Computers & Structures最新文献

In large-scale structural fire resistance tests, the interaction between the individual elements and the surrounding structure causes discrepancies in behaviour compared to single-element fire tests. Large-scale tests of real structures are challenging due to financial and time limitations. To bridge this gap, the concept of “Hybrid Fire Testing (HFT)” emerges, where a portion of the structural system (i.e., physical substructure) is experimentally tested while the remaining structure (i.e., numerical substructure) is analyzed numerically. The primary challenges in HFT involve ensuring stability throughout the analysis by considering the varying stiffness of the fire-exposed element during the test and establishing a versatile communication platform between the physical substructure (PS) and numerical substructure (NS) components. This paper presents a comprehensive HFT framework, implemented within a user-friendly software interface, facilitating both virtual and experimental testing. The software incorporates a new method addressing stability concerns by predicting PS stiffness during the test, achieving convergence within a limited number of iterations. Additionally, the framework includes a communication platform utilizing internet protocols (IP) and COM ports for rapid and easy connection to diverse experimental control systems and finite element software packages. The functionality of the developed software is validated through its successful application in an HFT conducted on a 3-story steel structure within a simulated environment. Both force-controlled and displacement-controlled approaches confirm the method’s adaptivity to the employed test procedures.

The paper presents a new type of weakly nonlinear two-scale model of controllable periodic porous piezoelectric structures saturated by Newtonian fluids. The flow is propelled by peristaltic deformation of microchannels which is induced due to piezoelectric segments embedded in the microstructure and locally actuated by voltage waves. The homogenization is employed to derive a macroscopic model of the poroelastic medium with effective parameters modified by piezoelectric properties of the skeleton. To capture the peristaltic pumping, the nonlinearity associated with deforming configuration must be respected. In the macroscopic model, this nonlinearity is introduced through homogenized coefficients depending on the deforming micro-configurations. For this, linear expansions based on the sensitivity analysis of the homogenized coefficients with respect to deformation induced by the macroscopic quantities are employed. This enables to avoid the two-scale tight coupling of the macro- and microproblems otherwise needed in nonlinear problems. The derived reduced-order model is implemented and verified using direct numerical simulations of the periodic heterogeneous medium. Numerical results demonstrate the peristaltic driven fluid propulsion in response to the electric actuation and the efficiency of the proposed treatment of the nonlinearity. The paper shows new perspectives in homogenization-based computationally efficient modelling of weakly nonlinear problems where continuum microstructures are perturbed by coupled fields.

This paper presents a new optimization framework in which the structural analyzer (isogeometric analysis–IGA) and data-driven surrogate model (deep neural network–DNN) are sequentially and repeatedly employed as the evaluation function in the optimization process of the computationally heavy problem of three-dimensional material distribution optimization in functionally graded (FG) plates. The optimization process starts with IGA normally, and the key point is to collect the evaluated candidates as data to build DNNs as surrogates predicting the plate behavior. Then, in the surrogate-assisted phase, based on the best predicted value, one more IGA analysis could be performed to find a new truly best candidate solution. This is also to track the surrogates' accuracy, which is another key feature of the proposed framework. When the prediction becomes less accurate, the optimization process is back to using IGA, more data is collected, and the whole procedure is repeated. Compliance minimization in FG plates under static bending is considered with various plate geometries. Numerical results confirm that the proposed recurrent optimization framework reduces up to 38% computational time whilst ensuring that the best candidate solution is always exact and of highest optimality.

To improve accuracy and convergence of biaxial wheel fatigue simulation with coupled nonlinearity, we propose a two-stage approach based on a composite tire model. The tire model is calibrated through an identification procedure, wherein the actual tire stiffness characteristics are matched, effectively addressing the difficulty in lack of tire structure and materials information. Based on the identified tire model, restart analysis algorithm is employed to decouple the biaxial simulation into a two-stage analysis, where wheel deformability is sequentially considered. At the first stage, large deformation of the loaded tire is calculated by modeling the wheel as a rigid part. Then the deformation and stress states of tire are maintained at the second stage, and the wheel elasticity is recovered for stress calculation. Compared to a single-stage direct method, the proposed method significantly reduces computational costs, while exhibiting only a minor stress discrepancy on the wheel rim. Finally, experimental results show that the present method not only ensures high accuracy in predicting stresses of the wheel disc, but also effectively reduces errors on the wheel rim region. It is convinced that the proposed method provides an efficient and reliable means for the comprehensive evaluation of wheel strength in biaxial fatigue tests.

The use of Reduced Dimensional Models (RDM) discretized like beams, plates and shell elements drastically decreases the computational cost of solving a full 3D elastic problem with a Finite Element Method (FEM). However, its kinematic assumptions are only applicable to bodies with regular sections or continuous layouts. For the correct analysis of irregular regions, it is necessary to rely on bi-dimensional or solid models that fully reproduce the geometry of the body and its behavior but have a much higher computational cost. The Mixing Dimensional Coupling (MDC) technique allows linking models discretized with elements of different topologies, allowing the possibility of considering the most cost-effective model in each region. This coupling takes place at the interface that delimits both models and relies on the equilibrium of work and reactions on its two faces. In this paper, the formulation is presented for coupling beams with laminar sections and 2D Plane-Stress (PS) models demonstrating its proper behavior. Finally, this coupling is used for defining a new beam element, the Beam-Like Reduced Order Model (BLROM), which is obtained from a Plane-Stress model of their longitudinal section.

