MODELING OF ELECTRICAL AND MAGNETIC TRANSIENT PROCESSES IN A SYSTEM WITH AN INDUCTOR AND A STATOR OF AC ELECTRICAL MACHINES

A brief review of the issue of diagnosing the condition of the insulation structure of electrical machines with classical distributed windings is presented. The rationale and relevance of developing and implementing new, innovative diagnostic and testing methods are provided. The proposed method possesses significant diagnostic features that are absent in other currently used methods. The main difference lies in the creation of high test overvoltages through a rapid change in magnetic fluxes in inductively coupled circuits, one of which is an external sensor-inductor, and the second is a part of the magnetic core with a defective or non-defective winding. A simple and convenient testing scheme is proposed, which can be implemented using appropriate measuring equipment. A comprehensive set of theoretical and experimental studies of the inductor-stator system with a winding was conducted using the Matlab Simulink software package and the field method in a two-dimensional problem formulation in the COMSOL Multiphysics software package. Experimental studies were also conducted on a real sample in the form of a stator of a low-power general-purpose induction motor with a random winding, utilizing the developed testing scheme and device. The theoretical and experimental studies show sufficient convergence of results for engineering purposes. The possibility of achieving test overvoltages exceeding 100 V per turn with low energy impact on defective zones is theoretically and experimentally substantiated, which ensures a minimal destructive effect during diagnostics. Substantial results were obtained for further research and development of the field of insulation condition diagnostics using inductive impulse methods. It is demonstrated that the proposed method will be useful for diagnosing the insulation condition of electrical machine windings in a disassembled state during manufacturing and repair processes.