The Infona portal uses cookies, i.e. strings of text saved by a browser on the user's device. The portal can access those files and use them to remember the user's data, such as their chosen settings (screen view, interface language, etc.), or their login data. By using the Infona portal the user accepts automatic saving and using this information for portal operation purposes. More information on the subject can be found in the Privacy Policy and Terms of Service. By closing this window the user confirms that they have read the information on cookie usage, and they accept the privacy policy and the way cookies are used by the portal. You can change the cookie settings in your browser.
Disordered solids exhibit intermittent avalanches when slowly driven by an external load. These avalanches are associated with plastic rearrangements of the atoms at the nanoscale that manifest as stress and energy drops in the loading curve. The complexity arising from their interactions through long‐range elastic fields and the disorder makes statistical approaches suitable for studying their behavior...
The prediction of the onset of fracture is a challenging issue in the mechanics of disordered materials. In this contribution, we show that the fracture process in network glasses, such as silica glass, turns out to be a complex phenomenon that originates from specific spots that have the size of a few hundred atoms only. We apply pure shear deformation to identify local rearrangement spots prone...
The one‐atom‐thick allotrope of graphite, C4 is called monolayer graphene and was discovered in 2004. It is well known for its fantastic electro‐mechanical properties. While transverse contraction and Poisson's ratio were previously studied only for homogeneous or crystalline two‐dimensional (2D) mono‐ and bilayer network structures under uniaxial tensile stress, we now numerically investigate these...
During a shear process the vibrational mode structure of a non‐crystalline model material will change under load. Thus, we expect an effect on the characteristic boson peak, which correlates with numerous features of disordered materials. In this paper, we perform shear deformation on two‐dimensional random network materials and investigate the distribution of their vibrational density of states (VDOS)...
The crude Monte Carlo method is computationally expensive. Hence, incorporating model order reduction methods enabling reliability analysis for high‐dimensional problems is necessary. However, this strategy may result in an inaccurate estimation of the probability of failure for rare events for two reasons. First, the model order reduction, represented by the proper orthogonal decomposition (POD)...
In this paper, we link the unique mechanical properties to structural aspects using a splitting method in order to disentangle nonlinear interactions of Stone–Wales defects as fundamental perturbations in a crystalline environment. We investigate the spatial features of the displacement field achieved by this splitting method and couple it to macroscopic material properties. The nonlinear interaction...
Markov Chain Monte Carlo simulations form an essential tool for exploring high‐dimensional target distributions. Metropolis developed a fundamental random walk algorithm which was improved by Hastings later. The result is known as the Metropolis‐Hastings algorithm, which enables the exploration of multi‐dimensional distributions. The main drawbacks of this algorithm are its high auto‐correlation and...
Cellulose polymers are widely used to fabricate green composites, implemented as fiber, matrix, and adhesive material between them. In this study, cellulose polymers are used as interphase material between spherical nanoparticles of hydroxylated alumina (Al2O3) and epoxy. Molecular dynamics simulations using the large‐scale atomic/molecular massively parallel simulator (LAMMPS) are utilized to investigate...
In this contribution, we present a non‐incremental solution procedure for the efficient treatment of elastoplastic problems. To this end, all time history data are decoupled into space and time, and the solution is obtained using the fixed‐point algorithm.
Methods from computational intelligence, such as (artificial) neural networks, have become an active research direction in mechanics, leading to the development of intelligent constitutive models, surrogate models, and meta elements. Therein, many neural network architectures are inspired by mechanical domain knowledge in the form of physics‐informed or physics‐guided neural networks. Complementary...
This paper investigates the transverse contraction of amorphous bilayer 2D Silica with varying network heterogeneities under uniaxial tensile loading using athermal molecular simulations in LAMMPS. It turns out that the transverse contraction increases nonlinearly with the x‐strain while it decreases with increasing network heterogeneity. We show that it is possible to engineer the heterogeneity of...
In this contribution, we compare two different neural network architectures to predict the response statistics of structures. The overall goal is a significant speed‐up of the numerically expensive Monte Carlo simulation. The first approach is based on a convolutional neural network that learns from the whole excitation history, whereas the second approach is based on a feed‐forward network architecture...
Evaluating the response statistics of nonlinear structures constitutes a key issue in engineering design. Hereby, the Monte Carlo method has proven useful, although the computational cost turns out to be considerably high. In particular, around the design point of the system near structural failure, a reliable estimation of the statistics is unfeasible for complex high‐dimensional systems. Thus, in...
Since Poisson's ratio has only been determined for crystalline forms of network materials, we investigate freestanding, amorphous monolayer 2D silica with varying ring size heterogeneity under tensile stress. Compared to the crystalline material, the relation between x‐ and y‐strain is slightly nonlinear. Thus, the Poisson's ratio is not constant, but a strongly oscillating function of the x‐strain...
In this contribution, we present a non‐incremental time integration scheme using the implicit Newmark approximations. The result is one space‐time equation that can be solved using the proper general decomposition ansatz.
A data‐driven “intelligent” meta‐element that reduces the dimensionality and accelerates nonlinear finite element computations is demonstrated on an elastoplastic continuum frame.
In this contribution, we present a fast prediction approach to estimate response statistics of the crude Monte Carlo simulation using artificial neural networks. Hereby, the neural network is trained within an initial response subset, based on which a forecast can be evaluated in an early state of the Monte Carlo simulation.
In order to investigate the behavior of 2D silica glass subjected to mechanical loading, we develop an algorithm for the realization of 2D newtork glasses consisting of corner‐sharing SiO3 triangles that build rings of different shapes and sizes. Our algorithm extends the network ring by ring in circular manner around the existing configuration. A random engine generates the size of each new ring...
The aim of the present study is to investigate how structural models with inelastic material behaviour can be enhanced by means of artificial intelligence. To this end, artificial neural networks are proposed to replace continuum mechanical models in structural dynamics. This concerns predictions of shock wave‐loaded plates and finite element simulations with neural network enhanced material modelling...
Structural integrity is an ubiquitous topic in research and daily living. Conventional monitoring strategies follow a preventive maintenance plan or require permanent installation of sensors. These steps are time and cost‐inefficient taking into account the high number of bridges to be monitored. In this study, a bridge structure is passively monitored using the acceleration signal of a bypassing...
Set the date range to filter the displayed results. You can set a starting date, ending date or both. You can enter the dates manually or choose them from the calendar.