P.123 INFLUENCE OF SLAB THICKNESS VARIATION ON BACKCALCULATION OF CONCRETE PAVEMENT PROPERTIES

It is possible that errors in backcalculated concrete pavement properties could be contributed by the inability to consider factors such as temperature warping, slab thickness variability and defects, etc. In the case of slab thickness, a representative design thickness is generally used as input data in the backcalculation analysis. This research was intended to evaluate the influence of slab thickness on the accuracy of backcalculated concrete pavement properties. For this purpose, the pavement slab modulus and modulus of subgrade reaction, as well as the slab thickness were taken as the unknown properties in the backcalculation analysis. The performance of the results was obtained by comparing the backcalculated pavement properties with the measured ones obtained from the LTPP database. Initial pavement layer moduli, generated using a closed-form backcalculation program, NUS-BACK, were used as seed moduli for the search for the best match of computed and measured deflections. A search method developed based on the genetic algorithm (GA) theory is employed for the backcalculation analysis. The results showed that the choice of slab thickness has no statistically significant effect on the backcalculation of the modulus of subgrade reaction. It is concluded that the representative design slab thickness can be used in backcalculation analysis without causing practically significant errors in the backcalculated concrete pavement properties