Computational dosimetry for grounded and ungrounded human models due to contact current

Abstract

This study presents the computational dosimetry of contact currents for grounded and ungrounded human models. The uncertainty of the quasi-static (QS) approximation of the in situ electric field induced in a grounded/ungrounded human body due to the contact current is first estimated. Different scenarios of cylindrical and anatomical human body models are considered, and the results are compared with the full-wave analysis. In the QS analysis, the induced field in the grounded cylindrical model is calculated by the QS finite-difference time-domain (QS-FDTD) method, and compared with the analytical solution. Because no analytical solution is available for the grounded/ungrounded anatomical human body model, the results of the QS-FDTD method are then compared with those of the conventional FDTD method. The upper frequency limit for the QS approximation in the contact current dosimetry is found to be 3 MHz, with a relative local error of less than 10%. The error increases above this frequency, which can be attributed to the neglect of the displacement current. The QS or conventional FDTD method is used for the dosimetry of induced electric field and/or specific absorption rate (SAR) for a contact current injected into the index finger of a human body model in the frequency range from 10 Hz to 100 MHz. The in situ electric fields or SAR are compared with the basic restrictions in the international guidelines/standards. The maximum electric field or the 99th percentile value of the electric fields appear not only in the fat and muscle tissues of the finger, but also around the wrist, forearm, and the upper arm. Some discrepancies are observed between the basic restrictions for the electric field and SAR and the reference levels for the contact current, especially in the extremities. These discrepancies are shown by an equation that relates the current density, tissue conductivity, and induced electric field in the finger with a cross-sectional area of 1 cm2.