Titanium alloy component surface integrity milling parameter optimization method

Abstract

The invention provides a titanium alloy component surface integrity milling parameter optimization method. The titanium alloy component surface integrity milling parameter optimization method comprises the steps that a mapping relation model between surface state characteristics and milling parameters is established; sensitivity analysis is conducted on the established mapping relation model between the surface state characteristics and the milling parameters, and a surface state characteristic stable process parameter domain omega1 is determined; selecting a parameter processing component in the surface state characteristic stable process parameter domain omega1, performing a fatigue life test, and establishing a component fatigue life prediction model based on the surface state characteristics; establishing a surface state feature constraint condition by taking the fatigue life as a criterion, and obtaining a surface integrity process parameter domain omega2; and with high efficiency as a target, process parameter domain optimization is carried out in the surface integrity process parameter domain omega2, and a surface integrity efficient precision milling process parameter domain omega3 is obtained.

Description

Technical Field

[0001] The present application belongs to the technical field of surface integrity efficient precision milling parameter optimization, and specifically relates to a method for optimizing milling parameters for the surface integrity of a titanium alloy component. Background Art

[0002] As a high-strength alloy, titanium alloy has excellent properties such as light weight, high strength, high temperature resistance and fatigue resistance. It is an important high-strength alloy material for key components in the manufacture of major equipment in the aerospace field. The content of titanium alloy in commercial aircraft engines accounts for 30% and 40% of their weight respectively.

[0003] Titanium alloy, as a typical high-strength alloy for aviation, is a typical difficult-to-process material. The low thermal conductivity hinders the rapid consumption of heat during the cutting process. The heat is concentrated in the cutting deformation zone and a small range near the cutting edge. The cutting force per unit area is large and the cutting temperature is high, which can easily cause tool collapse and edge chipping, and easily cause surface morphology defects and surface tissue damage, destroying the surface integrity of the component, thereby seriously affecting its fatigue performance.

[0004] Milling processing must not only ensure that the surface roughness and shape and position accuracy of the component meet the requirements of the process design drawings, but also ensure that the surface state characteristics of the component processing meet the service performance requirements.

[0005] The existing milling parameter optimization methods only consider machining efficiency, dimensional accuracy and position accuracy, but do not consider the integrity of the machined surface. Therefore, establishing a high-efficiency milling process parameter optimization model and solution method for surface integrity, and proposing a high-efficiency precision milling parameter optimization scheme for surface integrity based on fatigue as the criterion and surface state characteristics as the constraint is the guarantee for achieving the long life and high reliability requirements of aviation components. Summary of the invention

[0006] Purpose of the invention: To provide a method for optimizing and controlling the parameters of efficient and precise milling of titanium alloy components to improve the surface integrity, so as to solve the contradiction among the machining efficiency, surface state characteristics and fatigue life of titanium alloy components during machining.

[0007] The present application provides a method for optimizing milling parameters of a titanium alloy component surface integrity, the method comprising the following steps:

[0008] Step 1: Determine the initial process parameter domain Ω0 based on research and literature review, design a milling process test within the initial process parameter domain Ω0, test the surface state characteristics of the specimen after processing, summarize, statistically analyze, and nonlinearly fit all test results, and establish a mapping relationship model between surface state characteristics and milling parameters;

[0009] Step 2: Conduct sensitivity analysis on the established mapping relationship model between surface state characteristics and milling parameters to determine the stable process parameter domain Ω1 of surface state characteristics;

[0010] Step 3: Select parameter processing components in the surface state characteristic stable process parameter domain Ω1, and conduct fatigue life tests to establish a component fatigue life prediction model based on surface state characteristics; take fatigue life as the criterion, establish surface state characteristic constraint conditions, and obtain the surface integrity process parameter domain Ω2;

[0011] Step 4: With high efficiency as the goal, the process parameter domain is optimized within the surface integrity process parameter domain Ω2 to obtain the surface integrity high-efficiency precision milling process parameter domain Ω3.

[0012] Preferably, the initial process parameter domain Ω0 includes v c ,f z ,a p ,a e .

[0013] Preferably, the surface state characteristics include surface stress concentration factor, surface residual stress, and surface microhardness, and the specific characterization symbols are: K stx Represents the surface stress concentration factor of the specimen along the feeding direction, K sty Represents the surface stress concentration factor of the specimen along the cutting width direction, σ rx represents the surface residual stress of the specimen along the feeding direction, σ rx represents the surface residual stress of the specimen along the cutting width direction, and HV represents the surface hardness of the specimen.

[0014] Preferably, the surface state characteristic constraint conditions include: surface state characteristic stable process parameter domain constraint conditions, fatigue life constraint conditions, surface stress concentration constraint conditions, surface residual stress constraint conditions, surface microhardness constraint conditions and damage layer depth constraint conditions.

[0015] Preferably, the step 1 comprises:

[0016] Step 1.1 Determine the initial process parameter domain Ω0 for titanium alloy component processing;

[0017] Step 1.2: In the initial process parameter domain Ω0, the full factorial response surface method is used to perform the experimental design;

[0018] Step 1.3 Perform a milling test according to the response surface test design in step 1.2 to test the surface state characteristic parameters of the test piece processed by each set of process parameters;

[0019] Step 1.4 establishes a relationship model between the surface state characteristics of the specimen and the process parameters according to the process parameters and the test data of the surface state characteristic parameters of each set of parameters;

[0020]

[0021] In the formula, SS is the surface state characteristic, a i (i=1,2,3,4,5) is the fitting coefficient.

