機(jī)械外文文獻(xiàn)翻譯--專用機(jī)床多軸箱的可靠性分析【中文3216字】【PDF+中文WORD】,中文3216字,PDF+中文WORD,機(jī)械,外文,文獻(xiàn),翻譯,專用,機(jī)床,軸箱,可靠性分析,中文,3216,PDF,WORD International Journal of Science Vol.3 No.5 2016 ISSN:1813-4890 194 Reliability Analysis on Multi-Axle Box of Dedicated Machine Qingchen Li,Yinghua Liao,Fengjiang Li School of Mechanical Engineering,Sichuan University of Science&Engineering,Zigong 643000,China Abstract In a dedicated machine tool,the box is one of the most important parts of the multi axle box,its deformation will directly affect the position accuracy of the parts in the multi axle box,and then affect the accuracy and reliability of the entire multi axle box.The finite element analysis of the deformation of the box is usually based on certain parameters.Considering the randomness of the modulus of elasticity,the density and the cutting force in practice,the finite element model of the box is established in the finite element analysis software ANSYS,the modulus of elasticity,the density and the cutting force are regarded as random input variable,and the maximum deformation is regarded as output variable,uses Monte Carlo method to analyze the reliability of the box,and gets the reliability of the box under allowable deformation.A feasible analysis method is provided for reliability design of box and similar parts.Keywords box;reliability;Monte Carlo method;ANSYS 1.Introduction In a dedicated machine tool,the box is one of the most important parts of the multi axle box.The axles,bearings,bearing covers and other parts of multi axle box are respectively arranged on the box through different installation forms.The box also fixes the whole multi axle box on the machine tool bed by its own structure.The box not only transmits the force and torque between part with part,part with machined workpiece in the multi axle box,but also plays an important role in ensuring the position accuracy of each rotating axle and the relative position accuracy between rotating axle with machine tools guide rail.The box will deform under the action of external force,its deformation will directly affect the position accuracy of the parts in the multi axle box,and then affect the accuracy and reliability of the entire multi axle box.If the deformation of the box is too large,beyond the allowable range,it may causes the failure of the multi axle box12.The finite element analysis of the deformation of the box is usually based on the certainty parameters,that is to say,the material properties,boundary conditions,loads and other parameters are all assured.However,there are many uncertainties in these parameters,which brings some new problems to the deterministic finite element analysis.On the one hand,with the development of the finite element technology,a lot of high precision elements are put forward,which makes the calculation accuracy more and more high.On the other hand,there is a large statistical randomness in the material parameters,boundary conditions and loads,which makes the high precision of the high precision element greatly reduced due to the influence of the statistical randomness.In order to solve this problem,it is necessary to regard the material properties,loads and other uncertain factors as random variables,the carries on reliability probability analysis of box3.In this paper,the box of a dedicated machine tools drilling multi axle box is