ABSTRACT Tunnels are frequently subjected to blast‐induced waves from adjacent construction activities, making their dynamic analysis critically important. However, the conventional plane‐wave assumption adopted in many studies is often inadequate for near‐field source scenarios, where the curvature of the incident wavefront cannot be neglected. This study investigates the scattering of cylindrical P‐ and SV‐waves by a circular lined tunnel embedded in an unsaturated porous medium with an imperfect interface. By using the wave function expansion method, analytical solutions are derived for the stress and displacement fields in the unsaturated substrate and elastic lining. Numerical examples are presented to examine the effects of incident‐wave frequency, degree of saturation, wave‐source distance, and interfacial stiffness on the dynamic stress concentration factor (DSCF). The results show that saturation strongly affects the DSCF under a high‐frequency P‐wave, whereas its influence under an SV‐wave is limited. These findings provide a theoretical basis for evaluating lined tunnels subjected to near‐field blast‐induced waves in partially saturated ground.
ABSTRACT The liquefaction susceptibility of saturated sands is commonly evaluated using macroscopic indices such as relative density and confining pressure, whereas the role of particle morphology remains less well understood. In this study, a coupled computational fluid dynamics‐discrete element method (CFD‐DEM) framework is used to investigate particle‐shape effects on liquefaction under a unified deposition and cyclic loading protocol. Three idealized assemblies composed of 1‐sphere, 2‐sphere, and 3‐sphere particles are compared. The results show that particle shape affects packing state, permeability, depositional fabric, and load‐bearing skeleton stability. Although the 2‐sphere assembly has the densest initial packing, it exhibits the fastest excess pore‐pressure buildup and the lowest liquefaction resistance. The 1‐sphere assembly shows the slowest pore‐pressure accumulation and the highest liquefaction resistance, while the 3‐sphere assembly exhibits an intermediate response. The higher permeability of the 1‐sphere assembly may partly promote pore‐pressure dissipation, but the similar permeabilities of the 2‐sphere and 3‐sphere assemblies cannot explain their markedly different liquefaction responses. Micromechanical analyses show that the 2‐sphere assembly undergoes more rapid fabric and contact‐network degradation, whereas the 3‐sphere assembly maintains greater stability through stronger geometric interlocking. These results indicate that liquefaction resistance is governed by the coupled effects of hydraulic characteristics and shape‐induced microstructural evolution rather than by packing density or permeability alone.
ABSTRACT Integrating distributed temperature sensing (DTS) and distributed acoustic sensing (DAS) data with numerical simulations to analyze inter‐well interference and production responses remains challenging. This study employs a coupled thermal–hydraulic–mechanical simulation framework to model heat transfer, fluid flow, and geomechanical responses associated with the initiation and propagation of multiscale fractures. To represent fractures across different scales, the framework integrates commercial simulation capabilities with external scripts, including a new method for simulating natural‐fracture reactivation and closure and a dynamic scheme for updating matrix permeability associated with non‐Darcy gas flow. The results show that frac hits can significantly affect fracture growth, thermal behavior, and flowback performance at nearby perforation intervals, with their impacts depending on proximity, intensity, time since formation, and offset‐well production status. These findings have direct implications for fracture‐stimulation design, indicating that stimulation spacing, treatment sequencing, and offset‐well operating conditions should be considered together to manage fracture interference, maintain effective connectivity, and improve flowback performance. Low‐frequency DAS observations are used to constrain frac‐hit timing and location, while DTS responses are directly compared with simulated temperature behavior in a field example. Near‐wellbore DAS observations further provide qualitative constraints on relative flowback distribution. The simulations reproduce key features observed in the field DTS data and provide physical explanations for their complex temperature responses. Overall, this study demonstrates a practical framework for integrating fibre‐optic observations with coupled multiphysics simulations to improve fracture diagnosis, support stimulation‐design decisions, and assess treatment efficiency and stimulated reservoir volume under different frac‐hit scenarios.
ABSTRACT The classical theory of porous media is based on assumed micro‐homogeneity of the solid phase (skeleton). Based on the drained macroscale elastic stiffness, it is possible to analytically compute auxiliary properties such as the Biot‐Willis and Skempton parameters. The more general case of a micro‐heterogeneous solid phase, which is addressed in this paper, requires computational homogenization of the fluid‐solid interaction. Linearized kinematics is adopted, which is a necessary (but not sufficient) assumption for direct upscaling. Linear elasticity with arbitrary anisotropy is then assigned to the solid skeleton, while the fluid motion is modeled as compressible Stokes' flow. The RVE problem is established based on assumed periodicity for the fluctuation parts of solid displacement, pore pressure and fluid velocity. Although the macroscale problem is transient (quasi‐static), we approximate the RVE‐problem as (micro‐)stationary. The macroscale problem reveals the classical structure with displacements and pore pressure as the unknown fields. It is then possible to rigorously show the particular symmetry that the same Biot‐Willis coefficient appears in the “effective stress principle” as well as in the fluid storage function, thereby generalizing the classical model feature. In the numerical examples, we first verify the classical results for a micro‐isotropic and micro‐homogeneous solid phase. We thereby consider a simplified periodic micro‐architecture, which lends itself to an expected orthotropic macroscale response. Next, we consider a micro‐heterogeneous solid skeleton with inclusions, in which case the classical analysis does not apply.
The existing studies have reported that the surface of Mars shows the presence of materials having igneous origin, including basalt and regolith-type coarse soils. These findings are affected by significant uncertainties and, therefore, the selection of simulants for various exploration applications, including surface mobility of rovers, should take into account the potential complications that might arise if unexpected materials are found. The present study focusses on the micromechanical characterisation of a regolith-type soil (Eglin sand), comparing the results obtained with those determined for a quartz sand (Leighton Buzzard sand), that is also used as a Mars-soil simulant. Eglin sand is constituted by a complex composition and the various minerals identified were tested and analysed separately. The testing activities included the characterisation of the particle shape and surface morphology, and the determination of particle strength, contact response, contact stiffness and inter-particle friction. Differences in terms of particle morphology and surface properties were recorded for the two sands studied, and, along with grain mineralogy, these have been found to have an impact on the mechanical response, affecting both particle strength and inter-particle behaviour. This study aims to provide realistic soil data and information to be used in numerical simulations performed using the Discrete Element Method, which is often adopted for the analysis of surface mobility problems.