[0022] Preferably, in step 2, a sensitivity analysis is performed on the established mapping relationship model between the surface state characteristics and the milling parameters according to formula (2);

[0023] The input variables x1, ..., x p , …, x n x in p Divided into m levels M1,…,M q , …, M m , according to the sensitivity curve, for the input variable x p Calculate the sensitivity values ​​in [M1,M2], [M2,M3], …, [M m-1 ,M m ] The amplitude of change in the m-1 horizontal intervals is recorded as A1, A2, ..., A m-1 ;

[0024] Calculate the m-1 sensitivity value change amplitudes A1, ..., A j , …, A m-1 The average value of is A0;

[0025] If A j >A0, then the interval is an unstable domain interval;

[0026] If A j <A0, then the interval is a stable domain interval;

[0027] Obtain the stable process parameter domain Ω1 of the input variables: [x 1min ,x 1max ]、[x 2min ,x 2max ], …, [x nmin ,x nmax ];

[0028]

[0029] In the formula, n is the number of variables.

[0030] Preferably, in step 3, a component is processed by selecting parameters in the surface state characteristic stable process parameter domain Ω1, and a fatigue life test is performed to establish a component fatigue life prediction model based on surface state characteristics, as shown in formula (3):

[0031]

[0032] Where N f is fatigue life, N f0 is the fatigue life of the matrix material, K st is the stress concentration factor, HV is the microhardness value, HV0 is the matrix microhardness value, σ r is the residual stress value / MPa, σ b is the tensile strength of the material / MPa, a i , i=0,1,2,3 are model coefficients;

[0033] When the stress concentration factor is 1, the microhardness value is the microhardness of the matrix, and the residual stress is equal to the tensile strength of the material, N f Equal to N f0 ;

[0034] Taking fatigue life as the criterion, the constraint conditions of surface state characteristics are established, and the surface integrity process parameter domain Ω2 is obtained by solving; among them, the constraint conditions of process parameters and surface state characteristics, as well as the fatigue life criterion conditions are as follows:

[0035] ① Establish the constraints of the stable process parameter domain of surface state characteristics:

[0036] v cmin ≤v c ≤v cmax (4)

[0037] f zmin ≤f z ≤f zmax (5)

[0038] a pmin ≤a p ≤a pmax (6)

[0039] a emin ≤a e ≤a emax (7)

[0040] Where: v cmax is the maximum value of milling speed, v cmin is the minimum value of milling speed, f zmax is the maximum feed per tooth, f zmin is the minimum feed per tooth, a pmax is the maximum cutting depth, a pmin is the minimum cutting depth, a emax is the maximum value of the cutting width, a emin is the minimum value of cutting width;

[0041] ② Fatigue life constraints:

[0042] N f ≥ N f0 (8)

[0043]

[0044] Where N f is fatigue life, N f0 is the fatigue life of the matrix material, K st is the stress concentration factor, HV is the microhardness value, HV0 is the matrix microhardness value, σ r is the residual stress value / MPa, σ b is the tensile strength of the material / MPa, a i (i=0,1,2,3) is the model coefficient; when the stress concentration factor is 1, the microhardness value is the microhardness of the matrix, and the residual stress is equal to the tensile strength of the material, N f Equal to N f0 ;

[0045] ③Surface stress concentration constraint conditions:

[0046] K stmin ≤K st (v c ,f z ,a p ,a e )≤K stmax (10)

[0047] Where: K st (v c ,f z ,a p ,a e ) is the established mapping relationship model between milling parameters and surface stress concentration, K stmin is the minimum value of the surface stress concentration factor, K stmax is the maximum value of the surface stress concentration factor;

[0048] ④Surface residual stress constraint conditions:

[0049] σ rmin ≤σ r (v c ,f z ,a p ,a e )≤σ rmax (11)

[0050] Where: r (v c ,f z ,a p ,a e) is the established mapping relationship model between milling parameters and surface residual stress, σ rmin is the minimum allowable value of surface residual stress or the maximum allowable value of residual compressive stress, σ rmax is the maximum allowable value of surface residual stress or the minimum allowable value of residual compressive stress;

[0051] ⑤Surface microhardness constraints:

[0052] HV min ≤HV(v c ,f z ,a p ,a e )≤HV max (12)

[0053] Where: HV(v c ,f z ,a p ,a e ) is the established mapping relationship model between milling parameters and surface microhardness, HV min HV is the minimum allowable value of surface microhardness. max is the maximum allowable value of surface microhardness;

[0054] ⑥ Damage layer depth constraint conditions:

[0055] According to the residual stress cosine attenuation function prediction model, as shown in formula (13), when the residual stress value at depth h is less than or equal to the matrix residual stress, the depth h at this time is considered to be the residual stress influence layer depth value, which is numerically equal to the damage layer depth D; the inverse function calculation method is used to calculate the residual stress influence layer depth h, that is, the damage layer depth D is obtained, as shown in formula (14);

[0056] Ae -λh cos(wh+θ)+σ0≤σ0 (13)

[0057] Let σ rsb =Ae -λh cos(wh+θ)≤0, then:

[0058] h=f -1 (σ rsb ) (14)

[0059] Where: A, λ, ω, θ are the model control factors in the residual stress prediction model. The values ​​of each model control factor under given milling parameters can be obtained according to the established relationship model between milling parameters and model control factors.

