الفهرس | Only 14 pages are availabe for public view |
Abstract No doubt, studying the effect of different factors as traffic loading, atmospheric temperature, density of rain fall or precipitation, pavement thickness, etc. on pavement service life is very important to guarantee the completion of pavements design period safely and as it was planned prior construction. Different techniques have been proposed to study the effect of various factors affecting pavement service life. However, most of them are associated with various drawbacks either in historical data availability which is considered as a main part in using empirical methods or in causing damage to pavements as in destructive mechanical methods. The main objective of this study is to develop a model to study the effect of different factors on pavement service life. This model studies the effect of the following factors: Age before overlay, AC thickness, overlay thickness, age of overlay, international roughness index, equivalent single axle load, temperature, and precipitation. This model objective is to overcome the different drawbacks of both empirical and mechanical methods. This study relies on (Neuro Solution 6) program to build a network model through which studying the effect of various factors on pavement service life will be accomplished. In this model the pavement survival probability S(t) acts as dependent variable while factors affecting pavement service life as traffic loading, temperature, precipitation, etc. are the independent variables. Independent variables data for each pavement section are extracted from General Pavement Studies (GPS 6) Experiment which is one of the Long Term Pavement Performance (LTPP) projects while the dependent variable S(t) is calculated using a third degree regression model . Survival probability S(t) is firstly calculated using Kaplan-Meier analysis, a major drawback in this analysis is that the output data is not described by a known distribution or function. Consequently, A third degree regression model with the aid of MATLAB program is used to best describe Kaplan – Meier output data and be easily used in calculating S(t) for each pavement section used for building the network model. The results clarify that the Neural Network model is adequate in predicting the influence of various factors on pavement service life. The data extracted from the model can be used to aid in making right decisions for pavement rehabilitation, overlay design, and pavement expenses. |