taken as an example,the reliability of the box is studied by using the finite element software ANSYS.International Journal of Science Vol.3 No.5 2016 ISSN:1813-4890 195 2.Reliability analysis theory For the case of the multi axle boxs box,its reliability means the probability of completing the specified function under the prescribed conditions and the prescribed time interval.Its failure can be considered that the maximum deformation is more than the allowable deformation,which leads to the processing precision is not required.Failure probability Pf can be expressed as:0)(0)(0)(xgfdXxfxgPP (1)In the formula,g(x)is the structures function under limit state,the random variables X=(X1,X2,Xn)in it express random parameters.When g(x)0,the box is in safe condition.When g(x)=0,the box is in the limit state.When g(x)0,the box is fail.Generally,it is difficult to directly establish the expression of f(x).In ANSYS,the reliability of the box can be studied by Monte Carlo method45.Monte Carlo simulation methods basic idea as follows:Firstly,builds a stochastic model which is corresponding to the researched object,forms a random variable and its digital feature(such as probability,expectation.etc).Then in order to get plenty of random variables sample value,a great deal of random experiments had been done.Lastly,calculates the digital features estimative figure by using the statistic feature.Based on this idea,set the object function Z=g(X1,X2,Xn),among them Xi is random variable,it subjects to a certain distribution.Takes N times random sampling for X,gets N groups of sample.Then substitutes the value of j group into the object function,gets the value of Zj(j=1,2,n).If there are Nj groups of N which made Zj0,the probability of structural failure is:NNPff (2)In the formula,N is the total number of samples,Nf is the number of failure.3.Steps of reliability analysis 3.1 Build the model of box Before using ANSYS to study the reliability of the box,we must first establish the structure model of the box,the structural models quality will affect the efficiency and accuracy of the finite element analysis to a large extent6.Because the ANSYS modeling operation is very complicated,so this paper uses the three-dimensional modeling software Proe to carry on the modeling to the box,then imports the model into ANSYS.The fillet,chamfer,screw holes and other small structure will be divided into many elements when analyze in ANSYS,which will greatly increase the analysis time.The results of the finite element analysis will not produce large changes after ignoring these small features.So,in the process of building structure model,the box is simplified properly.The simplified box includes the subject box and box cover two parts.The model of subject box is shown in figure 1,the model of box cover is shown in figure 2.Figure 1 Model of subject box International Journal of Science Vol.3 No.5 2016 ISSN:1813-4890 196 Because the subject box and the box cover are connected by screws,there will not be relative sliding produced between them,so the two parts of the box are regarded as an integral part of the finite element analysis.The model of box is shown in figure 3.Figure 2 Model of box cover Figure 3 Box model of multi axle box 3.2 Perform reliability analysis When using ANSYS reliability analysis module Prob Design to carry on the reliability analysis to the box body,the following three steps are included:generating analysis file,performing reliability analysis and dealing with results7.