[0060] D1 p0 (15)

[0061] ​Where: D1 is the depth of the damaged layer caused by fine milling, a p0 The depth of material removal during the finishing process.

[0062] Preferably, in step 4, an efficient objective function as shown in formula (16) is established, and the process parameter domain is optimized within the surface integrity process parameter domain Ω2 by using the MATLAB mathematical optimization tool to obtain the surface integrity efficient precision milling process parameter domain Ω3:

[0063] maxQ=f(v c ,f z ,a p ,a e ) (16).

[0064] This application has the following technical effects:

[0065] The present invention aims at the fine milling process. By designing a response surface process test within the domain of initial process parameters, a relationship model between titanium alloy milling process parameters and surface state characteristics is established, and the stable process parameter domain of surface state characteristics is determined; through fatigue life tests, a blade fatigue life prediction model that relies on stress concentration and considers strain hardening and stress sensitivity is proposed; finally, with high efficiency as the goal and fatigue life as the criterion, process parameter optimization is completed within the stable process parameter domain of surface state characteristics. In order to verify the effect of the present invention, TC17 titanium alloy blades are processed using optimized parameters and conventional parameters, and the surface state characteristics and fatigue tests of the blades are performed. Compared with the conventional parameters, the processing efficiency of the blades processed by the parameters obtained by the titanium alloy component surface integrity efficient precision milling parameter optimization method is increased by 4.9%, and the surface stress concentration coefficient of the processed blades is reduced, the residual compressive stress and microhardness are high, the fatigue life is increased by 45%, and the fatigue resistance is significantly improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 This is the absolute sensitivity analysis diagram of surface state characteristics to cutting speed;

[0067] Figure 2 This is the absolute sensitivity analysis diagram of surface condition characteristics to feed per tooth;

[0068] Figure 3 This is the absolute sensitivity analysis diagram of surface state characteristics to cutting depth;

[0069] Figure 4 This is the absolute sensitivity analysis diagram of surface state characteristics to cutting width;

[0070] Figure 5 Iteration curve graph for genetic optimization algorithm;

[0071] Figure 6This is a comparison diagram of the gradient distribution of blade residual stress along the depth direction;

[0072] Figure 7 This is a gradient comparison diagram of the blade microhardness along the depth direction;

[0073] Figure 8 This is a comparison chart of blade fatigue life. DETAILED DESCRIPTION

[0074] See also Figure 1-Figure 8 The present invention adopts the following technical solution: a method for optimizing and controlling the parameters of efficient and precise milling of titanium alloy component surface integrity, which specifically comprises the following steps:

[0075] Step 1: Determine the initial process parameter domain Ω0(v based on research and literature review c ,f z ,a p ,a e ), design the milling process test within the initial process parameter domain Ω0; test the surface state characteristics of the specimen after processing, summarize, statistically analyze, and nonlinearly fit all the test results, and establish a mapping relationship model between the surface state characteristics and the milling parameters. Surface state characteristics (including surface stress concentration factor, surface residual stress, surface microhardness, the specific characterization symbol is: K stx Represents the surface stress concentration factor of the specimen along the feeding direction, K sty Represents the surface stress concentration factor of the specimen along the cutting width direction, σ rx represents the surface residual stress of the specimen along the feeding direction, σ rx represents the surface residual stress of the specimen along the cutting width direction, and HV represents the surface hardness of the specimen.

[0076] Step 2: Conduct sensitivity analysis on the established mapping relationship model between surface state characteristics and milling parameters to determine the stable process parameter domain Ω1 of surface state characteristics.

[0077] Step 3: Select parameter processing components in the surface state characteristic stable process parameter domain Ω1, and conduct fatigue life tests to establish a component fatigue life prediction model based on surface state characteristics; using fatigue life as the criterion, establish surface state characteristic constraint conditions to obtain the surface integrity process parameter domain Ω2.

[0078] Step 4: With high efficiency as the goal, the process parameter domain is optimized within the surface integrity process parameter domain Ω2 to obtain the surface integrity high-efficiency precision milling process parameter domain Ω3.

[0079] Furthermore, the relationship model between the surface state characteristics and the process parameters to be established in step 1 is shown in formula (1):

[0080]

[0081] In the formula, SS is the surface state characteristic, a i (i=1,2,3,4,5) is the fitting coefficient.

[0082] Furthermore, the specific method of step 1 includes the following steps:

[0083] Step 1.1 Determine the initial process parameter domain Ω0 for titanium alloy component processing;

[0084] Step 1.2: In the initial process parameter domain Ω0, the full factorial response surface method is used to perform the experimental design;

[0085] Step 1.3 Perform milling test according to response surface experimental design, and calculate the surface stress concentration factor K of the test piece after each set of parameters processing. st , residual stress σ r , test and calculate the microhardness HV;

[0086] Step 1.4: According to the process parameters and the surface state characteristic parameters of the specimen processed by each set of parameters, a relationship model between the specimen surface stress concentration factor, surface residual stress, surface microhardness and process parameters is established:

[0087]

[0088] In the formula, K stx is the surface stress concentration factor of the specimen along the feeding direction, K sty is the surface stress concentration factor of the specimen along the cutting width direction, σ rx is the surface residual stress of the specimen along the feeding direction, σ rx is the surface residual stress of the specimen along the cutting width direction, HV is the surface hardness of the specimen; a ij (i,j=1,2,3,4,5) are the fitting coefficients.

