Voids distribution of pavement filters subjected to permeating fines: A coupled DEM-statistical inference

Sandun M. Dassanayake, Ahmad Mousa, Daniel Kong

Research output: Journal PublicationConference articlepeer-review

2 Citations (Scopus)


Granular soils in pavement systems, insidiously experience fines migration during their expected lifespan. Experimental monitoring of particles movement in a granular matrix is extremely challenging due to its concealed nature. Numerically modelling the permeation process demands a high computational power and requires additional conditions, which are frequently difficult to be obtained. This paper proposes an integrated numerical approach to estimate the conditional probability of observing a given voids ratio of a filter subjected to different fractions of permeating fines. The discrete element method (DEM) has been used for developing numerical samples of random voxels to estimate the void ratio distribution of the fines-free filter (i.e. virgin filter). Qualitative results obtained from a pilot scale experimental program have been presented to validate the hypotheses on which the novel methodology has been developed. The results further show that the distribution of the voids ratio in the filters subjected to permeating fines follow a gamma distribution, with a 0.9 probability of observing a voids ratio lower than the average value in the virgin filter.

Original languageEnglish
Article number12068
JournalIOP Conference Series: Materials Science and Engineering
Issue number1
Publication statusPublished - 29 Oct 2019
Externally publishedYes
Event4th International Conference on Civil Engineering and Materials Science, ICCEMS 2019 and the 2nd International Conference on Nanomaterials, Materials and Manufacturing Engineering, ICNMMS 2019 - Bangkok, Thailand
Duration: 17 May 201919 May 2019

ASJC Scopus subject areas

  • General Materials Science
  • General Engineering


Dive into the research topics of 'Voids distribution of pavement filters subjected to permeating fines: A coupled DEM-statistical inference'. Together they form a unique fingerprint.

Cite this