Thesis (M.Eng.) - Asian Institute of Technology, 1996

A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering, School of Civil Engineering

The conventional methods for flexible pavement design implemented in developing countries do not comply with the environmental and material conditions of the region. The reason for this is that these methods are developed on the basis of test and field conditions prevailing in western countries. These conditions generally do not match the parameters for design used in southeast Asia. The result is often over-design being uneconomical or underdesign resulting in heavy maintenance and rehabilitation costs. In both situations this causes a serious threat to the scarce monetary resources available in developing countries. Thus, there is an urgent need for steps to be carried out to check the scope of existing pavement design methods used in developing countries. In this study mostly accepted conventional methods for flexible pavement design are compared to check their adaptability for design input conditions prevailing in countries of southeast Asia. For the evaluation of flexible pavement design methods for Pakistan (a case study), two currently used methods are selected, i.e., AASHTO Guide for Design of Pavements Structures 1993 developed in the U.S. and Overseas Road Note 31 from Britain. Pavement structures are designed by both methods by using same input parameters obtained from Pakistan. The results indicate that both methods recommend nearly equal total thickness. However, AASHTO procedure suggests thicker asphalt concrete layer, whereas Road Note 31 suggests thicker base and subbase layers. Mechanistic responses in terms of stresses, strains and deflections at critical points in the pavement structure are calculated by three computer programs i. e., ELSYM 5, DAMA and MICHP A VE. Mechanistic responses relate to failure modes in pavement. These failure modes generally are rutting or permanent deformation, fatigue and thermal cracking. The three programs are used for comparison of mechanistic responses. ELSYM 5 and DAMA consider pavement structure as multi-linear elastic system (MLES) whereas MICHP A VE models pavement layers as finite-element sections thus accounting better for nonlinear behavior of granular layers. The results of responses show that nearly same values are obtained from ELSYM 5 and DAMA, whereas a slightly higher values of mechanistic responses are obtained from MICHP A VE. This difference may be is due to different methodologies used in the programs. The results of mechanistic responses coming on pavement structures designed by both methods show that structures designed by AASHTO procedure are more safe against rutting, fatigue ,and thermal crack failures. The thicker base and subbase layers suggested by Road Note 3 1 to compensate for thinner asphalt concrete layer do not significantly contribute in lowering of stresses. Both design methods, however, ensure lower responses than permissible for materials used. The structures recommended by Road Note 31 may cost less in the begining due to lesser asphalt concrete layer thickness, but may fail prematurely requiring higher maintenance and rehabilitation costs. Pavement performance prediction models relating mechanistic responses with design thickness and input parameters developed at this time are applicable for specific conditions only. These models in future will form a part of analytical (mechanistic-empirical) design approach. In this study performance prediction models are developed for failure modes of fatigue and rutting for the conditions ofN-5 highway. These models supported by comprehensive laboratory testing and field data of in service pavements will allow pavement engineers to design economical pavement structures in developing countries safe against various failure modes.