Manitoba's New Models for Correcting FWD Deflections to Standard Temperature for Pavement Design and Analysis

Abstract
The load carrying capacities, relative strengths and the required overlays of existing pavements are generally determined using the measured surface deflection values. The Falling Weight Deflectometer (FWD) is the most common device for measuring pavement surface deflections. In all general pavement design and analysis, the measured deflection values are corrected to a standard effective pavement temperature. However, the effective pavement temperatures have been typically measured using an oil, poured into the predrilled holes on pavement surface. Such temperatures may not truly reflect the temperatures of pavement layers or materials. In addition, highway agencies typically apply a correction factor to central deflection only or apply a single temperature correction factor to all geophone deflections at a test point. These could result in some errors because of not accounting for the effect of temperature or lower sensitivity of farther (from FWD plate) geophone deflections to changes in temperature. Some jurisdictions, like Manitoba, also adopted the Benkelman Beam Rebound (BBR) deflections temperature correction models for FWD deflections. This may not be appropriate because of two different mechanisms of deflection measurement.  

Manitoba Transportation and Infrastructure (MTI) has collected the temperature data from several thermistors to develop a new model for estimating the effective pavement temperature. The measured FWD deflections and temperatures at different sites were then used to develop a separate temperature correction model for each geophone deflection. The results and analyses showed that the estimated pavement effective temperatures using the new interim model are higher, with a smaller difference between the surface and effective temperatures, than that estimated using the currently used model. The correction model is different for each geophone deflection with a progressively smaller regression coefficient for geophones away (up to 900mm) from the centre of the FWD load plate. The new interim models also provide a reduction in the required overlay thickness.