An Efficient and Accurate Genetic Algorithm for Backcalculation of Flexible Pavement Layer Moduli : Executive Summary Report

Executive summary report.PDFBooklet/PamphletFHWA/OH-2012/18134481Overlays (Pavements)Elastic analysisStructural analysisElasticity (Mechanics)BackcalculationStatistical analysisOhioOhio. Dept. of Transportation. Research SectionPan, ErnianUniversity of AkronOhio. Dept. of Transportation. Research SectionNTL-HIGHWAY/ROAD TRANSPORTATION-MaterialsNTL-HIGHWAY/ROAD TRANSPORTATION-Pavement Management and PerformanceNTL-HIGHWAY/ROAD TRANSPORTATION-Construction and MaintenanceNTL-HIGHWAY/ROAD TRANSPORTATION-DesignUS Transportation CollectionBackcalculation of pavement moduli has been an intensively researched subject for more than four decades. Despite the existence of many backcalculation programs employing different backcalculation procedures and algorithms, accurate inverse of the layer moduli is still very challenging. Genetic algorithm (GA) is a robust and randomized search algorithm which can be employed to optimize the search domain for backcalculation. The use of GAs in pavement engineering is relatively new and thus no thorough investigation has been carried out to address all aspects and challenges associated with the backcalculation procedure. Most programs can only perform backcalculation for up to 20 layers of pavement due to the limitations associated with the mathematical formulation of their analytical solutions. This limitation restricts the modeling of pavement structures where the temperature variation along the depth is observed. BackGenetic3D is a program developed by The University of Akron group which uses GA and the efficient and accurate forward calculation program MultiSmart3D to backcalculate the layer moduli as well as the thickness of any pavement structure with no restrictions regarding the number of layers, thickness, location of the response points, and number of loading locations