[0089] Furthermore, the specific method for determining the surface state characteristic stabilization process parameter domain Ω1 in step 2 is as follows:

[0090] The established mapping relationship model between surface state characteristics and milling parameters is subjected to sensitivity analysis according to the following formula.

[0091]

[0092] In the formula, n is the number of variables.

[0093] The method for determining the stable process parameter domain Ω1 is as follows: the input variables (x1, ..., x p , …, x n ) pDivided into m levels (M1, ..., M q , …, M m ), according to the sensitivity curve, the input variable x p Calculate the sensitivity values ​​in [M1,M2], [M2,M3], …, [M m-1 ,M m ] and other m-1 horizontal intervals, denoted as A1, A2, ..., A m-1 ; Calculate the m-1 sensitivity value change amplitudes A1, ..., A j , …, A m-1 The average value of is A0; if A j >A0, then the interval is an unstable domain interval; if A j <A0, then the interval is a stable domain interval; obtain the stable process parameter domain Ω1 of the input variable: [x 1min ,x 1max ]、[x 2min ,x 2max ], …, [x imin ,x imax ].

[0094] Furthermore, the specific method for determining the surface integrity process parameter domain Ω2 in step 3 is as follows:

[0095] Step 3.1 Select parameter processing components in the surface state characteristic stable process parameter domain Ω1, select one specimen in each group of tests to perform surface state test, and other specimens to perform fatigue life test;

[0096] Step 3.2 Based on the surface state test results and fatigue life test result data, the relationship is established based on the proposed blade fatigue life prediction model that relies on stress concentration and considers strain hardening and stress sensitivity:

[0097]

[0098] Where N f is fatigue life, N f0 is the fatigue life of the matrix material, K st is the stress concentration factor, HV is the microhardness value, HV0 is the matrix microhardness value, σ r is the residual stress value / MPa, σ b is the tensile strength of the material / MPa, a i (i=0,1,2,3) is the model coefficient. When the stress concentration factor is 1, the microhardness value is the microhardness of the matrix, and the residual stress is equal to the tensile strength of the material, N f Equal to N f0 .

[0099] Step 3.3: Solve the surface integrity process parameter domain Ω2 and establish the condition based on fatigue life:

[0100] N f ≥ N f0 (5)

[0101]

[0102] Where N f is fatigue life, N f0 is the fatigue life of the matrix material, K st is the stress concentration factor, HV is the microhardness value, HV0 is the matrix microhardness value, σ r is the residual stress value / MPa, σ b is the tensile strength of the material / MPa, a i (i=0,1,2,3) is the model coefficient. When the stress concentration factor is 1, the microhardness value is the microhardness of the matrix, and the residual stress is equal to the tensile strength of the material, N f Equal to N f0 .

[0103] Step 3.4 uses fatigue life as the criterion to solve and obtain the surface integrity process parameter domain Ω2. It is also necessary to establish the following surface state characteristic constraints.

[0104] ① Establish the surface state characteristic stable process parameter domain constraint conditions as follows:

[0105] v cmin ≤v c ≤v cmax (7)

[0106] f zmin ≤f z ≤f zmax (8)

[0107] a pmin ≤a p ≤a pmax (9)

[0108] a emin ≤a e ≤a emax (10)

[0109] Where: v cmax is the maximum value of milling speed, v cmin is the minimum value of milling speed, f zmax is the maximum feed per tooth, f zmin is the minimum feed per tooth, a pmax is the maximum cutting depth, a pmin is the minimum cutting depth, aemax is the maximum value of the cutting width, a emin is the minimum cutting width.

[0110] ②The surface stress concentration constraint condition is:

[0111] K stmin ≤K st (v c ,f z ,a p ,a e )≤K stmax (11)

[0112] Where: K st (v c ,f z ,a p ,a e ) is the established mapping relationship model between milling parameters and surface stress concentration, K stmin is the minimum value of the surface stress concentration factor, K stmax is the maximum value of the surface stress concentration factor.

[0113] ③The surface residual stress constraint condition is:

[0114] σ rmin ≤σ r (v c ,f z ,a p ,a e )≤σ rmax (12)

[0115] Where: r (v c ,f z ,a p ,a e ) is the established mapping relationship model between milling parameters and surface residual stress, σ rmin is the minimum allowable value of surface residual stress or the maximum allowable value of residual compressive stress, σ rmax It is the maximum allowable value of surface residual stress or the minimum allowable value of residual compressive stress.

[0116] ④The surface microhardness constraint conditions are:

[0117] HV min ≤HV(v c ,f z ,a p ,a e )≤HV max (13)

[0118] Where: HV(v c,f z ,a p ,a e ) is the established mapping relationship model between milling parameters and surface microhardness, HV min HV is the minimum allowable value of surface microhardness. max It is the maximum allowable value of surface microhardness.

[0119] ⑤The constraint condition of the damage layer depth is:

[0120] According to the residual stress cosine attenuation function prediction model, as shown in formula (14), when the residual stress value at depth h is less than or equal to the matrix residual stress, the depth h at this time is considered to be the residual stress influence layer depth value, which is numerically equal to the damage layer depth D. The inverse function calculation method is used to calculate the residual stress influence layer depth h, that is, the damage layer depth D is obtained, as shown in formula (14).

