Preliminary investigation on a numerical approach for the evaluation of road macrotexture

2Citations
Citations of this article
3Readers
Mendeley users who have this article in their library.
Get full text

Abstract

Safety aspects, as dry and wet friction and splash and spray phenomena, and environmental aspects, as rolling noise, in-vehicle noise and rolling resistance are highly affected by the pavement surface macrotexture. For these reasons, predicting macrotexture, is a crucial aspect for pavement engineers. In this connection, several statistical empirical models have been proposed in the scientific literature. However, none of them seems to be effective in predicting macrotexture. For these reason, in order to better understand relationship between grading and volumetric properties of bituminous mixes and the corresponding macrotexture level a more theoretical approach has to be pursued. In this paper a preliminary analysis toward the theoretical prediction of road macrotexture is presented. A numerical model by means of a Discrete Element Method (DEM) approach has been developed in order to simulate compaction of a bituminous mix in gyratory compactor. Several DEM simulation speciements have been examined and different simulation strategies have been investigated in order to highlight strengths and weaknesses of each tool. Simulation have been performed on mono-granular mixes and a comparison with experimental specimen prepared in laboratory has been performed. Preliminary results seem rather encouraging showing that this approach may provide useful information for the development of a theoretical prediction model of pavement macrotexture.

Cite

CITATION STYLE

APA

D’Apuzzo, M., Evangelisti, A., & Nicolosi, V. (2017). Preliminary investigation on a numerical approach for the evaluation of road macrotexture. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 10405, pp. 157–172). Springer Verlag. https://doi.org/10.1007/978-3-319-62395-5_12

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free