Model-based detection of local defects in gears

Abstract

The paper is presenting a methodology for developing model-based method of gear fault diagnostics. First the simulation model of the helical gearbox is discussed allowing analysis of the phenomena taking place during teeth mating in the presence of manufacturing and assembly errors. It includes observation of influence of errors on the generated signals. The model was initially used to analyze the teeth contact in the presence of pitch errors and later to verify the sensitiveness of proposed diagnostic methods and their availability to detect the fatigue damages of teeth. The common feature of discussed approach is the direct use of time signal processing algorithms, and in contrary to the methods based on spectral analysis it allows precise localization of gear defects like pitting and tooth fracture associating them to the particular pinion or gear teeth. Their advantage is the simplicity and speed of action that is of great significance for the implementation in the autonomous transmission diagnostic systems and diagnostic systems working online. Presented methods of signal processing were first tested on a simulation model of the gear assembly and later verified during the experiments on a back-to-back test stand.

New approach of model based detection of early stages of fuel injector failures

“…Model-based diagnostics is hugely popular. The frequency [30] and time [14,23] domain are typically used for the mathematical description. Mathematical models are used to estimate the combustion processes and to determine composition of combustion gases generated in the cylinder during a single cycle, however the injector diagnostics relying on the observation of its behaviour in the course of the actual work is not an easy task [17,19].…”

“…In model-based diagnostics [23] the mathematical model represents the proper operation of the device and the corresponding simulated quantities are compared to those measured. The emerging differences called residuals are hints about the existence of errors.…”

Novel Higher-Order Spectral Cross-Correlation Technologies for Vibration Sensor-Based Diagnosis of Gearboxes

“…The frequency bands, chosen for the integration on the IWSC3 and the IWSC4 features, are associated with mechanical properties of the gearbox under test. Although these relations are difficult to assess and such an assessment should be a subject of a specific relatively complicated modelling via model-driven diagnostic research, e.g., [ 61 , 62 ]. This assessment would also require an estimation of a transfer function between a gear’s meshing teeth and sensor location under operating load conditions.…”

Detecting Gear Surface Defects Using Background-Weakening Method and Convolutional Neural Network

“…In the research on fault diagnosis during gear runtime, Mączak and Jasiński [15] discussed the simulation model of the helical gearbox and analyzed a phenomenon during the tooth-meshing process in the presence of manufacturing and assembly errors. This work proposed a kind of gear fault diagnosis method based on the model.…”