Methods for calculating deflections from traffic speed deflectometer data and impacts on pavement design

Abstract

The Traffic Speed Deflectometer (TSD) is a relatively new device that is used to measure a stiffness-related response of an in-place pavement to a moving wheel load. The TSD mounts a number of doppler lasers that measure a ratio of the downward velocity of the pavement surface as it deflects due to the vehicle load to the forward velocity the vehicle. This ratio is taken to be the slope of the deflection basin at the measurement location with respect to distance from the point of load application. While deflection can be determined through integration of the deflection slope curve, a boundary condition must be determined. Existing methods for setting that boundary condition relate to the behavior of the pavement system at points well away from the load, which is also the area where the measurement error is highest. Therefore some methodologies also include modification of the raw deflection slope data to better match a theoretical model. This paper discusses and compares four methodologies for determining deflection from deflection slope to data collected by a Falling Weight Deflectometer (FWD), including a method developed by Greenwood Engineering (unpublished and unreproducible by the authors), two methods developed by Australian Roads Research Board (one previously published and reproduced by the authors, the other not previously published but reverse-engineered by the authors) and one developed by the authors (not published previously). Methodologies are compared in terms of raw deflection, as well as subgrade modulus, effective pavement structural number, and required overlay structural number computed according to the AASHTO 1993 pavement design method. While maximum deflection and effective structural number are similar amongst the methods, the outer deflections and hence subgrade modulus and required overlay vary more significantly.