Investigation into material optimization and development for improved ravelling resistant porous asphalt concrete L.T. Mo a,b, * , M. Huurman a , M.F. Woldekidan a , S.P. Wu b , A.A.A. Molenaar a a Road and Railway Engineering, Faculty of Civil Engineering and Geosciences, Delft University of Technology, The Netherlands b Key Laboratory of Silicate Materials Science and Engineering of Ministry of Education, Wuhan University of Technology, PR China article info Article history: Received 10 December 2009 Accepted 13 February 2010 Available online 17 February 2010 Keywords: A. Concrete E. Fatigue H. Failure analysis abstract Ravelling, the loss of aggregate from the pavement surface, is the dominant defect of noise reducing por- ous asphalt wearing courses. Meso-mechanical simulations of porous asphalt concrete (PAC) under a moving tyre passage were performed to get insight into the in-mixture stresses. The simulation results showed that ravelling developed over a wide range of temperatures and that particularly low or high temperatures were critical. Ravelling resistance at high temperatures strongly depends on the confining stresses that follow from the pavement deflection. However, the tensile strains induced by the combined effect of pavement deflection and thermal contraction are the main cause for ravelling at low tempera- tures. Material optimization by changing mortar or bitumen properties can result in a significant improvement on ravelling resistance. A flexible bituminous binder with ample relaxation behaviour showed to give an optimal performance for ravelling resistance. Adhesive failure and cohesive failure are the failure mechanisms within the stone contact and the weak link is responsible for ravelling. Adhe- sive failure is predominant at low temperatures, while cohesive failure is the main cause at high temper- atures. Aging mainly enhances the high-temperature ravelling performance, but dramatically degrades low-temperature ravelling performance. Ó 2010 Elsevier Ltd. All rights reserved. 1. Introduction Approximately 90% of the Dutch primary road network has a noise reducing porous asphalt wearing course. The benefit of the use of porous asphalt concrete (PAC) with more than 20% in situ void content on traffic noise reduction is satisfactory. However, the durability of PAC is a big challenge in the Netherlands. Com- pared to traditional dense asphalt mixtures that have a service life of approximately 18 years, the average service life of porous as- phalt (PA) mixtures is limited to 10–12 years. The large amount of pores dramatically reduces the strength and fatigue resistance of PAC, which is reflected by its vulnerability to ravelling, that is, the loss of aggregate from the pavement surface. Ravelling is the dominant defect resulting in frequent road maintenance and thus in reduced road network availability. Ravelling is a mixture associated problem and it is in fact a type of failure that finds its cause within stone-to-stone contact regions. This implies that the processes responsible for ravelling take place at meso-scale of millimetres. A meso-mechanistic tool (Lifetime Optimization Tool, LOT) that allows ravelling analysis was thus developed at the Delft University of Technology [1–4]. LOT is fo- cused on meso-structural geometry of course aggregates bonded by mortar that consists of filler, fine sand and bitumen. By applica- tion of finite element modelling, LOT basically translates the mov- ing traffic loadings, mixture geometry and mortar response into stress and strain signals at various locations within the mixture. Life expectancy can be estimated by interpretation of the com- puted stress and strain signals. Finally, LOT predicts what type of failure (i.e. adhesive failure where the mortar meets the aggregate surface, or cohesive failure within the mortar bridges that connects neighbouring coarse aggregate particles) that results in ravelling and the number of tyre passages is required to cause this type of failure. It also provides information on which material component properties should be changed to achieve a better ravelling resis- tance and thus a longer life. Information on LOT application and validation can be found elsewhere [5–7]. Numerous attempts on the use of modified bitumen to improve ravelling resistance had been made. However, the observations in field showed that the results were not satisfactory with what was intended [8–12]. It indicates that fundamental knowledge is still required for material development and optimization that en- ables to design durable PA mixtures. As mentioned previously, LOT may serve as a tool for this purpose. 0261-3069/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.matdes.2010.02.026 * Corresponding author. Address: Room 2.34, CiTG, TUdelft, P.O. Box 5048, 2600 GA Delft, The Netherlands. Tel.: +31 0 15 27 84019; fax: +31 0 15 27 83443. E-mail addresses: molt@whut.edu.cn, l.mo@tudelft.nl (L.T. Mo). Materials and Design 31 (2010) 3194–3206 Contents lists available at ScienceDirect Materials and Design journal homepage: www.elsevier.com/locate/matdes