基于表面形貌的滑动磨合磨损预测理论与方法研究
发布时间：2018-08-18 20:45
【摘要】：磨合是机器系统必须经历的重要磨损阶段,磨合磨损是机加工表面经过动态磨损达到稳定低磨损率表面的过程,它对防止发生咬合、获得低磨损率的稳定持续工作状态、延长使用寿命等都有重要影响。摩擦副表面形貌变化是磨合过程中的重要特征表现,磨合后表面形貌直接影响摩擦副稳定工作状态下的性能。此外,与稳定磨损相比,磨合过程磨损量大,磨合磨损量直接影响摩擦副的使用寿命。由于滑动摩擦过程中摩擦副表面接触面积较大,表面形貌对滑动摩擦副的影响较为显著。探讨滑动摩擦副磨合磨损表面形貌的变化规律,研究磨合磨损前后摩擦副表面形貌的关联关系,研究基于磨合前初始表面形貌的磨合磨损特性包括磨合后表面形貌与磨损量的预测理论和方法,对于摩擦副表面形貌优化设计,以保证摩擦副工作性能、和使用寿命都具有重要意义。本文主要工作与创新如下： 以接触力学为理论基础,建立了滑动摩擦副的磨合磨损解析模型。基于接触力学与混合润滑特征,依据Greenwood-Williamson摩擦模型与表面微形貌凸峰结构受力原理,同时考虑摩擦副黏着接触,并以摩擦系数稳定为磨合结束判据,提出并建立了一种基于滑动摩擦副的磨合磨损解析模型。 理论和实验论证了滑动摩擦副磨合前后表面形貌的相关性,为基于摩擦副表面形貌的磨合磨损特性预测提供了依据。基于磨合磨损解析模型,模拟磨合磨损过程并分析表面形貌的变化规律,论证了磨合前后表面形貌的相关性。同时,试验分析了不同工况与初始表面形貌对磨合后表面形貌的影响,实验验证了磨合前后表面形貌的相关性。 基于最大相关最小冗余特征选择方法,提出了磨合磨损预测模型的表面形貌特征参数有效提取方法。由于区域法表面形貌评定参数众多,将全部参数进行磨合磨损预测建模不但使模型复杂度过高,且势必造成表面表征的冗余性,因而,必须对区域法表面形貌评定参数进行选择,提取出与磨合磨损特性相关的特征参数。通过区域法表面形貌评定参数的特性分析,本文提出了基于最大相关最小冗余特征选择,面向磨合磨损预测模型的特征参数集合的有效提取方法。 提出了基于最小二乘支持向量机,建立磨合磨损特性包括表面形貌与磨损量预测模型的有效建模方法。根据提取到的表面形貌特征参数集合,建立能够预测磨合表面形貌参数与磨合磨损量的磨合磨损预测模型。该方法基于磨合前后表面形貌的相关性论证结论,考虑到解析磨合模型不能有效表征磨合过程的复杂性,将磨合过程作为黑箱过程进行机器学习,避免由于对磨合过程的简化造成的偏差,建立磨合过程输入变量(工况与表面形貌)与磨合过程输出变量(表面形貌或磨合磨损量)之间的相互关系,以实现基于摩擦副表面形貌的磨合磨损特性预测。。 实验测试验证了磨合磨损特性预测建模方法的有效性和准确性。
[Abstract]:Running-in is an important wear stage that the machine system must go through. Running-in wear is a process in which the machined surface achieves a stable low wear rate surface through dynamic wear. It has an important influence on preventing occlusion, obtaining a stable and continuous working state with low wear rate and prolonging service life. In addition, compared with stable wear, the wear amount during running-in process is larger, and the wear amount during running-in directly affects the service life of the friction pair. The relationship between the surface topography of friction pairs before and after running-in wear is studied. The theory and method of predicting the surface topography and wear amount after running-in based on the initial surface topography before running-in wear are studied. In order to ensure the working performance and service life of the friction pair, the design is of great significance.
On the basis of contact mechanics, an analytical model of running-in wear of sliding friction pairs is established. Based on contact mechanics and mixed lubrication characteristics, according to Greenwood-Williamson friction model and the stress principle of surface micro-topography convex structure, the adhesive contact of friction pairs is considered, and the stability of friction coefficient is taken as the end criterion of running-in. An analytical model of running in wear based on sliding friction pair is presented.
The correlation between the surface topography of sliding friction pairs before and after running-in is theoretically and experimentally demonstrated, which provides a basis for predicting the running-in wear characteristics based on the surface topography of sliding friction pairs. The influence of different working conditions and initial surface morphology on the surface morphology after running-in was analyzed. The correlation between the surface morphology before and after running-in was verified by experiments.
