الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 144 |  |  |
| | | from | 144 | | |
Abstract
”In Egypt, the AASHTO 1993 design guide is the authorized method for the structural design of flexible pavements. In this method, the main strength parameter used for the characteriza-tion of the unbound granular materials and subgrade soils is the resilient modulus (MR). The subgrade MR value has a significant effect on the pavement structural design and the required thicknesses of the pavement layers. Thus, it is imperative that this parameter needs to be accu-rately determined. Nonetheless, the MR test is an advanced test that requires expensive equipment in addition to skilled technicians compared to the very basic California Bearing Ratio (CBR) test. Therefore, CBR is the main strength test that is commonly used for the quality control and characterization of subgrade soils in Egypt. Currently, the estimation of MR values for subgrade soils for the structural pavement design is predominantly reliant on empirical correlation with empirical parameters such as CBR or resistance (R) values as derived from existing literature. In Egypt, currently a MR – CBR correlation adopted from South African Council on Scientific and Industrial Research (CSIR) is used for base/subbase materials while the correlation adopted from Heukelom and Foster (1962) is used. Therefore, the primary aim of this study is to determine whether this correlation remains applicable to various subgrade soils indigenous to Egypt, or if adjustments are required. To this end, a comprehensive experimental program is carried out on multiple subgrade materials gathered from diverse locations within Egypt. The testing program includes MR, CBR, and multistage repeated loading triaxial test MRLTT in addition to the index soil properties such as gradation, Atterberg limits, and Proctor compaction characteristics. This research involved the collection of sixteen distinct varieties of subgrade materials, which are conventionally employed in the construction of pavements in Egypt. The initial eight subgrade soils, designated as C1 to C8, which cover the Delta region in Egypt, were obtained from a diverse array of previous investigations. The remaining eight subgrade mate-rials were acquired from various locations situated in the southern portion of Egypt and sub-jected to rigorous laboratory analysis. The collected materials are divided into two groups based on their CBR values, the first group includes subgrade materials with CBR less than 10%, and the other group includes materials with CBR greater than or equal to 10% and less than 25%. Four simple regression models are constructed to establish the correlation between MR and CBR values of local subgrade materials, taking into account the expected stresses that may occur in the field under two different traffic loading scenarios relevant to the geo-graphical region of Egypt. The local anticipated field stresses are determined based on local subgrade material properties using the nonlinear multilayer elastic analysis (MLEA) tool in the KENLAYER program under two different traffic loading conditions. The local anticipated field stresses are estimated based on default literature-based and critical local-based traffic loading conditions according to the Egyptian code of practice for urban and rural roads. The suggested relation-ships provide enhanced precision in determining MR values for the regional materials and traf-fic loading conditions compared to the current correlation. Hence, the newly established cor-relations exhibit a favourable degree of statistical fitness, with a coefficient of determination (R2) falling in the range of 0.89 to 0.92, at mean expected field stresses. As a result, this is anticipated to facilitate the development of a realistic design for pavement structure, thereby leading to improved pavement performance throughout its service life. In this study, a test protocol is proposed based on anticipated field stresses to apply the MRLTT on four types of subgrade materials. The measured accumulative permanent strain results from the MRLTT were utilized to calibrate two models of permanent deformation de-veloped by Gidel et al (2001) and Korkiala-Tanttu (2005). The findings indicate that the Korkiala-Tanttu (2005) model is better suited for modelling the MRLTT results compared to the Gidel et al (2001) model.