Early detection of damage to rolling bearings by electrical impedance measurement

该项目的目的是确定一种基于阻抗的损坏指示器，用于早期检测滚动轴承损坏。与基于结构声的损伤检测相比，这种早期损伤检测的优势在于损伤指示器已经在接触表面微观变化的情况下更早响应。在表面损伤发生之前，阻抗的变化可以提供有关即将发生的初始损伤的信息。在进行具有严密监控的阻抗的一系列测试时，测量本身不会导致电感损伤。由于缺乏常规的阻抗测量，无法使用可用的寿命测试。必须使用一种测量方法，该方法可以区分润滑剂和老化的影响，并在低于润滑膜突破电压的测量电压下运行。为了解释测量结果并将其与滚动接触的变化相关联，必须以非破坏性的方式定期检查接触对的表面特性。根据确定的粗糙度和阻抗特性的时间序列，基于可测量的损伤指标，并与滚动轴承表面的变化相关联，开发了经验预测方法。此外，还使用了人工智能方法，以便在有足够的结构化数据可用时，基于黑箱方法对即将发生的故障进行预测，并将这些预测与进一步的测试程序进行比较。在初步测试的基础上，在工作计划中分析了两个基本研究问题，首先是在定义的测试计划中针对一种尺寸的深沟球轴承：RQ1：滚动轴承阻抗的特性变化能否在标称寿命内显示，以及这些是否可以作为即将发生的滚动轴承损坏的损坏指示器。滚动轴承或从其派生的指示器的阻抗是否显示作为累积损伤的函数的特性曲线？RQ2：哪些唯象模型解释了阻抗或派生的统计指标的特性变化？为了在样本情况下调查变量多样性和运行条件，在相同的工作温度下进一步研究：FF3：不同轴承类型和负荷角的阻抗变化或导出的统计指标是否相似？FF4：磨合过程是否影响阻抗的特性变化？FF5：阻抗的特性变化是否与尺寸无关？本项目不研究滚动轴承的纯机械疲劳行为。使用制造商指定的承载能力来评估阻抗的变化。

The aim of the project is to identify an impedance-based damage indicator for the early detection of rolling bearing damage. The advantage of such an early damage detection compared to a damage detection based on structure-borne sound is the earlier response of the damage indicator already in case of microscopic changes of the contact surface. Before surface damage occurs, the change in impedance can provide information about the impending initial damage.Test series with closely monitored impedance are carried out without the measurements themselves causing electrically induced damage. The use of available lifetime tests is not possible due to the lack of regular impedance measurements. A measuring method must be used which can differentiate between lubricant and aging influences and which operates at measuring voltages below the breakthrough voltage of the lubricant film.In order to interpret the measurements and to correlate them with changes in the rolling contact, the surface properties of the contact partners must be regularly examined in a non-destructive manner. With the determined roughness and the time series of the impedance characteristics, empirical prognosis approaches are developed on the basis of measurable damage indicators and correlated with changes in the surface of the rolling bearings. In addition, artificial intelligence methods are used in order to make forecasts based on a black box approach for an imminent failure if a sufficient amount of structured data is available and to compare these with the further test procedure.Based on preliminary tests, two fundamental research questions are analyzed within the work program, firstly for deep groove ball bearings of one size in a defined test program:RQ 1: Can characteristic changes in the impedance of a rolling bearing be demonstrated over the nominal life and can these be used as a damage indicator for imminent rolling bearing damage. Does the electrical impedance of the rolling bearing or an indicator derived from it show a characteristic curve as a function of the cumulative damage?RQ 2: Which phenomenological models explain the characteristic changes of the impedance or the derived statistical indicators?In order to investigate variant diversity and operating conditions in sample cases, further questions are investigated at the same operating temperature:FF 3: Are the impedance changes or the derived statistical indicators similar for different bearing types and load angles?FF 4: Does the running-in procedure influence the characteristic change of the impedance?FF 5: Is the characteristic change of the impedance independent of the size?The purely mechanical Fatigue behavior of rolling bearings is not investigated within the scope of this project. The load carrying capacity specified by the manufacturers is used to evaluate the change in impedance.