window carrier for displaying the test plan, loading the collected milling force data, and displaying the regression equation coefficients. The subsystem function is implemented by the main window class CMillingDlg. Each button corresponds to a message response function, including new OnNew(), OnOpen(), OnSave(), OnPrint(), OnQuit(), OnCallData(), OnDataHand(), Empirical Formula OnExperienceFormula(), significance test OnSignificanceLevel() and OnMatchCapability(). Data display in the main window class is borne by the test design class (CTestDesign). It inherits from the static class CStatic and has the following functions: 1 Pass the name of the test piece entered in the test condition class (CTestCondition) and the level of each factor to this class. And display; 2 select the test material, transfer the corresponding test data and display; 3 in the manual input data mode can easily edit the force value data, check the legitimacy of the input data; 4 pass the coefficients of the regression equation obtained by the data processing Go to this class and show. Since in CTestDesign you need to reference the member array level[12] of CMillingDlg, which is the parent of CTestDesign, you need to get the parent pointer in this class with the code (CMillingDlg*)pMillingDlg=(CMillingDlg*)(this -> GetParent()), where pMillingDlg is the parent class pointer to be obtained, so using the code pMillingDlg->level[i] can easily access any of these elements. This is an efficient way to transfer data between classes. This software enables the sharing of relevant information such as the name of the input test piece. After inputting each factor level value in the test condition dialog box and exiting, the originally displayed code value (1, -1, 0) in the main window is replaced and displayed by the actual value, so it is convenient for data processing at the test site. . Data processing subsystem data processing subsystem uses orthogonal regression algorithm to process milling force test data and has the following functions: 1 Transfer (or manual input) test data from the file and verify the validity; 2 Use regression analysis method to calculate milling force data Processing, to obtain the regression equation coefficient; 3 on the regression equation significant test; 4 on the regression equation for the loss of quasi-test; 5 Solve various experimental materials empirical formula. The material property class CMaterialProperty is defined, each test material is instantiated as an object of this class, and intermediate results and final results obtained after the transferred test data are processed (including the coefficient bj of the regression equation and the empirical formula index ki The level of significance, the sum of squared deviations, the sum of the squared regression deviations, the sum of the squared residual deviations, the F value of each factor, the F value of the regression equation, the t-value of the regression test, etc.) are registered in this object until the system needs it. When called. Serialized storage provided by MFC saves data from memory variables to the hard disk. Overload your own Serialize () function in the class, when you need to store or read data, you can call the Serialize () function directly in the program; data processing and inspection is completed by the function OnDataHandle (); the regression equation to solve the empirical formula The index is the inverse process of transforming empirical formulas into linear regression equations. It is implemented by the function OnExperienceFormula(); the generation and display of the empirical formulas are implemented by the CExperienceExpress class; the significance test and the results of the miscalculation tests are generated and displayed by the CSignificanceLevel class and the CMatchCapability class. In order to adapt to changes in the amount and type of materials, the CTableCtrl class is used, which can be viewed by clicking on the tab.
Figure 2 Milling force data acquisition and analysis system block diagram
Table 1 Factor Level Indicator Level Level Horizontal Level Zero Level ae 70.00 50.00 59.16 d0 125 80 100 fz 0.032 0.02 0.025 ap 3 1 1.732 Print Milling Force Test Report Subsystem Print Mill Force Test Report Subsystems are derived from the CReportView of the CScrollView class. Class implementation. Using the ScrollView function of the CScrollView class and the printing program, you can achieve flicker-free visual effects when editing and display switching, and a WYSIWYG milling force test report printing method. The concrete implementation method is to add code in the OnDraw() function of the CReportView class, draw the entire report on its device environment, and add the grid control code in the OnLButtonDown() function, so that the user can click, enter and edit in the corresponding position. Print the data required in the Mill Force Test Report dialog box. The print program can call the OnDraw() function code itself, except that it is output in the printer, not on the screen. The test report is composed of multiple pages. Each page is provided with a separate starting point coordinate variable. It is possible to set the print content of each page and the layout position on this page as required. 4 Application example In order to study the milling performance of 40MnV non-quenched and tempered steel, the orthogonal milling method was used to test the milling force when machining 40MnV. The milling force test data was collected by the industrial PC control acquisition card. The software was used to process the test data, and an empirical formula was established between the milling force and the milling amount when machining 40 MnV. Milling force data acquisition and analysis system consists of X53K vertical milling machine, HT-3B1 strain milling dynamometer, YD-28 dynamic resistance strain gauge, ACL-8112PG data acquisition card and PCA-6155 industrial computer, as shown in Figure 2. . The level of each factor determined according to the commonly used range of experimental factors and actual conditions is shown in Table 1. The milling test was arranged according to the orthogonal table L8 (27). The test conditions, test results and data processing results are shown in Table 2.
Table 2 Milling force test data and calculation table in transverse feed direction Factor x0 x1 (ae) x2 (d0) x3 (fz) x4 (ap) Fe (N) 1 1 70.000 125.000 0.032 3.000 347.9 2 1 70.000 125.000 0.020 1.000 72.1 3 1 70.000 80.000 0.032 1.000 168.6 4 1 70.000 80.000 0.020 3.000 338.7 5 1 50.000 125.000 0.032 1.000 92.0 6 1 50.000 125.000 0.020 3.000 174.3 7 1 50.000 80.000 0.032 3.000 450.5 8 1 50.000 80.000 0.020 1.000 88.4 9 1 59.161 100.000 0.025 1.732 130.8 10 1 59.161 100.000 0.025 1.732 150.2 11 1 59.161 100.000 0.025 1.732 150.8 bj 3.3779 0.1010 -0.2165 0.2332 0.5674 - From Table 2, the regression equation is y^=3.3779+0.1010x1-0.2165x2+0.2332x3+0.5674x4 where: x1= (lnae-lnae0)/∆aae=5.944lnae-24.253x2=(lnd0-lnd00)/∆d0=4.482lnd0-20.638x3=(lnfz-lnfz0)/∆fz=4.255lnfz-15.697x4=(lnap-lnap0) /∆ap=1.821lnap-1 Substituting x1~x4 into the above equation, y^=ln(Fe/z)=8.47+0.60lnae-0.97lnd0+0.99lnfz+1.03lnap takes an inverse number to the equation above, ie, 40MnV is processed. Non-quenched and tempered steel in the direction of the direction of the milling force of the empirical formula for the Fe =4789.31ae0.60d0-0.97fz0.99 a P1.03z 5 Conclusion Based on the Windows 98 milling force test data processing software can be orthogonal regression method for the study of milling forces, the entire test process without manual recording and calculation of data, to achieve real-time test data and direct output complete test report The goal. Through the milling force test of the as-cast aluminum alloy materials ZL106, DFL-1 and 40MnV non-quenched and tempered steel, it is proved that the software has a friendly interface and is easy to use. It can quickly and accurately derive the empirical formula of the milling force of the test material. The significant influence of factors and the merits of the regression equation are of great practical value.
Small Gift
Imitation bird Feather Bird Ornament color bird garden animal sparrow swallow pigeon hot sale home decoration ornament.Simulation bird feather feather bird binding bird model Taoism small animal teaching aids, pastoral style decoration, fake bird.Wedding bird lovers bird love bird lover gift simulation bird Wedding Dress Hat bird feather bird home decoration.
Small Gift,Easter Butterfly Crafts,Carnival Gifts,Christmas Presents
Xintong Craft Gift Co.,LTD , https://www.jlxtcraft.com