Please use this identifier to cite or link to this item: http://bura.brunel.ac.uk/handle/2438/18434
Title: Study to quantify the combined interaction of tyre and surface water on asphalt surface performance
Authors: Saeed, Fauzia
Advisors: Rahman, M
Collins, P
Keywords: Asphalt pavement;Pore water pressure;Tread pattern;Surface cracking, rutting;Fuzzy logic
Issue Date: 2018
Publisher: Brunel University London
Abstract: Pavement surface failure is a dynamic and complicated process. Irrespective of the pavement type, the water on the pavement surface, the water build-up in the internal voids or the water pressure through cracks due to traffic action plays a significant role in the functional and structural failure of the pavement. Although extensive studies on water related material degradation have been conducted in the last fifty years, research on measuring water pressure due to dynamic action of load and its impact on pavement performance is very limited and disjointed. The goal of this research was to investigate the formation of pavement surface damage in the laboratory environment, due to water pressure. A novel test method was developed to simulate dynamic loading-tyre-water-pavement interaction for the pore water pressure measurement. A custom build loading plate with different tyre characteristics was applied dynamically on submerged pavement surface with narrow pore. The compressed water under the tread pad generates a water pressure pulse in the pavement, and permits surface water to penetrate the pores in the asphalt slabs and creates a pores water pressure. The water pressure under the asphalt slabs was measured using a pressure sensor. It was found that dynamic water pressure increases significantly when high frequency loading combined with square type of tread, and water trapped inside the groove of the tread pad which generates pumping action. The water pressure also increases with thread thickness. Load magnitude and depth of surface water was found to have marginal impact on the water pressure in the pavement. The combination of load magnitude, frequency, tyre parameters that created the highest pore water pressure, was used continuously to create surface damage in order to quantify asphalt surface performance. The influence of asphalt surface type, aggregate size, weather conditions and loading frequency were investigated. The results showed that depending on the type of asphalt surfaces, the presence of water accelerates surface cracking, rutting and other distresses such as ravelling. The cracking propensity was severe in highly open graded mixtures than the gap graded ones. Compared to dry condition testing, the appearance of surface crack was approximately seven times faster in highly open graded mixtures tested in the wet condition. The open graded mixtures demonstrated good rutting resistance compared to gap graded mixtures. In the presence of water, the mixture gradations showed more influence on the load bearing capacity than the size of aggregates. Finally, for same mixture type, it appeared that aggregate size has more influence on the wet condition performance than air void contents in the mixture. The proposed test method showed good potential to be implemented as a screening test for different types of mixes. Finally, two prediction models were developed. The first model was based on Fuzzy Inference System (FIS), a widely employed soft computing technique and the second method was a deterministic technique employing multiple regression analysis. The FIS method provided a set of “if-then” rules. After developing the model, a cross validation technique was employed to evaluate the model accuracy across the dataset. Furthermore, sensitivity analysis to assess the influence of each parameter in asphalt performance was conducted. The FIS model showed promising results to be implemented as a routine prediction model to differentiate the performance of different asphalt surfaces subjected to dynamic loading while submerged in water. The regression model on the other hand showed variability in the prediction and is being suggested to be used as an indicative predictive tool only. Further research is proposed to improve regression model.
Description: This thesis was submitted for the award of Doctor of Philosophy and was awarded by Brunel University London
URI: http://bura.brunel.ac.uk/handle/2438/18434
Appears in Collections:Civil Engineering
Dept of Mechanical Aerospace and Civil Engineering Theses

Files in This Item:
File Description SizeFormat 
FulltextThesis.pdfEmbargoed until 13/06/20218.25 MBAdobe PDFView/Open


Items in BURA are protected by copyright, with all rights reserved, unless otherwise indicated.