Search

A rational constitutive model of crushable soils is proposed based on a novel evolution law for the grading index, which incorporates the effect of packing density on the particle crushing phenomena through the coordination number as an intermediate variable. The evolution law of the grading index reasonably combines the increase in the coordination number due to an increase in packing density; and the increase in the crushing resistance due to the increase in the coordination number. The effect of particle crushing on the stress–strain response is incorporated by lowering the critical state specific volume with the changing grading due to crushing. The validity of the proposed constitutive model was verified with an extensive series of elementary tests and DEM. The proposed model properly predicted two seemingly contradictory experimental evidence for crushable soils that: densely packed soil exhibits lesser crushing than loosely packed one under the same effective stress paths; densely packed soil is more likely to crush than loosely packed one when it is sheared. This approach can predict the stress–strain response of crushable soils under different packing densities and stress paths, offering valuable insights for comprehending the crushing responses. [ABSTRACT FROM AUTHOR]

• The machine learning (ML) models are utilized to predict compression parameters from CPTu's results. • The models overcome the difficulties in predicting compression index and swelling index. • The ML models with optimization of architecture outperform the classical ML models. • Large database was gathered, and sensitive analysis is performed. • Feature analysis is conducted, and a case study confirm the effectiveness of the ML models. Soil compression parameters are significant factors for determining settlement to ensure safety of the civil engineering structures. These parameters are strictly evaluated through laboratory tests using samples collected from drilling boreholes. Currently, the testing procedures require both time and labour, and extensively increased construction's cost in some cases such as in offshore structures. Therefore, it is necessary to establish a robust and reliable method, which can easily be used to obtain these parameters. In this paper, we employ Machine Learning (ML) models to relate field-testing results of piezocone penetration test (CPTu) to the compression parameters. A large database for this study is considered from five different construction projects, including two roads, a factory, and two massive container ports. Based on the database, a sensitivity analysis of the dataset and a feature analysis are performed. Furthermore, in order to select an appropriate ML method, several models are employed such as artificial neural network (ANN), deep neural network (DNN), DNN optimized with genetic algorithms (GA), and DNN optimized with particle swarm optimization (PSO). Comparisons between these ML models are also performed with remarkable better performances from the optimized DNN models than the classical models of ANN or DNN. Finally, validations are carried out using data from Nguyen Son (NS) road project to confirm the effectiveness and reveal the promising potential applications of the proposed methods. [ABSTRACT FROM AUTHOR]

[Display omitted] • For the ·CH 3 + i -C 3 H 7 OH reaction, the formation of CH 4 + (CH 3) 2 Ċ-OH products was the most feasible. Its predicted rate constant at 500 K, k R1 = 1.49. × 10-16 cm3 molecule-1 s−1, agrees closely with the experimental data given by P. Gray and A. A. Herod with 1.03 × 10-16 cm3 molecule-1 s−1. • The expression of the predicted rate constant of the most speedy reaction in the ·CH 3 + n -C 3 H 7 OH reaction system was k Re1 = 1.59 × 10-25 × T 4.03 exp(-3430.7/ T) cm3 molecule-1 s−1 at 298–2000 K. Contribution of k Re1 (CH 3 + n -C 3 H 7 OH) to the total rate constant is not superior as k R1 (CH 3 + i -C 3 H 7 OH). • The CH 3 + i -C 3 H 7 OH reaction is faster than CH 3 + n -C 3 H 7 OH reaction in the same conditions. The mechanisms and kinetics of the reactions of methyl radical with n/i -propanol (n/i -C 3 H 7 OH) were investigated in detail using density functional theory and coupled cluster theory with rate constant prediction. The thermodynamic properties and the schematic potential energy surfaces of the two reactions were calculated by single point calculations at CCSD(T)/6–311++G(3df,2p) level based on geometries optimized at the B3LYP/6–311++G(3df,2p). The rate constants for various product channels of the two reactions in the temperature range of 298–2500 K were predicted by transition state theory and Eckart tunneling correction. The results of the analysis of schematic potential energy surfaces, thermokinetics parameters, and rate constants showed that the abstraction H-atom from the carbon atom which was bonded with an oxygen atom in the n/i -propanol molecules is the most favorable of these reaction systems. [ABSTRACT FROM AUTHOR]