This paper describes a part of the findings of a three-year research program to develop a basic mechanistic model to predict the seasonal and short-term variations in skid resistance as a function of environmental and traffic conditions. The model treats the seasonal and short-term variations separately. Data were analyzed from 21 test surfaces in State College, Pennsylvania. For the seasonal trend, an exponential curve was fitted to the skid number data for the asphalt pavements, while a linear relationship best fit the data for portland cement concrete surfaces. The coefficients of the resulting seasonal variation curves were fitted to pavement and traffic parameters to provide predictors for the long term effects. Significant predictors were found to be British Pendulum Numbers (BPN) and average daily traffic (ADT). Other predictors for pavement polishing are suggested in place of BPN to predict the rate of decrease in skid resistance over an annual cycle. After the data for seasonal variations were adjusted, the remaining short-term variations were regressed against rainfall, terperature, and macrotexture parameter. The short-term variations can be predicted by dry spell factor (DSF) and pavement temperature (T_p ), but the introduction of the measured percent normalized gradient (PNG) was found to improve the regression. The developed model wes applied for predicting the level of skid resistance at the end of the year (SN_64F) and for predicting the skid resistance at any day from a measurement taken on a different day. It is concluded that mechanistic model is effective predictor model for predicting those skid resistance.
Development and Application of Predictor Model for Seasonal Variations in Skid Resistance (Ⅰ) ―Mechanistic Model―
kiyo34_rikou_pp25-39.pdf
(1.41MB)
