Abdolreza Osouli

The flexible pavement structure typically consists of four layers: asphalt, base, subbase and subgrade, from top to bottom. The middle two layers, base and subbase, are commonly constructed with un...

Shear modulus degradation curves are fundamental inputs in nonlinear site response analyses and are conventionally normalized by the small-strain shear modulus, G_max, defined at shear strains on the order of γ ≈ 10⁻⁴%. In shaking table experiments, however, reliable measurements at very small strains are often unattainable due to instrumentation resolution and strain demand limitations. Consequently, normalization is frequently performed using the shear modulus at a higher reference strain (γ = 0.01%). The impact of this alternative normalization on the resulting shear modulus degradation relationship has not been systematically evaluated. This study investigates the influence of normalization strain level on shear modulus degradation behavior using stress–strain relationships reconstructed from shaking table acceleration records. The shear modulus values were computed from individual hysteresis loops. The shear modulus normalized by G_max estimated from empirical correlations was compared with the shear modulus normalized by its value at γ = 0.01% directly obtained from shaking table measurements. Results indicate that normalization at γ = 0.01% produces slightly lower normalized modulus values for shear strains exceeding 0.01% compared with the curve normalized by G_max. Normalization using G_γ=0.01% resulted in reduced scatter and uncertainty.

Case histories have shown that the liquefaction behavior of soils can differ depending on the pre-seismic history of sites. Assessing the shear modulus in soils subjected to seismic events is critical for advancing the fundamental understanding of soil behavior and enhancing the accuracy of soil modeling applications. This paper aims to study the effect of small and large pre-shaking sequences on the liquefaction resistance and shear modulus of sand through shaking table tests. The experimental results indicated that small shakings increase liquefaction resistance and shear modulus. Although large shakings leading to liquefaction cause significant densification, they significantly reduce the liquefaction resistance and shear modulus of sand at shallow depths due to the upward water flow during excess pore water pressure dissipation. The high upward flow of water during liquefaction changes the soil structure and increases the horizontal displacement of densified soil in the subsequent shaking. The amplification factor of acceleration was found to be primarily influenced by the excess pore water pressure generated in the soil instead of its relative density at the start of shaking. This paper presents the variations in Ru with shear strain and the relationship between a normalized shear modulus and shear strain considering the pre-shaking history of sand for shallow depths.

Bioswales are commonly constructed along roadways to control stormwater runoff. Many factors can affect the performance of a bioswale such as the size of the bioswale and its associated drainage area, rainfall characteristics, site conditions, soil properties, and deterioration of the bioswale’s condition over usage. Transportation agencies and engineering communities need a reliable and convenient method for predicting the effectiveness of bioswale. Although available software tools can be used to model and analyze design options, input values for a large number of variables and highly skilled modelers are required to handle these sophisticated modeling tools. The objective of this study was to develop a simplified and easy-to-use mathematical model for predicting the effectiveness of bioswales through empirical predictions of stormwater runoff as a function of four key parameters: area ratio (bioswale surface area to its drainage service area), rainfall depth, rainfall intensity, and sediment accumulation (build-up) on bioswale’s surface area. A PCSWMM model was developed to simulate the physical conditions of a field-scale bioswale. This PCSWMM tool was also used to simulate an idealized (conceptual) catchment model that represents common highway geometries and characteristics. A total of 72 scenarios were simulated on various combinations of the four studied parameters: area ratio (9%, 13%); rainfall depth (2.54, 5.08, 7.62, 10.16 cm); rainfall intensity (2.54, 5.08, 10.16 cm/h); and sediment accumulation (0, 0.25, 1.78 cm). Half of the total scenarios (i.e., 36 scenarios) were used to develop a new simplified mathematical model, and the other 36 scenarios were used to calibrate and validate this newly developed model. The analysis revealed a reasonable correlation (R² = 0.967) between modelled predictions and PCSWMM-simulated results, indicating the newly developed mathematical model can serve as an adequate alternative for simulating bioswales’ performance for stormwater runoff control.

Regular optimization techniques have been widely used in landslide-related problems. This paper outlines two novel optimizations of artificial neural network (ANN) using grey wolf optimization (GWO) and biogeography-based optimization (BBO) metaheuristic algorithms in the Ardabil province, Iran. To this end, these algorithms are synthesized with a multi-layer perceptron (MLP) neural network for optimizing its computational parameters. The used spatial database consists of fourteen landslide conditioning factors, namely elevation, slope aspect, land use, plan curvature, profile curvature, soil type, distance to river, distance to road, distance to fault, rainfall, slope degree, stream power index (SPI), topographic wetness index (TWI) and lithology. 70% of the identified landslides are randomly selected to train the proposed models and the remaining 30% is used to evaluate the accuracy of them. Also, the frequency ratio theory is used to analyze the spatial interaction between the landslide and conditioning factors. Obtained values of area under the receiver operating characteristic curve, as well as mean square error and mean absolute error showed that both GWO and BBO hybrid algorithms could efficiently improve the learning capability of the MLP. Besides, the BBO-based ensemble surpasses other implemented models.