China
Africa
Explore chapters and articles related to this topic
Surrogate modelling of concrete girders using artificial neural network ensemble
Published in Airong Chen, Xin Ruan, Dan M. Frangopol, Life-Cycle Civil Engineering: Innovation, Theory and Practice, 2021
Computational models are becoming more and more complex in many engineering fields and therefore more computationally demanding. It is a new and growing research area connected with the use of computational methods and devices to analyze the mechanical models. In computational mechanics, the computational model – mathematical function – is typically defined using finite element method (FEM). This brings high computational burden especially when calculating structures made of materials with nonlinear response. The situation is even worse when reliability analysis of the structure is performed. This means to carry out a set of numerical simulations which number is increasing with a decrease of expected failure probability. In order to overcome above mentioned high computational burden the methods based on substituting the initial expensive model by a simpler model, which is fast to evaluate, are desirable. Such approach is called surrogate modelling.
Classical and Nonclassical Treatment of Problems in Elastic-Plastic and Creep Deformation for Rotating Discs
Published in Satya Bir Singh, Alexander V. Vakhrushev, A. K. Haghi, Materials Physics and Chemistry, 2020
A. Temesgen, S. B. Singh, Pankaj Thakur
In the current research, solid mechanics is a fundamental discipline which addresses as yet poorly understood phenomenon in the mechanical response and failure ofmaterials and structures. Research in solid mechanics is essential not only for basic understanding of mechanical phenomena but also to advance engineering methodology in a host of areas throughout mechanical and structural technology. Advances in the subject are central to assuring safety, reliability, and economy in design of structures, devices, machines, and complete systems, and hence to the continued development of power generation technologies such as fusion, nuclear and gas turbine power, aerospace and surface transportation vehicles, earthquake-resistant design, offshore structures, orthopedic devices, and materials processing and manufacturing technologies. Nowadays, hot research areas in solid mechanics are material mechanics, computational mechanics, dynamics, instabilities in structures, biomechanics, manufacturing, optimal design of engineering structural components, and so on.
Computational Methods in Cardiovascular Mechanics
Published in Michel R. Labrosse, Cardiovascular Mechanics, 2018
F. Auricchio, M. Conti, A. Lefieux, S. Morganti, A. Reali, G. Rozza, A. Veneziani
In this respect, the terrific development of new devices and algorithms for image and data retrieval and processing has allowed the migration from (over)simplified, idealized descriptions to high-fidelity patient-specific models, in a critical, still ongoing process of merging measurement and conceptualization. Meanwhile, the progressive improvement of computational resources provides the infrastructure to perform numerical simulations of complex dynamics in reasonable times. Complementary to these advances is the development of novel specific modeling techniques and solvers in the field of computational mechanics, in an exciting process involving mathematics, engineering, computer science, and biomedical knowledge. This process already has an impact not only on research but also on clinical practice. As a matter of fact, as the example of the HeartFlow company (Taylor et al., 2013) demonstrates, mathematical and computational modeling can be more than research tools and can be integrated in products for the clinical market.
Influence of aggregate morphology on the mechanical performance of asphalt mixtures
Published in Road Materials and Pavement Design, 2018
Daniel Castillo, Silvia Caro, Masoud Darabi, Eyad Masad
Granular aggregates are the largest constituent by weight and volume in asphalt mixtures and, consequently, they play an important role in defining the overall mechanical response of these materials. Since the shape properties of these particles vary significantly, this phase greatly contributes to the heterogeneity of the mixtures and to the variability related to their behaviour and performance. However, quantifying the effect of these morphological characteristics on the response of asphalt mixtures is a difficult task, mainly due to practical and experimental constraints. In this context, computational mechanics provides a powerful tool to conduct this type of analyses through virtual experiments. As it will be explained in later sections, this paper uses different computational algorithms and finite elements (FE) modelling to evaluate the influence of the shape properties of aggregates on the mechanical behaviour and degradation of asphalt mixtures.
An overview on advances in computational fracture mechanics of rock
Published in Geosystem Engineering, 2021
Mojtaba Mohammadnejad, Hongyuan Liu, Andrew Chan, Sevda Dehkhoda, Daisuke Fukuda
Brittle and semi-brittle rock is very likely to experience observable crack growth at some stage of its life cycle under severe loading. Since the pioneering work by Griffith (1921), for many years, the mechanisms of the crack growth in brittle materials have been studied extensively under the assumptions of linear elastic fracture mechanics (LEFM). However, it was not until the mid-seventies that the fracture of ductile materials was first explored using elasto-plastic fracture mechanics (EPFM) principles. The complexity of the fracture process is even more complicated in a naturally heterogeneous brittle material such as rock and concrete. There are basically three types of investigation techniques in fracture mechanics, namely experimental, analytical and numerical. Computational fracture mechanics has long been used for determination of the stress intensity factors, and later has been expanded into the simulation of crack nucleation and propagation. Generally, rock fracture is essentially a dynamic process, at least in the final stage (Cox, Gao, Gross, & Rittel, 2005; Zhou, Lomdahl, Thomson, & Holian, 1996b), and not all of the numerical methods are capable of correctly capturing the cracking process, due to difficulties posed by time dependency of crack onset and rate dependency of crack velocity (Owen, Feng, Cottrell, & Yu, 2007). For a realistic simulation of the fracture process, numerical techniques are required to model crack onset and arbitrary crack growth, the correct crack length within a given time interval as well as the propagating directions. Recent advances in computational mechanics have facilitated a much better understanding of complex process, and accordingly numerical simulation of the fracture process has been the object of massive interests. Generally, computational mechanics can be classified mainly into continuum and discontinuum formulations. Continuum-based methods discretise the domain into elements and the domain is treated as a single continuous body using a mathematical formulation involving a constitutive law, balance principles, boundary conditions and initial conditions (Munjiza, 2004). The main continuum methods are finite element method (FEM), finite difference method (FDM), boundary element method (BEM), scaled boundary finite element method (SBFEM), extended finite element method (XFEM) and mesh-less methods. Discontinuum-based methods are relatively new, and they model the domain as a collection of discrete bodies that can move, rotate and interact. Accordingly, their mathematical formulation includes the law between particles and balance principles (Munjiza, 2004). Distinct element method, lattice model (LM), and molecular dynamics (MD) are the common discontinuum methods in the field of fracture analysis. In recent years, increasing attention has been paid on these techniques, which can bring together the advantages of the continuum-based and discontinuum-based methods. Attempts in this direction lead to the development of coupled methods, combined methods and multi-scale coupled methods.