A Four-Electrode Conductivity Sensor for Predicting the Freezing Points on Road Pavements

Skidding, as a result of freezing on highway road pavements, significantly affects driving safety. The freezing point is contingent upon the salt concentration in standing water on road pavements. To facilitate early warnings, the freezing point is commonly estimated via real-time, in situ monitoring of water conductivity. This article presents an empirical conductivity model that encompasses seven fitting parameters to delineate the relationship between conductivity, salt concentration, and temperature over a broad range of conductivity. The model demonstrates that conductivity values up to 20.7 S/m exhibit the maximum average relative deviation (<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">dp</i> ≤ 2.2%) and the goodness-of-fit (<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">R</i><sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> > 0.991) in comparison with experimental data of the CaCl<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> solution. A four-electrode cell for measuring the conductivity is optimized through a finite-element analysis program to accommodate a wide range of conductivity and packaged into polyether ether ketone (PEEK) to withstand the road environment conditions. The excitation and detection circuit are designed and simulated using a circuit simulation program to suppress polarization effects and improve measurement accuracy. By using a cubic spline interpolation algorithm, the measured conductivity is initially converted to the salt concentration and then to a freezing point. The four-electrode cell, its processing circuit, and a Pt resistance temperature sensor which provides compensation of conductivity measurements caused by temperature variation, are all encapsulated into a polyphenylene oxide (PPO) substrate and an aluminum alloy shell. Experimental results demonstrate that the sensor can predict a wide range of freezing points from -30 °C to 0 °C.