0885-8969 (c) 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information. This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TEC.2017.2648512, IEEE Transactions on Energy Conversion > REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 1 Abstract - This paper deals with the calculation of iron losses in a turgenerator using a magnetodynamic Finite Element (F.E.) Analysis accounting for the eddy currents in the damper bars. Two iron loss models, based on the Berttoti’s decomposition approach, are compared in the post-processing step of a F.E. calculation. The numerical model of the studied system is validated by comparing the calculation and the experiment for no-load conditions. Index Terms — Finite element, Turbogenerator, magnetic losses, iron losses, analytical model. I. INTRODUCTION HE limitation and reduction of greenhouse gas emissions and a rational and efficient use of energy are now a crucial issue of our society. One of the main thrusts for energy saving is the development of efficient electrical machines as these are widely present in industrial and domestic applications. Indeed, in Europe, electrical machines represent 70% of the industrial electrical consumption and 33% of the domestic electrical consumption. Thus, a few percent of improvement in the efficiency of electrical machines would result in a considerable saving of fossil energy. To reach this goal, there exist several approaches, such as the improvement of the design and optimization steps or the control strategy of the electrical machine accounting for energy saving during its operation [1]. But, the main improvement factor is obviously the first step that is related to the accurate design of the electrical machine with limited losses in the context of its operating conditions in an industrial environment. The operating constraints of these machines are even more severe as they are generally operating in variable speed conditions. In fact, the use of static converters contributes to the increase of iron losses due to the high harmonic content to which the magnetic core is subjected [2]. Moreover, these losses will not only burden the performances of the electromechanical device but also induce hot spots that are potential sources of electrical Revised Manuscript received November 03, 2016. (Write the date on which you submitted your paper for review.) M. Fratila and M. Dessoude are with EDF Lab Paris-Saclay, 7 Boulevard Gaspard Monge, 91120 Palaiseau, France (e-mail: mircea.fratila@edf.fr, maxime.dessoude@edf.fr) A. Benabou and A. Tounzi are with Univ. Lille, Centrale Lille, Arts et Metiers ParisTech, HEI, EA 2697 - L2EP – Laboratoire d’Electrotechnique et d’Electronique de Puissance, F-59000 Lille, France (e-mail: abdelkader.benabou@univ-lille1.fr, mounaim.tounzi@univ-lille1.fr). faults. Consequently, the designer needs accurate models to provide optimal electrical machines. One essential contribution to these models is the iron loss calculation that is still a challenging issue for modern electrical machines operating under high harmonic content. In the case of permanent magnet synchronous machines, in addition to the reduction of the efficiency, these harmonics induce losses in the rotor magnets that can be damageable in terms of irreversible demagnetization of the magnets due to the increase of temperature [2]-[5]. In the case of wound rotor synchronous machines, the rotor is generally equipped with damper bars subjected to losses due mainly to the stator and rotor slotting [6], [7]. In both cases, it is of importance to be able to quantify accurately the losses, and more specifically the iron losses, right from the design stage in order to be able to choose a strategy to reduce them. Classically, the modeling of the iron losses is based on the loss decomposition approach proposed by Bertotti [8]. This approach can be either implemented in the post-processing step or in the non-linear resolution (hysteresis) of a time stepping finite element analysis (F.E.A.). The use of a F.E. code to compute the iron losses in an electrical machine enables us to display the iron loss distribution and, so one can modify the structure or adapt the cooling system in order to reduce the losses and increase the efficiency. The 3D F.E.A., including a method that takes into account the hysteresis effects on the distribution of magnetic flux in the magnetic circuit is, intrinsically, the most accurate in determining such losses [9], [10]. Furthermore, the use of a hysteresis model allows one to take into account the effects of minor loops that are induced by the space harmonics or by the static converters providing the use of an adequate hysteresis model. However, this approach still remains quite delicate to systematically ensure the numerical convergence and requires excessive computational time with regard to the improvement that is expected. An alternative is to compute the magnetic losses in a post-processing step of the F.E.A. by using an analytical approach. The latter is by far the most used in the field of electrical engineering and the most used in industrial environments [7], [11]-[13], [17]. Its popularity is mainly due to the following reasons: - it is not necessary to use a hysteresis model in order to obtain the magnetic field and therefore an anhysteretic curve is sufficient, which ensure a fast numerical convergence. - the identification procedure for the analytical model is much simpler than those required by the hysteresis methods. Iron Loss Calculation in a Synchronous Generator Using Finite Element Analysis Mircea Fratila, Abdelkader Benabou, Abdelmounaïm Tounzi, Maxime Dessoude T