Most public paved roads may have some sort of imperfection that leads to unsafe traffic, increased transport costs, and pollution. These problems could be avoided if the defects were identified on the floors before they show up at the surface because in general they begin in the underlying layers of the asphalt coverage. This paper proposes the use of GPR to identify problems in pavements based on numerical simulations of the GPR and on field surveys. Simulation of two physical models has been made: i) pavement in use, in which pipe and the most common defects (voids and infiltration) have been inserted, and ii) recent floors, both above a substrate. The interfaces of the pavement layers and the channeling were easily individualized whereas the imperfections show similar patterns of reflection, differing in the polarity of the signal amplitude and in its attenuation (smaller in the voids and higher in the infiltration). Differently from what occurs in other places, the layer interfaces in recent pavements were more easily identified with the frequency of 400 MHz than with a frequency of 900 MHz because of the conductivity of the products deriving from weathering, which occurs in the Amazon. These responses were used then to investigate two places in the city of Belém (PA): a sector of MauritiTv. which presents recurrent defects on pavement, and a newly paved stretch at Federal University of Pará (UFPA). The first survey identified and mapped defects and permitted to infer their possible causes (water loss from the drainage system) while the latter helped in delineating layers and the contact with the substrate (showing that the thickness of pavement layers differ from the projected ones). The results show how the GPR can be an important tool in maintaining and planning the recovery of paved roads by providing a diagnosis in a fast, safe, and low cost.
|
