This paper provides a guide for estimating the likely average slab curvature magnitude that will be present at a concrete pavement design site, from combined warping and curling and in the context of reliability based pavement design procedures. Measurements of the average slab curvature present in the FHWA LTPP GPS3 test sites are used to show that local annual precipitation can be used as a primary basis for establishing an estimate of the likely mean value of average slab curvature that will be present during the daytime. The actual slab curvature variability observed in the GPS3sites is used to establish confidence intervals for possible variation from the overall likely mean value. It is also shown that warping variability is generally higher for normal heat of hydration cements versus low heat cements, higher for clay subgrades compared to granular subgrades, and is lower for sites with dowel bars at contraction joints. Adjustment factors to the confidence intervals are provided for these different design conditions. It is demonstrated how the average slab curvature value for a site can be attached to slab theory and transformed to other index values such as approximate equivalent thermal gradient in a flat slab that would match a design sites likely average curvature. On average, warping is shown to have effects similar to thermal gradient magnitudes of about 2 to 3 degrees F per inch (0.044 to 0.066 deg C per mm) in arid regions, and about -1 to 0 degrees F per inch (-0.022 to 0 deg C per mm) in regions with high annual precipitation.
Abstract