Journal of Engineering Mechanics

Soft network materials with biomimetic mechanical properties such as a negative Poisson’s ratio have important applications in tissue engineering, biomedical devices, and soft robotics. Several finite-element (FE)-based design strategies have been developed to produce network materials with prescribed mechanical properties. However, obtaining network designs with a prescribed negative Poisson’s ratio over large strain remains a challenge. Here, an optimization framework was developed using isogeometric analysis and a genetic algorithm for the design of soft missing rib structures with controllable negative Poisson’s ratios over large strains. The missing rib structures with six ligaments were optimized to achieve constant negative Poisson’s ratios ranging from $-0.1$ to $-0.6$ up to 70% tensile strain under plane stress condition. The optimization framework was employed to obtain a missing rib network design with deformation behavior closely matching that of cat’s skin up to 90% tensile strain. This optimized design was fabricated using a biocompatible material via liquid additive manufacturing and validated experimentally, demonstrating the potential of the soft missing rib designs for biomedical applications.

This paper puts forward a perspective or opinion that we can demonstrate Darcy’s law is valid at any scale where fluid can be modeled/analyzed as a continuum. Darcy’s law describes the flow of a fluid through a porous medium by a linear relationship between the flow rate and the pore pressure gradient through the permeability tensor. We show that such a linear relationship can be established at any scale, so long as the permeability tensor is expressed as a function of adequate parameters that describe the pore space geometry, fluid properties, and physical phenomena. Analytical models at pore scale provide essential information on the key variables that permeability depends on under different flow regimes. Upscaling techniques based on the Lippman-Schwinger equation, pore network models, or Eshelby’s homogenization theory make it possible to predict fluid flow beyond the pore scale. One of the key challenges to validate these techniques is to characterize microstructure and measure transport properties at multiple scales. Recent developments in imaging, multiscale modeling, and advanced computing offer new possibilities to address some of these challenges.

A novel exact analytical solution is derived for the equation $y^{(4)}+x^{n}y=0$ in the region $x\ge 0$, which is important for the analysis of piles in soil with stiffness varying with depth. To date, exact solutions for long piles are available only for the cases where $n=-4$, 0, and 1. For other values of the exponent $n$, solutions have formerly been obtained numerically, mainly by the finite-difference method or approximate analytical solutions. An inherent difficulty in obtaining solutions for long beams (which are used to model flexible piles) lies in the inability to isolate the regular, converging part of the solution over the singular part that diverges with increasing $x$. In this paper, an exact solution is derived for $n>-4$, focusing on the important case of semi-infinite beams. Key aspects of the problem such as the stiffness and flexibility matrices at the pile head, and the peak bending moments due to eccentrically acting lateral loads, are discussed. A novel approach for deriving Winkler spring moduli for combined force and moment loading is proposed and shown to provide good agreement with rigorous numerical continuum results.

The borehole pressure of the rock mass under blasting under consideration is essentially the pressure of the transmitted shock wave acting on the borehole wall. Its value directly affects the fragmentation degree and damage range of the rock mass around the borehole. The whole-process action mechanism of the shock wave in different media and interfaces for air-decoupled charge explosion was comprehensively analyzed, discontinuous Galerkin (DG) method was used to simulate and solve parameters of the multiple media flow field formed by the detonation products and air in the borehole. Two theoretical calculation models of the borehole pressure of air-decoupled charge with discontinuous interface conditions of the pressure and particle velocity were proposed and their algorithm flows were presented. Furthermore, the reliability and feasibility of the proposed models were verified by comparing with the numerical simulation results of the same example under five working conditions. It is suggested to use different theoretical models when using different decoupling coefficients. It is shown that there is an optimal value of the decoupling coefficient for air-decoupled charge, which makes the rock blasting obtain a high utilization rate of explosive energy and good blasting effect at the same time. This study can provide a theoretical basis for revealing the blasting rock-breaking mechanism and optimizing blasting designs.