(1)Generating analysis file.Analysis file refers to the cyclic document of the reliability analysis,it needs to be written by ANSYS parametric design language APDL.The rationality of the analysis file is an important guarantee for reliability analysiss calculation efficiency and correctness of the results.The creation of analysis file contains pretreatment,solving and extraction results 3 parts.The first step is to pre-treat the box:first,the box model is introduced,and the element type is SOLID185,the material parameter E=1.4e11,the Poissons ratio is 0.25,the density is 7300N/m3.Then meshing the model(the element level is set to level 3).The finite element analysis model of the box is shown in Figure 4.International Journal of Science Vol.3 No.5 2016 ISSN:1813-4890 197 Figure 4 Finite element analysis model of box The second step is to solve,sets full constraints to the contact surface of box and machine tool bed,sets cutting force F=663.8N,and then solves by the Solver.The third step is to extract the results,extracts the maximum deformation of the box,sets to the parameter of U.(2)Performing reliability analysis.After entering the analysis file,you need to repeat the analysis file to achieve reliability analysis,this is the key link of reliability analysis.This stage mainly includes:the definition of input and output variables,the choice of analysis method,the implementation of the reliability cycle 3 steps.Whether static or dynamic analysis,the material is regarded as the ideal material,and its parameters are in good accordance with the theoretical value.But in reality,there is a small difference between the materials,the material parameters can not be absolutely the same.In addition,due to the external conditions such as temperature and other factors such as cutting tool,workpiece,the size of the cutting force is also changing.So,you can specify the modulus of elasticity of the material,the density and the cutting force as random input variables,its parameters are shown in Table 1.Table 1 Random parameters Distribution type Mean value Standard deviation Modulus of elasticity E(Pa)GAUSS 1.4E11 1.4E9 Density(N/m3)GAUSS 7300 100 Cutting force F(N)GAUSS 663.8 20 Next,sets the maximum deformation of the box as the output variable U.Then selects Latin Hypecube of Monte Carlo as analysis method to sample 500 times,Performs reliability analysis.(3)Dealing with results.When cycle completed,you can view the input and output variables sampling value,histogram and probability cumulative curve.4.Reliability analysiss results of the box The three input variables statistical histograms are shown in Figure 5,6 and 7 International Journal of Science Vol.3 No.5 2016 ISSN:1813-4890 198 Figure 5 Histogram of elastic modulus E Figure 6 Histogram of density Figure 7 Histogram of cutting load F In the figure,the horizontal coordinate indicates the value of each random variable,and the longitudinal coordinate is the relative frequency(the proportion that a values sample number for the International Journal of Science Vol.3 No.5 2016 ISSN:1813-4890 199 total sample number).The red line is the curve that fits the frequency of each value.It can be seen from the statistical histogram that the curves of the 3 random variables