The analytical damped dynamic stiffness formulation is developed for the dynamic response analysis of orthotropic viscoelastic plate built-up structures with a general frequency-dependent damping model. The governing differential equation in the frequency domain is established, which allows for the direct introduction of frequency-dependent damping models by considering internal (material) and external (environmental) damping. The adopted viscoelastic damping model is sufficiently general to describe various types of damping, including viscous or non-viscous, integer or fractional order models. Then, the exact damped dynamic stiffness formulations for both in-plane and out-of-plane vibrations of plate elements are developed. Arbitrarily distributed excitations can be applied to the plate nodal boundaries based on the analytical Fourier-type forward and inverse transforms. The dynamic response analysis of the viscoelastic plate is carried out, which verifies the accuracy and efficiency of this method within the broadband frequency range. The numerical results serve as a valuable reference and can be used as benchmark solutions. Accurate and profound comprehension of the dynamical behavior of viscoelastic plates is a key task in designing these structures, and also optimizing their vibrational behavior. This method offers a powerful tool for representing the broadband dynamics of viscoelastic plate structures, utilizing very few degrees of freedom.

Additive manufacturing (AM) has revolutionized the way we design and manufacture lightweight composite structures with complex geometries and extraordinary performance. In composite AM, fibers are often steered within the plane of printing and sometimes at predefined discrete angles. Hence, designing structures for AM must consider such manufacturing constraints along with the concurrent optimization of structures and fiber orientations. Herein, we propose a method that uses a penalized normal distribution (PND) function to design the fiber orientation based on predefined discrete angles. By discretizing a continuous design variable and penalizing the effective properties, the method effectively drives the design variable to converge to one of the target candidates with low deviations. Using only one design variable at each spot, the method is scalable and can be easily adapted as the number of candidates changes. By coupling the discrete angle optimization with structural optimization, the multiscale method concurrently optimizes the structural topology with fiber orientations considering AM constraints. Numerical examples demonstrate the advantages of this framework and its extension to solving 3D problems.

The rising cost of natural resources and environmental concerns motivate systematic design and manufacture of more efficient structures. For that purpose, topology optimization has been appealing, as well as working on an enlarged design space to include multi-material solutions. The resulting optimal designs can be materialized using multi-material additive manufacturing. In the present framework, multi-material printed parts or layouts can be envisaged as having better strength properties than single-material counterparts.

The maximum von Mises stress is minimized inside a design domain through topology changes and material selection. The selected composite material model encompasses either the classical arrange of two discrete materials with sharp interfaces, or their mixture controlled by the volume fraction of each base material to generate a Functionally Graded Material (FGM). An optimized continuous variation of properties makes the FGM appealing to mitigate stress concentrations. To adequately capture the physics of mixtures considering the FGM’s mechanical properties, one uses the RAMP interpolation scheme within the Hashin-Shtrikman bounds.

A set of plane stress benchmarks are proposed. It is shown that considerably lower stress peaks on the evaluated structures can be obtained on the account of introducing more than one solid phase, specifically in the case of FGM solutions.

Existing non-gradient topology optimization algorithms require numerous objective function evaluations for reinforcement design of damaged structures, resulting in huge computational costs. In this study, a non-gradient directed evolution (NGDE) topology optimization method is proposed for strengthening locally damaged RC structures. First, a topology description strategy for reinforcement material (RM) component is developed to reduce the number of design variables. Then, a directed generation criterion of RM samples is given by locating the regions with the maximum concrete damage from the previous iteration step. Subsequently, the random disturbance operation, incorporating mutation, crossover, and selection, is employed to enhance the diversity of RM samples. On this basis, the element removal and size adjustment strategies of RM components are presented to overcome the numerical instability. Finally, the applicability and effectiveness of the proposed method is illustrated by the numerical examples. Results show that the proposed method can effectively generate a reasonable RM configuration for strengthening damaged RC structures without relying on design sensitivity.

Two highly accurate and reliable eigen-solution techniques, the new homotopy perturbation method and the extended argument principle method, are proposed for analysing orthotropic viscoelastic plate built-up structures. These techniques are formulated to solve transcendental eigenvalue problems in modal analysis based on the analytical damped dynamic stiffness formulations. The homotopy perturbation method uses undamped real-valued eigenvalues and eigenvectors computed by the Wittrick-Williams algorithm as the exact initial solutions. The internal damping coefficient and external damping coefficient are set as convergence control parameters by using the homotopy method, and the initial solutions are updated through inverse iteration to efficiently obtain the final complex eigenvalues. Conversely, the extended argument principle method utilizes the dichotomy of mode count in the complex domain, based on the denominators of elemental dynamic stiffness matrices, to pinpoint complex eigenvalues. Validation against finite element solutions from COMSOL shows that while the extended argument principle method offers benchmark solutions, it is computationally intensive. In contrast, the proposed homotopy perturbation method presents a valuable tool in engineering applications due to its exceptional balance of accuracy and computational efficiency. This method facilitates rapid analyses and design parameter optimization within the context of viscoelastic plate structures.