A new isotropic plastic-damage cap model of concrete is proposed in this study to overcome the limitations (e.g., crack closure behavior) of the existing continuous surface cap (CSC) model in the hydrocode LS-DYNA. The explicit return mapping algorithm of the cutting plane is used in the proposed model to update the plastic behavior of concrete. A scalar damage variable based on the two different damage laws, i.e., brittle damage and ductile damage, is employed to represent different damage mechanisms of concrete under different stress states. Utilizing the concept of effective stress and the hypothesis of strain equivalence decouples the plastic-damage algorithms in the numerical integration process. The principal strain rules are employed to capture the crack closure feature and reversal load-bearing capacity of concrete. The proposed cap model is implemented into LS-DYNA as the user-defined material model. A significant number of simulations are performed to validate the rationality and accuracy of the newly developed plastic-damage cap models. It is found that the proposed plastic-damage cap model not only captures the basic mechanical behaviors and damage modes of concrete under different stress states but also is superior to the existing model in the predictions of the crack closure behavior and reversal load-bearing capacity of concrete under static and impact loadings.

To establish a physically direct link between the subsystem-level stiffness demand (i.e., the external frame system, internal core tube, link beam system, and outrigger-belt system) and that at the global system level of frame core-tube structural systems that have been commonly used in super high-rise buildings, this paper examines the analytical correlation of the dynamic characteristics between the modified flexure-shear coupled model (FSM-MS) and modified frame-tube-outrigger model (MFTOM) proposed previously. The MFTOM is developed with the appropriate distribution functions of linear density and subsystem stiffness in the frame-core tube system, and its approximate vibration periods are analytically determined by the rationally assumed mode shape and derived approximate mode shape derivative and curvature. From the perspective of the energy balance of the free vibration, the frequency amplification factor induced by outriggers is formulated. The effectiveness of the correlation established by the first two order vibration periods is systematically demonstrated from a series of cases, from the aspects of vibration periods and inter-story drift ratios. The results of the parameter analysis indicate that the influence of the outrigger on the structural shear-flexural stiffness ratio is mainly realized by the constraint on the flexural deformation of the core tube and the amplification of the axial deformation of the column. The proposed correlation not only provides a reliable theoretical basis for distributing the structural lateral stiffness to the subsystem stiffness in the preliminary design, but can also be used as a practical tool to adjust the overall lateral deformation shape of the frame-core tube system.

Application of an edge-based smoothed finite-element method (ESFEM) to vortex-induced vibration (VIV) of a circular cylinder in generalized Newtonian fluids is presented. The incompressible Navier–Stokes equations incorporating power-law and Carreau–Yasuda viscosity models are solved by the characteristic-based split scheme under the arbitrary Lagrangian–Eulerian description. The equation of motion of an elastically supported circular cylinder subjected to the generalized Newtonian fluid flows is advanced via the generalize-$\alpha$ method. The spatial discretization is based on a three-node triangular element that is particularly suitable for the ESFEM. New integration points are subsequently proposed in local smoothing domains to facilitate the weak-form approximation. The fluidic excitation acting on the submerged cylinder is also derived from the edge-based notion. Grid nodes are instantaneously rearranged by a cost-effective moving submesh approach. Especially, a mass source term is structured in the current context to satisfy geometric conservation law for the ESFEM. The tightly coupled mechanical system is settled through fixed-point iterative procedure. The present method is validated against available data for two non-Newtonian VIV examples.

Physics-informed neural networks (PINNs) have received increased interest for forward, inverse, and surrogate modeling of problems described by partial differential equations (PDEs). However, their application to multiphysics problem, governed by several coupled PDEs, presents unique challenges that have hindered the robustness and widespread applicability of this approach. Here we investigate the application of PINNs to the forward solution of problems involving thermo–hydro–mechanical (THM) processes in porous media that exhibit disparate spatial and temporal scales in thermal conductivity, hydraulic permeability, and elasticity. In addition, PINNs are faced with the challenges of the multiobjective and nonconvex nature of the optimization problem. To address these fundamental issues, we (1) rewrote the THM governing equations in dimensionless form that is best suited for deep learning algorithms, (2) propose a sequential training strategy that circumvents the need for a simultaneous solution of the multiphysics problem and facilitates the task of optimizers in the solution search, and (3) leveraged adaptive weight strategies to overcome the stiffness in the gradient flow of the multiobjective optimization problem. Finally, we applied this framework to the solution of several synthetic problems in one and two dimensions.