are very smooth,and the shape conforms to the characteristic of GAUSS distribution,it means the sample number 500 times is sufficient.The value of each input variable is shown in Table 2 Table 2 Value feature of input parameters Mean value Standard deviation Maximum value Minimum value Modulus of elasticity E(Pa)1.4e11 1.398e9 1.44e11 1.357e11 Density(Kg/m3)7299.9 99.86 7589.3 6990.3 Cutting force F(N)663.81 20.06 731.25 605.56 It can be seen from the table that each parameters value is very similar to the preset value,it means the value is reasonable and reliable.Sets the confidence probability to 95%,we can get the output variables sampling,histogram and probability cumulative curve as shown in Figure 8,9 and 10.1.20E-61.30E-61.40E-6U1.35E-61.25E-6 Figure 8 Sampling of output parameter U 1.228E-61.356E-61.259E-61.289E-61.320E-6U Figure 9 Histogram of output parameter U International Journal of Science Vol.3 No.5 2016 ISSN:1813-4890 200 1.20E-61.35E-61.23E-61.26E-61.29E-61.32E-6 Figure 10 Probability cumulative curve of output parameter U It can be seen from the figure that when the modulus of elasticity,density,and the value of the cutting force is no longer constant,the maximum deformation of the box U is no longer fixed,but approximate GAUSS distribution,mean value is about 1.2893 10-6mm,standard deviation is about 2.1436 10-8mm.In addition,we can continue to calculate the reliability of the box under a certain deformation according to the probability cumulative curve.As shown in Table 3,the reliability of the box under some typical allowable deformation is given.Table 3 Reliability of box under typical allowable deformations Allowable deformation(mm)1.25e-6 1.3e-6 1.35e-6 Reliability 43.04%69.83%99.76%Conversely,we can also calculate the allowable deformation under a certain reliability.For example,in order to ensure the reliability of the box is not less than 98%,the boxs allowable deformation can not be less than 1.335e-6mm.In the process of practical design and optimization,it is supposed not only to know which factors may affect the deformation,but also to understand the sensitivity of these factors to the deformation,then to determine the primary and secondary relationships to consider various factors of optimal design.Through the reliability analysis of ANSYS,these data can be clearly obtained.The sensitivity of the modulus of elasticity,density and cutting force to the deformation is shown in Figure 11.Figure 11 Sensitivity distribution map of parameters International Journal of Science Vol.3 No.5 2016 ISSN:1813-4890 201 It can be seen that in the three factors,the density of the material is the most important factor that affects the deformation of the box,followed by the modulus of elasticity,and cutting forces effect is very small.The reason for this phenomenon is that the variation range of cutting force F is small.5.Conclusion(1)When the modulus of elasticity,the density and the cutting force are used as the random input variables,the boxs maximum deformation U is no longer fixed,but approximate GAUSS distribution,mean value is about 1.2893 10-6mm,standard deviation is about 2.1436 10-8mm.(2)The reliability of the box under allowable deformations 1.35 10-6mm is 99.76%.Conversely,in order to ensure the reliability of the box is not less than 98%,the boxs allowable deformation can not be less than 1.335e-6mm.