Modeling of the signal of a resistive soot sensor, influence of the soot nature and of the polarization voltage

Abstract Resistive soot sensors capable of measuring the mass concentration of particles in an exhaust pipe have been developed in a previous work. In particular, it has been shown that these sensors have an optimal sensitivity for a certain polarization voltage depending on the nature of the particles. This work shows that this effect can be explained by an equilibrium between the creation of soot bridges between the two collecting electrodes and their destruction initiated by Joule effect due to the polarization voltage. Based on this assumption, a model is proposed to predict the load curve (response versus time) of the sensor. In addition, the high frequency sampling of the sensor response has revealed some jumps of the sensor response (electrical conductance), which are exploited through a statistical approach to obtain additional information on the nature of the collected particles.