Ain Shams Engineering Journal (Jun 2021)
Finite element method for optimal transformer connection based on induction motor characteristics analysis
Abstract
The Voltage Sags (VSs) may cause irrecoverable impacts on the sensitive loads, such as moving into plugging mode and high pulsations of the torque. So, the various effects of VSs on these loads should be studied, considering all conditions. The non–linear performance of the transformer causes the calculation of VS problem to be complicated. Therefore, a robustness approach should be chosen to solve the problem. In this paper, a novel classification of VS is carried out by using the non–linear solver of the transient Finite Element Method (FEM). In the proposed method, the non–linear behavior of the iron core of the transformer applied using the Jiles–Atherton (JA) vector hysteresis model to increase the precision of the results. To obtain a comprehensive classification, different connections of transformer windings as well as various types of short circuits are considered, and the influence of these classes on the Induction Motor (IM) performance is rendered. To show the accuracy of the proposed method, the VS results of other methods and FEM are compared with IEC standard data by introducing an evaluation metric. The comparison outcomes verified the effectiveness of the proposed method, as the average of errors obtained 4.9%, which is lower than other approaches. Furthermore, an IM is considered to characterize its speed and torque variations. Based on the IM speed deterioration, the best–to–worst connections of the transformer windings are appointed. Finally, the simulation results demonstrate that under single phase SC, when the primary connection of the transformer is Star or grounded Star, the deterioration of speed is lower than others, which are related to the first to third classes. In essence, the key obtained result is that the results of VS calculations with FEM could be a proper database, because of its ability in solving the non-linear problems with minimum errors.