(3)In the three factors,the density of the material is the most important factor that affects the deformation of the box,followed by the modulus of elasticity,and cutting forces effect is very small.It is because the variation range of cutting force F is small.Acknowledgements This article is supported by the Innovation Fund of Postgraduate,Sichuan University of Science&Engineering(No.y2014035)and the Science and Technology Bureau of Zigong City(No.2013J19)and Sichuan University of Science&Engineering(No.2014PY04).References 1 Molodova M,Li Z,Dollevoet R.Axle box acceleration:Measurement and simulation for detection of short track defectsJ.Wear,2011,271(1):349-356.2 Lee J S,Choi S,Kim S S,et al.A mixed filtering approach for track condition monitoring using accelerometers on the axle box and bogieJ.Instrumentation and Measurement,IEEE Transactions on,2012,61(3):749-758.3 Der Kiureghian A,Ke J B.The stochastic finite element method in structural reliabilityJ.Probabilistic Engineering Mechanics,1988,3(2):83-91.4 Zhiyu Sun,Liangyu Chen.The Design Theory and Method of mechanical Reliability M.Beijing:Science Press,2003.5 Zhang Q L,Pei U.Random finite element analysis for stochastical responses of structures J.Computer&structures,1997,62(4):611-616.6 Lee Huihuang.Finite Element Simulation with ANSYS WorkbenchM.Taiwan,Schroff Development Corp,2010.7 Chen Puhui,Xiao Shanshan.Probabilistic Design Methodology for Composite Aircraft J.Journal of Nanjing University of Aeronautics and Astronautics.2012,44(5):683-693 【中文3216字】 International Journal of Science Vol.3 No.5 2016 ISSN: 1813-4890 專用機(jī)床多軸箱的可靠性分析 Qingchen Li, Yinghua Liao, Fengjiang Li 四川理工大學(xué)機(jī)械工程學(xué)院，中國(guó)自貢643000 摘要： 在專用機(jī)床中，箱體是多軸箱最重要的部件之一，其變形將直接影響多軸箱中部件的位置精度，從而影響整個(gè)多軸箱的精度和可靠性。箱體變形的有限元分析通?；谀承﹨?shù)?？紤]到實(shí)際中彈性模量，密度和切削力的隨機(jī)性，在有限元分析軟件ANSYS中建立了箱體的有限元模型，彈性模量，密度和切割力被認(rèn)為是隨機(jī)輸入變量，最大變形被認(rèn)為是輸出變量，采用蒙特卡羅方法分析箱體的可靠性，并獲得箱體允許變形的可靠性。 提供了一種可行的分析方法，用于箱子和類似零件的可靠性設(shè)計(jì)。 關(guān)鍵詞： 箱子； 可靠性；蒙特卡羅方法；ANSYS 1、前言： 在專用的機(jī)床中，箱體是多軸箱最重要的部件之一。多軸箱的軸，軸承，軸承蓋等部件分別通過(guò)不同的安裝形式安裝在箱體上。該箱還通過(guò)其自身的結(jié)構(gòu)將整個(gè)多軸箱固定在機(jī)床上。該箱不僅在多軸箱內(nèi)部分與加工件之間傳遞力和扭矩，而且在確保每個(gè)旋轉(zhuǎn)軸的位置精度以及旋轉(zhuǎn)軸與機(jī)床導(dǎo)軌的相對(duì)位置精度之間起著重要的作用。 箱體在外力作用下會(huì)變形，其變形將直接影響多軸箱內(nèi)部件的位置精度，影響整個(gè)多軸箱的精度和可靠性。如果箱體的變形太大，超出允許范圍，則可能導(dǎo)致多軸箱的故障。 箱體變形的有限元分析通常基于確定性參數(shù)，也就是說(shuō)，材料性質(zhì)，邊界條件，載荷等參數(shù)都是有保證的。然而，這些參數(shù)存在許多不確定性，這對(duì)確定性有限元分析帶來(lái)了一些新的問(wèn)題。 一方面，隨著有限元技術(shù)的發(fā)展，提出了很多高精度元件，使計(jì)算精度越來(lái)越高。另一方面，材料參數(shù)，邊界條件和載荷具有較大的統(tǒng)計(jì)隨機(jī)性，這使得由于統(tǒng)計(jì)隨機(jī)性的影響，高精度元件的高精度大大降低。為了解決這個(gè)問(wèn)題，有必要將材料性質(zhì)，載荷等不確定因素作為隨機(jī)變量，進(jìn)行箱子的可靠性概率分析[3]。在本文中，以專用機(jī)床鉆多軸箱為例，使用有限元軟件ANSYS對(duì)箱體的可靠性進(jìn)行了研究。 2、 可靠性分析理論 對(duì)于多軸箱箱體的情況，其可靠性意味著在規(guī)定的條件和規(guī)定的時(shí)間間隔內(nèi)完成指定功能的概率?？梢哉J(rèn)為其失效最大變形大于允許變形，這樣就不需要加工精度。故障概率Pf可表示為： （1） 在公式中，g（x）是極限狀態(tài)下的結(jié)構(gòu)函數(shù)，其中隨機(jī)變量X =（X1，X2，... Xn）表示隨機(jī)參數(shù)。當(dāng)g（x）> 0時(shí)，箱子處于安全狀態(tài)。當(dāng)g（x）= 0時(shí)，箱子處于極限狀態(tài)。當(dāng)g（x）<0時(shí)，箱子失敗。通常，很難直接建立f（x）的表達(dá)式。在ANSYS中，盒子的可靠性可以通過(guò)蒙特卡羅方法進(jìn)行研究。蒙特卡羅模擬方法的基本思想如下：首先構(gòu)建一個(gè)與研究對(duì)象相對(duì)應(yīng)的隨機(jī)模型，形成隨機(jī)變量及其數(shù)字特征（如概率，期望等）。然后要做大量的隨機(jī)實(shí)驗(yàn)以獲得大量隨機(jī)變量的樣本值。最后，使用統(tǒng)計(jì)特征來(lái)計(jì)算數(shù)字特征的估計(jì)值。 基于這個(gè)想法，設(shè)置對(duì)象函數(shù)Z = g（X1，X2，... Xn），其中Xi是隨機(jī)變量，它遵循一定的分布。對(duì)X進(jìn)行N次隨機(jī)抽樣，得到N組樣本。