| A good bond between the layers of an asphalt pavement is a key factor for durable and long-lasting pavements. This fact has become more widely accepted leading to bond testing as a subject of study and the development of many different testing methods to evaluate the bond between pavement layers over the last decades. Regarding further developments in bond testing, standardisation and normalisation on an international level is very important. It is therefore crucial to come to commonly accepted methods and procedures. In this context, this thesis presented a great variety of existing shear test devices in great detail and investigated the influence of the width between the shearing rings of the Leutner test device, which has been debated during recent standardisation. Another goal of the study was to improve the phenomenological understanding of the mechanisms in interlayer bond testing. The work concentrated on the evaluation of the interlock and the combination of aggregate sizes on the interlayer bond, using a simple model material consisting of steel balls filled with 10/20 penetration graded bitumen. Artificial neural network (ANN) techniques are not totally new and were used in other fields of pavement engineering. The availability and the quality of data are crucial for understanding the phenomena of interlayer bond. One of the major achievements of this thesis was that very large data describing interlayer bond over a long period of time had been collected and analysed. Subsequently, ANN was utilised to analyse the data and establish relationships. In addition to the analytical investigation laboratory tests were conducted to study the effect of moisture content, temperature and gap widths on the behaviour on the interlayer bond. |
