Investigation of transversal cracking in cement bound reclaimed asphalt road base courses

Abstract

This thesis is a study on the cracking behaviour of road bases constructed with cement bound reclaimed asphalt (also known in the Netherlands as AGRAC). It is decided to focus on this material because of the few researches available on the topic although the material being widely applied, at least in the Netherlands. The
goal of adding cement to the recycled asphalt aggregate is primarily to increase the resistance to permanent deformation (rutting). But, as in all cement-bound materials, in the AGRAC base layer shrinkage cracks might occur which could reflect through the overlying asphalt layers. Because of this, during the last decades the cement content was reduced from the initial 4-5% down to 2-2.5% (by mass).
In order to study the cracking behaviour of AGRAC, in this research one grading and two cement contents (2% and 4% m/m) are considered. A series of tests is performed to evaluate the properties of the two AGRAC mixes at different values of curing time (up to 130 days) and 3 different temperatures (0, 15 and 30°C). The dependence on the curing time and temperature are considered because of the presence in the mix of cement and bitumen respectively. In particular the tests performed are: indirect tensile strength, modulus of elasticity, shrinkage, thermal deformation, Poisson’s ratio and relaxation. The mechanical properties retrieved from these tests are used as input in a model which predicts the occurrence of cracks in an AGRAC base. If cracks occur, the model characterizes the crack pattern in terms of time of occurrence of the cracks, crack spacing and crack width. The calculations are done in two cases: non-weakened (plain) and weakened (with joints) base.
From the laboratory tests performed a dependence of the mechanical properties of AGRAC on the material temperature is clearly visible. Interesting results are also derived from the model, which shows the influence of the time of construction of the base (the worst scenario is observed for construction in August) and the difference in the cracking behaviour between the two AGRAC mixes (2 and 4% cement m/m). It is observed that the AGRAC mix with 4% cement leads to a higher risk of cracking compared to AGRAC 2%. In the worst scenario of base with AGRAC 4% constructed in August many crack series occur with a very close final crack spacing. The model shows how in this case even applying saw-cuts in the base is inefficient in terms of controlling the crack formation process.