然后將j組的值代入對(duì)象函數(shù)，得到Zj（j = 1,2，... n）的值。如果N中有Nj組使Zj <0，結(jié)構(gòu)失效的概率為： （2） 在公式中，N是樣本總數(shù)，Nf是失敗次數(shù)。 3.可靠性分析步驟 3.1建立箱子模型 在使用ANSYS研究箱體的可靠性之前，首先必須建立箱體的結(jié)構(gòu)模型，結(jié)構(gòu)模型的質(zhì)量將在很大程度上影響有限元分析的效率和精度。因?yàn)锳NSYS建模操作非常復(fù)雜，所以本文使用三維建模軟件Proe對(duì)框進(jìn)行建模，然后將模型導(dǎo)入ANSYS。在ANSYS中分析時(shí)，圓角，倒角，螺孔等小結(jié)構(gòu)將分為多個(gè)要素，大大增加了分析時(shí)間。有限元分析的結(jié)果在忽略這些小特征之后不會(huì)產(chǎn)生很大的變化。所以在建筑結(jié)構(gòu)模型的過(guò)程中，箱子被簡(jiǎn)化了。簡(jiǎn)化箱包括主題箱和箱蓋兩部分。主題箱的模型如圖1所示，箱蓋的型號(hào)如圖2所示。由于主體箱和箱蓋通過(guò)螺絲連接，因此它們之間不會(huì)產(chǎn)生相對(duì)滑動(dòng)，因此箱子的兩個(gè)部分被認(rèn)為是有限元分析的組成部分。箱型號(hào)如圖3所示。 圖1：主題箱模型 圖2：箱蓋模型 圖3：多軸箱的箱模型 3.2進(jìn)行可靠性分析 當(dāng)使用ANSYS可靠性分析模塊Prob Design對(duì)箱體進(jìn)行可靠性分析時(shí)，包括以下三個(gè)步驟：生成分析文件，執(zhí)行可靠性分析和處理結(jié)果。 （1） 生成分析文件。分析文件是指循環(huán)文件的可靠性分析，需要用ANSYS編寫(xiě)的參數(shù)設(shè)計(jì)語(yǔ)言APDL。分析文件的合理性是可靠性分析計(jì)算效率和結(jié)果正確性的重要保證。分析文件的創(chuàng)建包含預(yù)處理，求解和提取結(jié)果3部分。 第一步是對(duì)箱子進(jìn)行預(yù)處理：首先引入箱型，元件類型為SOLID185，材料參數(shù)E = 1.4e11，泊松比為0.25，密度為7300N / m3。然后網(wǎng)格化模型（元素級(jí)別設(shè)置為級(jí)別3）。箱體的有限元分析模型如圖4所示。 第二步是解決，對(duì)箱體和機(jī)床的接觸面設(shè)置完全限制，設(shè)定切削力F = 663.8N，然后由Solver求解。第三步是提取結(jié)果，提取盒子的最大變形，設(shè)定為U的參數(shù)。 （2）進(jìn)行可靠性分析。 進(jìn)入分析文件后，需要重復(fù)分析文件實(shí)現(xiàn)可靠性分析，這是可靠性分析的關(guān)鍵環(huán)節(jié)。 這個(gè)階段主要包括：輸入和輸出變量的定義，分析方法的選擇，可靠性循環(huán)的執(zhí)行3個(gè)步驟。 無(wú)論是靜態(tài)還是動(dòng)態(tài)分析，材料被認(rèn)為是理想材料，其參數(shù)與理論值相符。但實(shí)際上材料之間存在很小的差異，材料參數(shù)不能絕對(duì)相同。另外，由于外部條件如溫度等因素，如切削刀具，工件，切削力的大小也在變化。因此，您可以指定材料的彈性模量，密度和切割力作為隨機(jī)輸入變量，其參數(shù)如表1所示。 圖4：箱的有限元分析模型 接下來(lái)，將箱的最大變形設(shè)定為輸出變量U。 然后選擇蒙特卡羅的拉丁超立方體作為分析方法循環(huán)500次，進(jìn)行可靠性分析。 （3） 處理結(jié)果。 循環(huán)完成后，可以查看輸入和輸出變量的采樣值，直方圖和概率累積曲線。 表1隨機(jī)參數(shù) · 分布型 · 平均值 · 標(biāo)準(zhǔn)差 彈性模量E(Pa) · 高斯 · 1.4E11 · 1.4E9 密度ρ(N / m3) · 高斯 · 7300 · 100 切削力f(N) · 高斯 · 663.8 · 20 4. 可靠性分析的結(jié)果 三個(gè)輸入變量的統(tǒng)計(jì)直方圖如圖5,6和7所示： 圖5：彈性模量E直方圖 圖6：密度ρ直方圖 圖7切削力F直方圖 在該圖中，水平坐標(biāo)表示每個(gè)隨機(jī)變量的值，縱坐標(biāo)是相對(duì)頻率（總樣本數(shù)的值的樣本數(shù)的比例）。紅線是適合每個(gè)值的頻率的曲線。從統(tǒng)計(jì)直方圖可以看出，3個(gè)隨機(jī)變量的曲線非常平滑，形狀符合GAUSS分布的特征，這意味著樣本數(shù)量500次就足夠了。 各輸入變量的值如表2所示。 表2輸入?yún)?shù)的值特征 平均值 標(biāo)準(zhǔn)差 最大值 最小值 彈性模量E(Pa) 1.4E11 1.398E9 1.44E11 1.357E11 密度ρ(Kg/ m3) 7299.9 99.86 7589.3 6990.3 切削力f(N) 663.81 20.06 731.25 605.56 從表中可以看出，每個(gè)參數(shù)值與預(yù)設(shè)值非常相似，這意味著該值是合理可靠的。 將置信概率設(shè)置為95％，我們可以得到輸出變量的抽樣，直方圖和概率累積曲線，如圖8,9和10所示。 圖8：輸出參數(shù)U的采樣 圖9：輸出參數(shù)U的直方圖 圖10：輸出參數(shù)U的概率累積曲線 從圖中可以看出，當(dāng)彈性模量，密度和切削力的值不再恒定時(shí)，箱的最大變形U不再固定，而近似GAUSS分布，平均值約為1.2893 ×10-6mm，標(biāo)準(zhǔn)偏差約為2.1436×10-8mm。另外，我們可以根據(jù)概率累積曲線，在一定的變形下繼續(xù)計(jì)算箱子的可靠性。如表3所示，給出了一些典型允許變形下箱體的可靠性。 表3：典型容許變形箱體的可靠性 允許變形（mm） 1.25e-6 1.3e-6 1.35e-6 可靠性 43.04% 69.83% 99.76% 相反，我們也可以在一定的可靠性下計(jì)算出允許的變形。 例如，為了確保箱子的可靠性不低于98％，箱子的允許變形量不能小于1.335e-6mm。 在實(shí)際設(shè)計(jì)和優(yōu)化的過(guò)程中，不僅要知道哪些因素可能會(huì)影響變形，還要了解這些因素對(duì)變形的敏感性，然后確定主要和次要關(guān)系，以考慮各種因素的最優(yōu)設(shè)計(jì)。 通過(guò)ANSYS的可靠性分析，可以清楚地得到這些數(shù)據(jù)。 彈性模量，密度和切削力對(duì)變形的敏感度如圖11所示。 圖11：參數(shù)靈敏度分布圖 可以看出，在三個(gè)因素中，材料的密度是影響箱體變形的最重要因素，其次是彈性模量，切削力的影響非常小。 這種現(xiàn)象的原因是切割力F的變化范圍小。 5、 結(jié)論 （1）當(dāng)彈性模量，密度和切削力用作隨機(jī)輸入變量時(shí)，箱體的最大變形U不再固定，但近似GAUSS分布，平均值約為1.2893×10-6mm，標(biāo)準(zhǔn)偏差 約為2.1436×10-8mm。 （2）允許變形1.35×10-6mm箱體的可靠性為99.76％。 相反，為了確保箱子的可靠性不低于98％，箱子的允許變形不能小于1.335e-6mm。 （3）在三個(gè)因素中，材料的密度是影響箱體變形的最重要因素，其次是彈性模量，切削力的影響非常小。 這是因?yàn)榍邢髁的變化范圍小。 致謝 本文由四川理工大學(xué)（編號(hào)y2014035）研究生創(chuàng)新基金，自貢市科技局（2013J19）和四川理工大學(xué)（2014PY04）支持。