[0121] Ae -λh cos(wh+θ)+σ0≤σ0 (14)

[0122] Let σ rsb =Ae -λh cos(wh+θ)≤0, then:

[0123] h=f -1 (σ rsb ) (15)

[0124] Where A, λ, ω, and θ are the model control factors in the residual stress prediction model. The values ​​of each model control factor under given milling parameters can be obtained according to the established relationship model between milling parameters and model control factors.

[0125] D1 p0 (16)

[0126] Where: D1 is the depth of the damaged layer caused by fine milling, a p0 The depth of material removal during the finishing process.

[0127] In other embodiments of the present application, a titanium alloy component surface integrity efficient precision milling parameter optimization control method comprises the following steps: determining the initial process parameter domain Ω0(v c ,f z ,a p ,a e ​), design a milling process test within the initial process parameter domain Ω0, test the surface state characteristics of the specimen after processing, summarize, organize, statistically analyze, and nonlinearly fit all test results to establish a first relationship model; perform sensitivity analysis on the established mapping relationship model between surface state characteristics and milling parameters, and determine the surface state characteristic stable process parameter domain Ω1; select parameter processing components in the surface state characteristic stable process parameter domain Ω1, and conduct fatigue life tests to establish a component fatigue life prediction model based on surface state characteristics; take fatigue life as the criterion, establish surface state characteristic constraint conditions, and obtain surface integrity process parameter domain Ω2; take high efficiency as the goal, perform process parameter domain optimization in the surface integrity process parameter domain Ω2, and obtain surface integrity high-efficiency precision milling process parameter domain Ω3.

[0128] The present invention establishes a relationship model between surface state characteristics and process parameters by adopting titanium alloy milling test, obtains the process parameter domain of stable surface state characteristics of the processed specimen, and then establishes a relationship model between fatigue life and surface state characteristics through blade processing and fatigue test; finally, based on the high efficiency goal, the blade fatigue life requirement is established as the criterion, and the surface state characteristic range is used as the constraint condition to calculate and obtain the blade surface integrity high-efficiency precision milling parameters. The blade is processed by the surface integrity high-efficiency precision milling parameters, which improves the processing efficiency, optimizes the surface state characteristics, and increases the fatigue life of the blade while meeting the requirements of the blade's dimensional accuracy and shape and position accuracy.

[0129] Step 1: Determine the initial process parameter domain Ω0(v based on research and literature review c ,f z ,a p ,a e ), design the milling process test in the initial process parameter domain Ω0. Test the surface state characteristics of the specimen after processing (including surface stress concentration factor, surface residual stress, surface microhardness), summarize, statistically analyze, and nonlinearly fit all test results, and establish a mapping relationship model between surface state characteristics and milling parameters. The specific method is as follows:

[0130] Step 1.1 Under the selected machine tool, tool, and coolant conditions, the initial process parameter domain Ω0(v c ,f z ,a p ,a e ), design the milling process test within the initial process parameter domain Ω0. Based on the actual application object, the material of this case is TC17 titanium alloy, and the initial process parameter domain Ω0 is determined to be v c ∈[75,395]m / min,f z∈[0.02,0.06]mm / z,a p ∈[0.05,0.25]mm,a e ∈[0.1,0.5]mm.

[0131] Step 1.2: In the initial process parameter domain Ω0, the experimental design is completed by combining the four-factor and five-level full factor response surface test method with the single factor test. The milling processing test is carried out using a fixed machine tool, tool, and cooling and lubrication method, and a new tool is used for each set of parameters.

[0132] Step 1.3 Test the surface characteristics of each set of process test pieces. Use a roughness tester to test the surface roughness of each set of test pieces after processing, and calculate the surface stress concentration factor K based on the test results. st , using X-ray stress testing system to measure the residual stress σ r The microhardness HV is tested using a microhardness tester.

[0133] Step 1.4: According to the process parameters of each group of tests and the surface state characteristic parameters of the processed specimens, a relationship model between the surface state characteristics of the specimens and the process parameters is established, as shown in formula (17);

[0134]

[0135] In the formula, K stx is the surface stress concentration factor of the specimen along the feeding direction, K sty is the surface stress concentration factor of the specimen along the cutting width direction, σ rx is the surface residual stress of the specimen along the feeding direction, σ rx is the surface residual stress of the specimen along the cutting width direction, and HV is the surface hardness of the specimen.

[0136] Step 2: Conduct sensitivity analysis on the established mapping relationship model between surface state characteristics and milling parameters to determine the stable process parameter domain Ω1 of surface state characteristics.

[0137] Step 2.1 Based on the concept of stable parameter domain, sensitivity analysis is performed on the mapping relationship model between surface state characteristics and milling parameters established by equation (17) to obtain the absolute sensitivity between surface state characteristics and milling parameters, as follows: Figure 1 to Figure 4 shown.

[0138] Step 2.2 Based on the absolute sensitivity change, the sensitivity amplitudes of cutting speed, feed rate, cutting depth and cutting width in different parameter ranges are calculated respectively, as shown in Table 1. The change amplitude includes the stress concentration factor, surface residual stress and surface microhardness change amplitude. Calculate the average value based on multiple change amplitudes: According to the definition of the stable process parameter domain, it can be seen that the sensitivity change amplitude within the stable process parameter domain must be less than the average value of the sensitivity amplitude of all intervals of the parameter factor.

