热流固多相耦合作用下采煤机截割部齿轮系统故障机理研究

Due to harsh working environment in coal mining processing the gear systems of coal cutters are vulnerable to failures. A simple fault, such as a crack in the gear system, would knock off the whole mining production line for a couple of days. Hence, it is imperative to monitor the machine condition to prevent break-downs.. One of the most important but challenging issues in mechanical fault diagnosis is failure mechanism. To address this issue, this project aims to investigate the failure mechanism of gear systems in shearer cutting parts of coal cutters and propose a quantitative manner to assess the system operation condition.. This project contains three phases including: (a) modeling of a multi-degree of freedom (DOF) gear tribo-dynamic model with consideration of the thermal-fluid-solid coupling effect, (b) investigation of gear failure mechanism, and (c) development of fault identification state space for fault quantitative assessment.. In the first phase, a new gear tribo-dynamic model will be developed. This new model couples a gear lubrication-friction model and a 6 DOF gear dynamic model. The factors of the dynamic load, temperature field, non-newtonian fluid, contact surface roughness will be included in the lubrication-friction model, whose outputs will be used to calculate the time variant meshing stiffness, damping and backlash for feeding the 6 DOF dynamic model.. Then, in the second phase, we will investigate the gear failure mechanism using the established tribo-dynamic coupling model. The tribological behavior and dynamic response of the model will be analyzed to reveal the fault root and development mode. The fault indicators will be developed in theoretical to assess the health condition of the gear systems.. Thirdly, we will develop a manifold learning approach for constructing a fault identification state space using practical accessible physical variables. In this approach, the images and static electrostatic signal of the wear particles and the vibration signals of the gear system will be recorded. The tribological fault features and time-frequency features will be extracted by advanced signal processing technologies. The spectral regression based kernel fisher discriminant (SRKFD) will be proposed to nonlinearly project the fault feature space into a manifold subspace to form a new state space, where fault features can approximate to the theoretical fault indicators. Hence, based on this new state space, the quantitative assessment can be established to conduct the condition monitoring and fault diagnosis for the gear systems.. Thus, the findings of this project will provide theoretical basis and technical support for detect early fault of the coal mining machine gear systems in a timely, accurate and reliable manner.

故障机理研究是机械故障诊断的基础与难点。本项目致力于研究采煤机截割部齿轮系统故障机理，提出可靠的故障表征和定量评估方法。研究将立足于热-流-固多相耦合本质，通过研究齿轮啮合区润滑、摩擦磨损和动力学响应间相互耦合作用影响机制，提出热流固多相耦合作用下的齿轮系统摩擦-动力学耦合统一建模理论，揭示齿轮系统故障的形成机理与作用机制。.研究思路：1.考虑动载荷、温升效应、非牛顿流体、接触表面粗糙度等因素，建立有限长线接触齿轮啮合混合弹流润滑理论模型，分析齿轮润滑、摩擦和磨损；2. 基于润滑与摩擦分析结果推导齿轮啮合时变刚度、阻尼和侧隙等，建立齿轮摩擦-动力学耦合信息实时同步计算模型；3.分析齿轮故障机理与发展规律，获取表征故障有效理论变量，利用非线性投影技术构建故障定量评估方法。.从而，通过本项目的研究，为采煤机齿轮系统可靠的状态监测与故障诊断提供理论基础与方法支持。

项目通过研究采煤机齿轮系统在热流固多相耦合作用下的故障机理，建立了考虑摩擦副接触面宏观-微观多尺度效应的物理模型，并提出了控制体积法数值计算方法；探讨了微观流体在磁场作用下的温度场变化机理，提出了温度场数值计算方法；分析了扰流作用下流体与纳米颗粒间的热传递特性，设计了扰流试验模型；提出了基于机器学习的热流固多相耦合流体-颗粒建模方法；设计了考虑温度-负载因素的轴承实验台，通过理论计算、有限元模拟和试验验证，确定了轴承配置方案并完成样机制造；设计了齿轮系统典型故障实验与数据采集，提出了基于变工况条件下齿轮系统故障诊断技术。同时,对已有的工业级油膜轴承试验台进行了能量流分析，科学地评价了试验系统的能量效率，并提出了改进方法；对试验系统的物质流通性进行了定性分析，并在此基础上定量地描述了试验系统中各物质流动情况；优化了试验台直流电动机转速控制系统；设计了针对试验台的远程监控系统，建立了远程控制平台，实现了设备运行的自动化，为增强我国油膜轴承的自主研发创新能力提供了有力支撑。