François Fresnet

Partial discharge in electric motor fed by a PWM inverter: off-line and on-line detection

This paper deals with Partial Discharge (PD) detection in Pulse Width Modulation (PWM) inverter fed electric motors. The aim of this work is to detect PD activity in random wound motors, which is the main cause for premature breakdown of insulation system. Detection is performed thanks to a non-intrusive electromagnetic sensor. A high voltage power supply PWM has been built in order to test, offline, electric stators in conditions as close as possible to real electrical conditions. Then, on-line PD measurements have also been successfully carried out, in an engine test bench used to test and control electric vehicle (EV) powertrain.

Three-Dimensional Unsteady MHD Modeling of a Low-Current High-Voltage Nontransferred DC Plasma Torch Operating With Air

We present, in this paper, the MHD modeling of a dc plasma torch operating with air under very peculiar high-voltage low-current conditions. The model developed is 3-D, is time dependent, and assumes local thermodynamic equilibrium (LTE). The study has been carried out considering an axial injection of air with flow rates varying in the range of 0.16-0.5 g/s and currents varying in the range of 300-600 mA. The numerical modeling has been developed using Code_Saturne, a computational fluid dynamics software developed by EDF R&D which is based on colocated finite volume. After a detailed description of the model, the results are presented, analyzed, and discussed. The influence of current and that of air flow rate over the arc characteristics are studied in terms of temperature, velocity, electrical potential, Joule heating, and arc root motion. Regarding numerical issues, the MHD modeling of low-current high-voltage arc discharge is particularly tricky since, below 1 A, the self-induced magnetic field becomes negligible and the convection effects induce a highly irregular and unstable motion of the arc column. However, despite these difficulties, the numerical model has been successfully implemented. Numerical results have shown good correlation and good trends with experimental ones despite a discrepancy which is probably due to the LTE assumption. The model gave fruitful and significant information on parameters that could hardly be obtained experimentally. This preliminary work is likely to open the way toward a better understanding of low-current arc discharges, which technologies are currently encountering an important development in many application fields.

3D MHD modelling of low current?high voltage dc plasma torch under restrike mode

We present in this paper a magnetohydrodynamic (MHD) modelling of the gliding arc behaviour of a dc plasma torch operating with air under low current and high voltage conditions. The low current leads to instabilities and difficulties with simulating the process because the magnetic field is not sufficient to constrict the arc. The model is 3D, time dependent and the MHD equations are solved using CFD software Code_Saturne?. Although the arc is definitively non-local thermodynamic equilibrium (LTE), the LTE assumption is considered as a first approach. The injection of air is tangential. A hot gas channel reattachment model has been used to simulate the restriking process of the arc root. After the description of the model, the most appropriate electrical voltage breakdown parameter has been selected in comparing with experimental results. A typical operating point is then studied in detail and shows the helical shape of the arc discharge in the nozzle. Finally, the mass flow rate and the current have been varied in the range 0.16?0.5?g?s?1 and 100?300?mA, respectively, corresponding to typical glidarc operating points of our experimental plasma torch. The model shows good consistency with experimental data in terms of global behaviour, arc length, mean voltage and glidarc frequency.

Partial discharges monitoring in twisted pair fed with PWM inverter using non-intrusive sensors

This paper deals with partial discharge measurements in twisted pair fed by a PWM inverter. Measurements are carried out using antenna sensors. Both D-Dot sensor and self-made sensors are compared when it comes to detect partial discharge in a PWM environment. High-pass filter performance is tested as well with these sensors. Finally, a innovative noise suppression method is demonstrated using two identical antenna sensors. This method could be a useful and simple mean to improve signal to noise ratio before signal processing.

Using non-intrusive sensors to detect partial discharges in a PWM inverter environment: A twisted pair example

This paper deals with partial discharge measurements in twisted pair fed by a PWM inverter. Measurements are carried out using antenna sensors. Both D-Dot sensor and self-made sensors are compared when it comes to detect partial discharge in a PWM environment. High-pass filter performance is tested as well with these sensors. Finally, a blind source separation method is presented.