[0139] For example, the sensitivity analysis of stress concentration factor, surface residual stress, and surface microhardness to each process parameter is performed within the initial processing parameter range. According to the sensitivity curve, the input variables (x1, ..., x p , …, x n ) p Divided into m levels (M1, ..., M q , …, M m ), according to the sensitivity curve, the input variable x p Calculate the sensitivity values ​​in [M1,M2], [M2,M3], …, [M m-1 ,M m ] and other m-1 horizontal intervals, denoted as A1, A2, ..., A m-1 Mathematically speaking, sensitivity reflects the gradient of the output function to the design variables. When solving the sensitivity of one of the design variables, the other design variables take the intermediate values.

[0140] Step 2.3 According to Figure 1 The sensitivity change curve is shown in Table 1. Figure 2 As shown in the sensitivity change curve, the sensitivity change amplitude within the four horizontal ranges of each tooth feed is calculated; Figure 3 As shown in the sensitivity change curve, the sensitivity change amplitude within the four horizontal ranges of cutting depth is calculated; Figure 4 As shown in the sensitivity change curve, the sensitivity change amplitude within the four horizontal ranges of the cutting width is calculated.

[0141] Calculate the m-1 sensitivity value change amplitudes A1, ..., A j , …, A m-1 The average value of is A0; if A j >A0, then the interval is an unstable domain interval; if A j <A0, then the interval is a stable domain interval; obtain the stable process parameter domain Ω1 of the input variable: [x 1min ,x 1max ]、[x 2min ,x 2max ], …, [ximin ,x imax ].

[0142] Step 2.4 For different intervals of the four process parameters, the average value of the sensitivity change amplitude of the stress concentration factor, residual stress, and microhardness in each interval is smaller than the average value of the sensitivity change amplitude of all intervals, and the surface state characteristic stable process parameter domain Ω1 is determined as shown in Table 2, that is, v c =[195,395]m / min, f z =[0.03,0.06]mm / z, a p =[0.1,0.25]mm, a e =[0.2,0.5]mm.

[0143] Table 1 Sensitivity analysis results

[0144]

[0145] Table 2 Sensitivity change amplitude and stability parameter domain analysis results

[0146]

[0147] Step 3: Select parameter processing components in the surface state characteristic stable process parameter domain Ω1, and conduct fatigue life tests to establish a component fatigue life prediction model based on surface state characteristics; using fatigue life as the criterion, establish surface state characteristic constraint conditions to obtain the surface integrity process parameter domain Ω2.

[0148] Step 3.1 Specifically, in the surface state characteristic stable process parameter domain Ω1, five different milling parameters shown in Table 3 are selected to complete the fine milling blade profile process in the blade unit processing. Each set of parameters processes 4 blades, totaling 20 blades. The integral blade disc fine milling is carried out in the Liechti G-mill 1150 five-axis high-speed milling machining center. The tool is a Zhuzuan BR5×3×XD16 taper ball head cutter with four teeth, and the milling method is spiral milling.

[0149] Table 3 Blade milling parameters

[0150]

[0151] Step 3.2 For the above processed blades, select one blade in each group for surface state characteristic test, and the other three blades for vibration fatigue life test. According to the surface state characteristic test data and the vibration fatigue life test results, a fatigue life prediction model based on surface state characteristics is established, as shown in formula (18):

[0152]

[0153] Where: Nf is fatigue life, K st is the stress concentration factor, HV is the surface microhardness, σ r is the surface residual stress value.

[0154] Step 3.3 For this example, the fatigue life is greater than the matrix life as the criterion, and the fatigue life coefficient constraint condition is established, which can be expressed as:

[0155]

[0156] Step 3.4 For this example, establish the surface state characteristic stable process parameter domain constraint conditions, which can be expressed as:

[0157] 195≤v c ≤395 (21)

[0158] 0.03≤f z ≤0.06 (22)

[0159] 0.1≤a p ≤0.25 (23)

[0160] 0.2≤a e ≤0.5 (24)

[0161] Step 3.5 For this example, the surface state feature constraints are established in sequence, which can be expressed as:

[0162] The surface stress concentration factor constraint condition can be expressed as: K st ≤1.056(25)

[0163]

[0164] The surface residual stress constraint condition can be expressed as: r ≤-152.215(27)

[0165]

[0166] The surface microhardness constraint condition can be expressed as: HV ≥ 393.24 (29)

[0167]

[0168] The constraint condition of the damage layer depth can be expressed as:

[0169] h<0.05 (31)

[0170] Where h = f -1 (σ rsb ) (32)

[0171] σrsb =Ae -λh cos(wh+θ) and σ rsb ≤0 (33)

[0172]

[0173] Step 4: With high efficiency as the goal, the process parameter domain is optimized within the surface integrity process parameter domain Ω2 to obtain the surface integrity high-efficiency precision milling process parameter domain Ω3.

[0174] Step 4.1 In this implementation, based on the surface integrity high-efficiency precision milling parameter domain optimization model, with high efficiency as the optimization goal, the optimization objective function is established:

[0175]

[0176] Specifically, the ball-end cutter used in the test has a diameter of 10 mm and a number of teeth of 4. The optimization objective function can be converted into:

[0177]

[0178] Step 4.2 After establishing the optimization objectives and constraints, a mathematical optimization algorithm is used to optimize and solve the above nonlinear programming problem, and then the solution results are sorted to obtain the efficient and precise milling parameter domain Ω3 of the titanium alloy blade surface integrity.

