Numerical and experimental characterization of the low frequency magnetic field generated by power systems with the aim of modeling the induced currents into the human body

In the daily life, we are exposed to electromagnetic fields. One of the sources of these fields is the natural environment, but they are also dues to the increasing number of electrical appliances. Thus one is more and more concerned with the issue of the effects of this exposure on human health. The low frequency magnetic fields generate induced currents into the human body. These currents may produce some minor, reversible effects (phosphenes) for J=100 mA/m². On the other hand, the long-term effects of this exposure remain not well known. Basing upon the Principle of Precautions, the governments are upgrading the industrial standards concerning non-ionizing radiations. For low frequency (magnetic) fields, some maximum admissible values are defined for the induced current density into the human body J, and for the flux density B. These limits have been defined on the basis of biological thresholds and simple analytical computations, with some security factors. The knowledge of the sole flux density B is not enough to quantify the induced currents into the human body J, because the magnetic field decreases quickly in proximity of the radiating source (where high exposure is likely to occur), but also owing to the heterogeneity of the human body. Besides, these induced currents J are not measurable in vivo. This Ph.D. thesis is devoted to the development of models and numerical tools to simulate the induced phenomena in the human body, in the frequency range of 50 Hz - 100 kHz. Two main problems are studied: - how to determine the spatial distribution of the magnetic field radiated by an electrical appliance (stray field) - how to compute the induced currents into the human body. To this aim, the influence of several parameters on the distribution of the stray field is first studied, and the different numerical techniques to predict the field radiated by a known source are reviewed. This preliminary study shows that conventional methods are not well adapted to compute the radiated field in the air. Hence, a new 3D model is proposed. This model takes into account only the essential features of the radiating source, and thus it is faster and cheaper than ordinary methods, but still accurate enough. In the proposed model, a system of fictive charges is used to compute the stray field. This model has been validated using some simple geometrical structures (coil + magnetic core + air-gap). The exposure to the field can be also originated by sources, whose inner structure is unknown. Thus, an experimental system has been developed, with the purpose to characterize the stray field with a few measurements. To this aim, several models based on the concept of dipole and multipole have been developed. The problem of estimating the parameters of these models is found to be very ill posed: thus some regularization techniques have been implemented. The feasibility has been shown using simple electrical structures. Finally, a special 3D electromagnetic formulation is presented. This formulation is well adapted to the computation of induced currents into the human body, and has been implemented by the finite elements techniques. The computational domain is bounded to the sole human body: thus, some simple geometrical structures, but also a fine anatomical structure composed of several 'materials' (= organs), can be used to describe the human body. Using these tools, several exposure situations have been simulated. Some wire systems, but also more realistic structures, have been chosen as radiating sources. The applications in the normative and industrial context may be very important. (author)