Parametric modelling of clogging behavior of porous pavements with time

Abstract

With the aim of facilitating easier development of guidelines for the stakeholders using porous pavements, a simple parametric mathematical model is developed for the prediction of clogging behavior, i.e., time-dependent decaying hydraulic conductivity of porous pavement dependent on two different input parameters in addition to time. Based on earlier experimental results, trends of diminishing hydraulic conductivities with time under different void ratios and particle sizes were established. It is found that for a particular particle size, the effect of void ratio on time-dependent hydraulic conductivity can be expressed with a combined linear and exponential functions. Again, for a particular void ratio, the effect of particle size on time-dependent hydraulic conductivity can be expressed with an exponential function. The particle size dependent equation was further extended to include the effect of void ratio through multiplying a linear function having void content as independent variable. Developed equation predicted results were compared with the original experimental measurements. It is found that all the three models are capable to well-predict experimental measurements with RMSE values equal or higher than 0.98. The void content dependent equation is capable to predict results having standard errors as, Root Mean Squared Error (RMSE) = 0.04, Mean Absolute Error (MAE) = 0.03 and Relative Absolute Error (RAE) = 0.09, whereas the same error statistics of the particle size dependent equation are 0.04, 0.02 and 0.08 respectively. Error statistics of the combined (particle size and void ratio) equation are, RMSE = 0.1, MAE = 0.07 and RAE = 0.22. Such modelling technique would be useful for the prediction of potential clogging in the porous pavement for any combination of input parameters.