[0179] Step 4.3 In this implementation, the genetic optimization algorithm is used in MATLAB to optimize and solve the above nonlinear programming problem. The initial population size is 80, the crossover probability is 0.9, the mutation probability is 0.1, the elite ratio is 0.1, the maximum number of iterations is 500, and the genetic algorithm iteration curve is calculated. The optimization results are sorted and analyzed, and the surface integrity and high-efficiency precision milling parameters of TC17 titanium alloy blades are obtained as follows: c =395m / min, f z =0.04mm / z, a p =0.13mm, a e =0.2mm.

[0180] Step 4.4 Experimental verification of parameter optimization results. The TC17 titanium alloy blade surface integrity high-efficiency precision milling parameters and conventional parameters were used to complete the processing of the blade test piece, and the surface state test and fatigue life test were carried out to verify that the parameter optimization scheme can achieve the purpose of meeting the requirements of blade dimensional accuracy and shape and position accuracy while improving processing efficiency, enhancing surface state characteristics, and increasing blade fatigue life.

[0181] The optimized milling parameters and conventional milling parameters shown in Table 4 were used to complete the machining of the blade test piece. The material removal rates of the blades machined with two groups of different parameters were calculated respectively, and the surface state characteristics and fatigue life of the two groups of blades were tested.

[0182] Table 4 Comparative verification test plan

[0183]

[0184] Table 5 shows the results of material removal rate, surface roughness, surface residual stress, and surface microhardness of the blade. Compared with the conventional parameters, the processing efficiency of the optimized parameters is improved by 156%. The stress concentration factors of the blade processed by the optimized parameters in the feed direction and the cutting width direction are 1.0062 and 1.011, and the stress concentration factors of the blade processed by the conventional parameters are 1.0058 and 1.014; compared with the conventional parameters, the conventional parameters increase by 0.0004 in the feed direction and decrease by 0.003 in the cutting width direction.

[0185] Table 5 Comparison of surface characteristics

[0186]

[0187] The distribution of blade residual stress along the depth direction is shown in Figure 6 The analysis shows that the depth of the residual stress influence layer is 25-30 μm, and the consistency and stability of the surface residual compressive stress values ​​of the blade processed by the optimized parameters are greater than those of the conventional processing parameters.

[0188] The distribution of leaf microhardness along the depth direction is shown in Figure 7 The surface hardness of the blade processed by the optimized parameters is 408.23HV, the hardening rate is 6.03%, and the hardened layer depth is about 20μm; the surface hardness of the blade processed by the conventional parameters is 394.67HV, the hardening rate is 2.51%, and the hardened layer depth is about 15μm.

[0189] The fatigue life of blades processed with different process parameters is as follows Figure 8 The median fatigue life of the blade processed with the optimized parameters is 9.03×10 5 The median fatigue life of the blade processed with conventional parameters is 6.21×10 5 Compared with conventional parameters, the median fatigue life of blades processed with optimized parameters is increased by 45%.

Claims

1. A method for optimizing milling parameters of titanium alloy component surface integrity, characterized in that: The method comprises the following steps: Step 1: Determine the initial process parameter domain Ω0 based on research and literature review, design a milling process test within the initial process parameter domain Ω0, test the surface state characteristics of the specimen after processing, summarize, statistically analyze, and nonlinearly fit all test results, and establish a mapping relationship model between surface state characteristics and milling parameters; Step 2: Conduct sensitivity analysis on the established mapping relationship model between surface state characteristics and milling parameters to determine the stable process parameter domain Ω1 of surface state characteristics; Step 3: Select parameter processing components in the surface state characteristic stable process parameter domain Ω1, and conduct fatigue life tests to establish a component fatigue life prediction model based on surface state characteristics; take fatigue life as the criterion, establish surface state characteristic constraint conditions, and obtain the surface integrity process parameter domain Ω2; Step 4: With high efficiency as the goal, the process parameter domain is optimized within the surface integrity process parameter domain Ω2 to obtain the surface integrity high-efficiency precision milling process parameter domain Ω3.

2. The method according to claim 1, characterized in that The initial process parameter domain Ω0 includes v c ,f z ,a p ,a e .

3. The method according to claim 2, characterized in that The surface state characteristics include surface stress concentration factor, surface residual stress, and surface microhardness. The specific characterization symbol is: K stx Represents the surface stress concentration factor of the specimen along the feeding direction, K sty Represents the surface stress concentration factor of the specimen along the cutting width direction, σ rx represents the surface residual stress of the specimen along the feeding direction, σ rx represents the surface residual stress of the specimen along the cutting width direction, and HV represents the surface hardness of the specimen.

4. The method according to claim 3, characterized in that The surface state characteristic constraints include: surface state characteristic stable process parameter domain constraints, fatigue life constraints, surface stress concentration constraints, surface residual stress constraints, surface microhardness constraints and damage layer depth constraints.

5. The method according to claim 4, characterized in that The step 1 comprises: Step 1.1 Determine the initial process parameter domain Ω0 for titanium alloy component processing; Step 1.2: In the initial process parameter domain Ω0, the full factorial response surface method is used to perform the experimental design; Step 1.3 Perform a milling test according to the response surface test design in step 1.2 to test the surface state characteristic parameters of the test piece processed by each set of process parameters; Step 1.4 establishes a relationship model between the surface state characteristics of the specimen and the process parameters according to the process parameters and the test data of the surface state characteristic parameters of each set of parameters; In the formula, SS is the surface state characteristic, a i (i=1,2,3,4,5) is the fitting coefficient.

