Computationally efficient finite‐element‐based methods for the calculation of symmetrical steady‐state load conditions for synchronous generators

Abstract

This study presents an accurate and computationally efficient finite-element (FE) calculation and modelling methods for the simulation of symmetrical steady-state load conditions for synchronous generators. For that purpose, two iterative methods are presented, implemented and compared. Those are Newton's iterative method and the relaxation parameter iterative method, which were both applied to magneto-static and transient FEs simulations. Various methods for obtaining armature voltage and power angle are proposed and analysed from the computational efficiency and the accuracy point of view. All methods showed good convergence results for solving this particular problem. Calculation results were compared with measurements on two synchronous generators installed in power plants.

Computationally efficient steady‐state finite element simulation of multiphase PM AC machines

“…Another limitation of the proposed method is that it employs magnetostatic FE simulations, making it specifically suitable for steady‐state FE simulation of PM AC machines. This means that the method cannot be directly employed for the analysis of machines which operate with induced currents in the steady state, such as rotor bar currents in induction machines [23] or damper winding currents in synchronous machines [5]. However, the method can even be employed for steady‐state FE simulation of induction machines if rotor currents are calculated outside the FE model.…”

“…Design and analysis of modern high efficiency and high torque density multiphase permanent magnet (PM) AC machines are based on finite element (FE) modelling and simulation [1–4]. Main benefits of FE method in electrical machine design are related to accurate modelling of complex machine geometries and reliable assessment of machine performance in the presence of high iron core saturation [5–7]. However, long execution times of FE simulations still pose an obstacle for their broader application in many design problems.…”

“…Several papers propose computationally efficient ways of exploiting machine symmetry with the goal of reducing the overall duration of FE simulations [1–13]. There are two types of symmetries that can be used for this purpose – geometric and time course symmetry [6–8].…”

“…Time course symmetry, on the other hand, exploits the periodic behaviour of the machine magnetic and electric circuits with regard to rotor angular displacement in the steady state [6–8]. When time course symmetry is applied to FE simulation of electrical machines, time savings are achieved by decreasing the value of simulated angular displacement [1–13]. This means that the waveform of magnetic flux density in one stator tooth can be reconstructed from segments of flux density waveforms in all other teeth that are obtained from FE simulations conducted over a minimum electrical angle [1, 3, 4, 10].…”

Application of an Improved Ridgelet Process Neural Network for Predicting the Temperature Rise of Rotor Structure Optimization

“…Since the 20th century, finite element calculation methods have been widely applied in various fields of generator state analysis, with possibilities for solving complex engineering problems [11][12][13]. Hanic et al present accurate and computationally efficient finite element calculation and modeling methods for simulating symmetrical steady-state load conditions in synchronous generators [14]. The finite element calculations and experimental measurements are employed to detect faults in a two-pole synchronous generator.…”