The main research activities of the Institute of Electrical Machines, IEM for short, are
- measurement technologies,
- drive control and
- operation of electrical machines.
The IEM with its mechanical workshop is able to build prototypes for measurement purpose, running on more than 35 test benches in the IEM. Measurement technologies for several research issues, as well as measurement electronics, are developed in-house. All relevant machine types are considered in the Institute’s research. To cover the relevant research topics, four research groups are organized at the IEM.
Analysis and Modelling
The group “Analysis and Modelling” deals with parasitical effects in electrical machines. The group’s research focus lies on the characterization of particular physical phenomena, which are analyzed through analytical as well as numerical models. Among several industrial projects, the group also works on numerous DFG projects and participates in a DFG research group. One of the main research topics at the IEM is noise, vibration and harshness, NVH for short, of electrical drive systems. In this context, the researchers at the Institute develop overall system models, which consist of the various interacting system components. Through utilization of the models, different causes of noise excitations, for example power electronics, geometrical details of the magnetic circuit, or excitations originating from mechanical components of the complete drive train are analyzed. Moreover, design rules to avoid the mentioned causes can be derived with the IEM tools and applied at an early design stage of the drive system. Lifetime estimation and the development of innovative insulation systems are other focuses of this research group. Metrological lifetime tests are performed and the results are used to develop new insulation systems and to derive dynamic machine models. Winding test models, so called motorettes, are subjected theoretically as well as experimentally to different load scenarios to study different deterioration effects separately. These include mechanical, thermal and electrical deterioration effects. By modern, fast switching power electronics, high voltage changes dU/dt are fed to the winding system. This can lead to early irreversible failure of the winding system and thus unwanted insulation faults. To characterize the loads on the winding, specific measurement technologies and electronics are developed at IEM, which are used in the practical laboratory tests.In addition to insulation materials, the IEM also deals with the experimental characterization of both, ferromagnetic and permanent-magnetic materials.
This work led to novel material models, which characterize the real physica properties with high precision. Latest findings in this area of material models are the characterization of the material behaviour by micro-scale parameters
and by relevant processing parameters of the materials. Innovative cuttingedge models have been developed to simulate the influences of the manufacturing process, e.g. the process of punching or cutting of electrical steel. Grainoriented, GO for short, and non-grain-oriented, NO for short, materials are studied for diverse magnetizations. The existing infrastructure at IEM enables the metrological investigation of materials from dc-magnetizations up to magnetizing frequencies of 10 kHz. The measurement samples can also be measured under mechanical stress. In various projects, medium frequency power transformers with GO sheets are designed and the prototypes are built and measured or the impact of manufacturing tolerances on the operation behaviour of drive systems has been studied. Current work at the IEM focuses on the statistic distribution of manufacturing impacts and other uncertainties.
Simulation and Methodologies
The development of numerical methods on the algorithmic for solving electromagnetic field problems is the focus of the group “Simulation and Methodologies”. Over the years, an in-house software environment, which focus lies in the finite element method, has been developed in the IEM. In combination with other software products, multi-physical problems can be solved through coupled models. A particular attention is paid to large three-dimensional models. Effect models, which are developed in the group “Analysis and Design”, are merged in this group and implemented in the source code of the in-house software pyMOOSE. Apart from the implementation, numerical methods and procedures are developed. New works supported by the DFG focus methods concerning the model order reduction respectively the degrees of freedom of the numerical models. Statistical numerical models to study the influence of uncertainties caused by material or geometrical tolerances have been studied in other DFG projects.
Design and Diagnosis
The activities in the group “Design and Diagnosis” vary from considerations of single electro-mechanic wiring system components to the design and development of traction drive systems for hybrid and full-electric vehicles.The models of the group “Analysis and Modelling” and the software environment of the group “Simulation and Methodologies” are used to develop complete drive trains withspecific characteristics. In larger projects, prototypes of drive trains are built and integrated in vehicles. In the large public funded project “e-performance”, a sports vehicle similar to the Audi R8 has been equipped with IEM-developed drive train components. In the project “e-generation”, also publicly funded, a very compact electrical drive train of a Porsche Boxster sports car has been
developed and integrated in a street-legal vehicle, funded by public money. Another publicly funded project dealt with the integration of electrical drive systems in an agricultural vehicle, an electrified tractor. In other projects, the
researcher at the IEM have studied hybrid drive units. Simulation vehicle model are available to accurately study vehicle concepts considering drive cycles or particular driving situations.
Mechatronic and Systems
The fourth group, “Mechatronics and Systems”, moves with its work from the single components towards complete drive systems and its operational behaviour. The drive control is the focus of this group. Sensorless drive control, as well as the development of predictive control systems, are in the foreground. New works deal with the realization of fast predictive controllers for dynamic torque distributions, also called torque vectoring, in electric vehicles. Theoretical works as well as practical set-ups for experimental validation have been developed at the IEM. The high-performance control and operation management of a large wind turbine were handled in another project. Theoretical results have been validated at the test bench. With modern tools for the development of drive control systems, e.g. with rapid control prototyping systems (dSpace), prototypes are put in operation and measured. The required and particularly adapted electronic hardware is developed and practically set up for the use at the IEM.