Based on the method of maximum correlation and minimum redundancy feature selection, an effective method for extracting surface topography parameters of wear prediction model is proposed. It is necessary to select the parameters of regional surface topography evaluation and extract the characteristic parameters related to running-in wear characteristics.Based on the characteristic analysis of the parameters of regional surface topography evaluation, an effective method for extracting the feature parameters set of running-in wear prediction model is proposed.
An effective modeling method based on least squares support vector machine (LS-SVM) is proposed to establish a prediction model of running-in wear characteristics including surface topography and wear loss. Considering that the analytical running-in model can not effectively characterize the complexity of running-in process, the running-in process is regarded as a black box process for machine learning to avoid the deviation caused by the simplification of running-in process, and the input variables of running-in process (working condition and surface morphology) and the output variables of running-in process (surface shape) are established. In order to predict the running-in wear characteristics based on the surface morphology of the friction pair, the relationship between the wear morphology and the running-in wear amount was studied.
The experimental results verify the validity and accuracy of the prediction method of running in wear characteristics.
【学位授予单位】：华中科技大学
【学位级别】：博士
【学位授予年份】：2013
【分类号】：TH117
本文编号：2190626
[Abstract]:Running-in is an important wear stage that the machine system must go through. Running-in wear is a process in which the machined surface achieves a stable low wear rate surface through dynamic wear. It has an important influence on preventing occlusion, obtaining a stable and continuous working state with low wear rate and prolonging service life. In addition, compared with stable wear, the wear amount during running-in process is larger, and the wear amount during running-in directly affects the service life of the friction pair. The relationship between the surface topography of friction pairs before and after running-in wear is studied. The theory and method of predicting the surface topography and wear amount after running-in based on the initial surface topography before running-in wear are studied. In order to ensure the working performance and service life of the friction pair, the design is of great significance.
On the basis of contact mechanics, an analytical model of running-in wear of sliding friction pairs is established. Based on contact mechanics and mixed lubrication characteristics, according to Greenwood-Williamson friction model and the stress principle of surface micro-topography convex structure, the adhesive contact of friction pairs is considered, and the stability of friction coefficient is taken as the end criterion of running-in. An analytical model of running in wear based on sliding friction pair is presented.
The correlation between the surface topography of sliding friction pairs before and after running-in is theoretically and experimentally demonstrated, which provides a basis for predicting the running-in wear characteristics based on the surface topography of sliding friction pairs. The influence of different working conditions and initial surface morphology on the surface morphology after running-in was analyzed. The correlation between the surface morphology before and after running-in was verified by experiments.
Based on the method of maximum correlation and minimum redundancy feature selection, an effective method for extracting surface topography parameters of wear prediction model is proposed. It is necessary to select the parameters of regional surface topography evaluation and extract the characteristic parameters related to running-in wear characteristics.Based on the characteristic analysis of the parameters of regional surface topography evaluation, an effective method for extracting the feature parameters set of running-in wear prediction model is proposed.
An effective modeling method based on least squares support vector machine (LS-SVM) is proposed to establish a prediction model of running-in wear characteristics including surface topography and wear loss. Considering that the analytical running-in model can not effectively characterize the complexity of running-in process, the running-in process is regarded as a black box process for machine learning to avoid the deviation caused by the simplification of running-in process, and the input variables of running-in process (working condition and surface morphology) and the output variables of running-in process (surface shape) are established. In order to predict the running-in wear characteristics based on the surface morphology of the friction pair, the relationship between the wear morphology and the running-in wear amount was studied.
The experimental results verify the validity and accuracy of the prediction method of running in wear characteristics.
【学位授予单位】：华中科技大学
【学位级别】：博士
【学位授予年份】：2013
【分类号】：TH117
【参考文献】
相关期刊论文 前4条
1 韩德宝,关德林;船舶主机汽缸套—活塞环磨合过程的模拟试验[J];大连海事大学学报;2004年04期
2 张移山,孔宪梅,陈大融;磨合过程的摩擦学设计研究[J];清华大学学报(自然科学版);1998年04期
3 杨练根,谢铁邦,蒋向前,李柱;表面形貌的Motif评定方法及其发展[J];中国机械工程;2002年21期
4 葛世荣,陈国安;磨合表面形貌变化的特征粗糙度参数表征[J];中国矿业大学学报;1999年03期
相关博士学位论文 前1条
1 汪选国;销盘滑动磨损试验的仿真方法研究[D];武汉理工大学;2009年
，本文编号：2190626
本文链接：https://www.wllwen.com/kejilunwen/jixiegongcheng/2190626.html
