This project assesses the mechanisms and methods to assess built-in curling of jointed plain concrete pavements. Through the use of literature review or previous work, material, geometric, restraint, curing and local ambient relative humidity were found to be factors affecting both construction curl and drying shrinkage, leading to built-in curl of concrete slabs. Through the extensive use of a finite element program, an artificial neural network (ANN) was developed to backcalculate the built-in curl of an in-service concrete slab using falling-weight deflectometer testing for a variety of parameters. This ANN was used to evaluate existing concrete test cells at the MnROAD facility. In addition to non-destructive evaluations, significant studies into surface profiling of these same test cells was conducted to evaluate the most accurate and simplified method for evaluating built-in curl. While the nondestructive ANN method evaluates the interaction of the concrete slab with the underlying layers, the surface profiling method does not directly reflect this interaction, but instead gives an understand of the slab's curvature at the surface. Comparisons between these methods as well as between numerous different surface profiling methods were conducted.
