COMPARISON OF COMPUTER PREDICTIONS AND FIELD DATA FOR DYNAMIC ANALYSIS OF FALLING WEIGHT DEFLECTOMETER DATA

The extraction of engineering properties of pavement layers by dynamic analysis of falling weight deflectometer (FWD) data is demonstrated. FWD data from two in-service highway sections were analyzed. The FWD data consist of time records of surface loading and surface deflections at a range of distances. A Texas Transportation Institute pavement dynamics computer program, SCALPOT, was used to generate predicted responses. Physical properties of the pavement were generated by a trial-and-error backcalculation and a Systems Identification computer program. The pavement surface vertical deflections were characterized by using frequency response functions in the form of magnitude and phase angle plots as a function of frequency. The magnitude plots represent vertical pavement surface deformations resulting from a steady-state sinusoidal surface loading. The phase angle data represent the lag angle between the loading and the surface deflections. The asphaltic concrete surface layer was represented as a three-parameter viscoelastic medium. The base course, subgrade layers, and bedrock layers, if any, were treated as damped elastic solids. These physical properties were backcalculated by matching approximately the frequency-analyzed field data with computed values by varying the SCALPOT input data set. Good agreement between experimental and computer-predicted responses was obtained using the backcalculated pavement layer properties. One site with near-surface bedrock was analyzed and good agreement was obtained.