6. The method according to claim 5, characterized in that The step 2 performs sensitivity analysis on the established mapping relationship model between the surface state characteristics and the milling parameters according to formula (2); The input variables x1, ..., x p , …, x n x in p Divided into m levels M1,…,M q , …, M m , according to the sensitivity curve, for the input variable x p Calculate the sensitivity values ​​in [M1,M2], [M2,M3], …, [M m-1 ,M m ] The amplitude of change in the m-1 horizontal intervals is recorded as A1, A2, ..., A m-1 ; Calculate the m-1 sensitivity value change amplitudes A1, ..., A j , …, A m-1 The average value of is A0; If A j >A0, then the interval is an unstable domain interval; If A j <A0, then the interval is a stable domain interval; Obtain the stable process parameter domain Ω1 of the input variables: [x 1min ,x 1max ]、[x 2min ,x 2max ], …, [x nmin ,x nmax ]; In the formula, n is the number of variables.

7. The method according to claim 6, characterized in that The step 3 selects parameters to process components in the surface state characteristic stable process parameter domain Ω1, and performs fatigue life test to establish a component fatigue life prediction model based on surface state characteristics, as shown in formula (3): Where N f is fatigue life, N f0 is the fatigue life of the matrix material, K st is the stress concentration factor, HV is the microhardness value, HV0 is the matrix microhardness value, σ r is the residual stress value / MPa, σ b is the tensile strength of the material / MPa, a i , i=0,1,2,3 are model coefficients; When the stress concentration factor is 1, the microhardness value is the microhardness of the matrix, and the residual stress is equal to the tensile strength of the material, N f Equal to N f0 ; Taking fatigue life as the criterion, the constraint conditions of surface state characteristics are established, and the surface integrity process parameter domain Ω2 is obtained by solving; among them, the constraint conditions of process parameters and surface state characteristics, as well as the fatigue life criterion conditions are as follows: ① Establish the constraints of the stable process parameter domain of surface state characteristics: in cmin ≤in c ≤in cmax (4) f zmin ≤f z ≤f zmax (5) a pmin ≤a p ≤a pmax (6) a emin ≤a e ≤a emax (7) Where: v cmax is the maximum value of milling speed, v cmin is the minimum value of milling speed, f zmax is the maximum feed per tooth, f zmin is the minimum feed per tooth, a pmax is the maximum cutting depth, a pmin is the minimum cutting depth, a emax is the maximum value of the cutting width, a emin is the minimum value of cutting width; ② Fatigue life constraints: N f ≥ N f0 (8) Where N f is fatigue life, N f0 is the fatigue life of the matrix material, K st is the stress concentration factor, HV is the microhardness value, HV0 is the matrix microhardness value, σ r is the residual stress value / MPa, σ b is the tensile strength of the material / MPa, a i (i=0,1,2,3) is the model coefficient; when the stress concentration factor is 1, the microhardness value is the microhardness of the matrix, and the residual stress is equal to the tensile strength of the material, N f Equal to N f0 ; ③Surface stress concentration constraint conditions: K stmin ≤K st (v c ,f z ,a p ,a e )≤K stmax (10) Where: K st (v c ,f z ,a p ,a e ) is the established mapping relationship model between milling parameters and surface stress concentration, K stmin is the minimum value of the surface stress concentration factor, K stmax is the maximum value of the surface stress concentration factor; ④Surface residual stress constraint conditions: s rmin ≤σ r (v c ,f z ,a p ,a e )≤σ rmax (11) Where: r (v c ,f z ,a p ,a e ) is the established mapping relationship model between milling parameters and surface residual stress, σ rmin is the minimum allowable value of surface residual stress or the maximum allowable value of residual compressive stress, σ rmax is the maximum allowable value of surface residual stress or the minimum allowable value of residual compressive stress; ⑤Surface microhardness constraints: HV min ≤HV(v c , f z , a p , a e )≤HV max (12) Where: HV(v c ,f z ,a p ,a e ) is the established mapping relationship model between milling parameters and surface microhardness, HV min HV is the minimum allowable value of surface microhardness. max is the maximum allowable value of surface microhardness; ⑥ Damage layer depth constraint conditions: According to the residual stress cosine attenuation function prediction model, as shown in formula (13), when the residual stress value at depth h is less than or equal to the matrix residual stress, the depth h at this time is considered to be the residual stress influence layer depth value, which is numerically equal to the damage layer depth D; the inverse function calculation method is used to calculate the residual stress influence layer depth h, that is, the damage layer depth D is obtained, as shown in formula (14); Oh, oh -λh cos(wh+θ)+σ0≤σ0 (13) Let σ rsb =Ae -λh cos(wh+θ)≤0, then: h=f -1 (s rsb ) (14) Where: A, λ, ω, θ are the model control factors in the residual stress prediction model. The values ​​of each model control factor under given milling parameters can be obtained according to the established relationship model between milling parameters and model control factors. D1 p0 (15)​ Where: D1 is the depth of the damaged layer caused by fine milling, a p0 The depth of material removal during the finishing process.

8. The method according to claim 7, characterized in that In step 4, an efficient objective function as shown in formula (16) is established, and the process parameter domain is optimized within the surface integrity process parameter domain Ω2 by using the MATLAB mathematical optimization tool to obtain the surface integrity efficient precision milling process parameter domain Ω3: maxQ=f(v c ,f z ,a p ,a e ) (16)。

Images