Heterogeneous multiscale methods (HMM) capable of simulating asynchronously multiple scales concurrently are now tractable with the advent of exascale supercomputers. However, naive implementations display a large number of redundancies and are very costly. The macroscale model typically requires computations of a large number of very similar microscale simulations. In hierarchical methods, this is barely an issue as phenomenological constitutive models are inexpensive. However, when microscale simulations require, for example, high‐dimensional molecular dynamics (MD) or finite element (FE) simulations, redundancy must be avoided. A clustering algorithm suited for HMM workflows is proposed that automatically sorts and eliminates redundant microscale simulations. The algorithm features a combination of splines to render a low‐dimension representation of the parameter configurations of microscale simulations and a graph network representation based on their similarity. The algorithm enables the clustering of similar parameter configurations into a single one in order to reduce to a minimum the number of microscale simulations required. An implementation of the algorithm in the context of an HMM application coupling FE and MD to predict the chemically specific mechanical behavior of polymer‐graphene nanocomposites. The algorithm furnishes a threefold reduction of the computational effort with limited loss of accuracy.

We present the VECMA toolkit (VECMAtk), a flexible software environment for single and multiscale simulations that introduces directly applicable and reusable procedures for verification, validation (V&V), sensitivity analysis (SA) and uncertainty quantication (UQ). It enables users to verify key aspects of their applications, systematically compare and validate the simulation outputs against observational or benchmark data, and run simulations conveniently on any platform from the desktop to current multi-petascale computers. In this sequel to our paper on VECMAtk which we presented last year [ 1 ] we focus on a range of functional and performance improvements that we have introduced, cover newly introduced components, and applications examples from seven different domains such as conflict modelling and environmental sciences. We also present several implemented patterns for UQ/SA and V&V, and guide the reader through one example concerning COVID-19 modelling in detail. This article is part of the theme issue ‘Reliability and reproducibility in computational science: implementing verification, validation and uncertainty quantification in silico ’.

Graphene oxide (GO) shares many novel mechanical and electronic properties with graphene and has been applied extensively for uses in physics, engineering and medicine. Computational simulations of GO have widely neglected accurate characterisation by random functionalisation, forsaking steric strain and abandoning edge functional groups. Here, we show that molecular dynamics forcefield design using electronic structure calculations of hundreds of atoms of GO with accurate functionalisation shows good agreement with state-of-the-art ab initio molecular dynamics (AIMD) simulations. We find that the bespoke forcefield shows better agreement with previous AIMD and experimental results in terms of the interfacial water dynamics and ion adsorption. Namely, GO described by the bespoke forcefield is found to disrupt the hydrogen bonding network at the interface by playing a more dynamic role in accepting and donating hydrogen bonds from water. Furthermore, with the bespoke forcefield, we find preferential adsorption of ions to carboxyl functional groups and a similar mean adsorption half-life for Na+ and Cl− ions around GO. These findings are critical for future investigations of GO in complex environments in application ranging from desalination to protein adsorption for drug delivery.

Molecular dynamics simulation is now a widespread approach for understanding complex systems on the atomistic scale. It finds applications from physics and chemistry to engineering, life and medical science. In the last decade, the approach has begun to advance from being a computer-based means of rationalizing experimental observations to producing apparently credible predictions for a number of real-world applications within industrial sectors such as advanced materials and drug discovery. However, key aspects concerning the reproducibility of the method have not kept pace with the speed of its uptake in the scientific community. Here, we present a discussion of uncertainty quantification for molecular dynamics simulation designed to endow the method with better error estimates that will enable it to be used to report actionable results. The approach adopted is a standard one in the field of uncertainty quantification, namely using ensemble methods, in which a sufficiently large number of replicas are run concurrently, from which reliable statistics can be extracted. Indeed, because molecular dynamics is intrinsically chaotic, the need to use ensemble methods is fundamental and holds regardless of the duration of the simulations performed. We discuss the approach and illustrate it in a range of applications from materials science to ligand-protein binding free energy estimation. This article is part of the theme issue 'Reliability and reproducibility in computational science: implementing verification, validation and uncertainty quantification in silico'.

The current approach to materials discovery and design remains dominated by experimental testing, frequently based on little more than trial and error. With the advent of ever more powerful computers, rapid, reliable, and reproducible computer simulations are beginning to represent a feasible alternative. As high performance computing reaches the exascale, exploiting the resources efficiently presents interesting challenges and opportunities. Multiscale modeling and simulation of materials are extremely promising candidates for exploiting these resources based on the assumption of a separation of scales in the architectures of nanomaterials. Examples of hierarchical and concurrent multiscale approaches are presented which benefit from the weak scaling of monolithic applications, thereby efficiently exploiting large scale computational resources. Several multiscale techniques, incorporating the electronic to the continuum scale, which can be applied to the efficient design of a range of nanocomposites, are discussed. Then the work on the development of a software toolkit designed to provide verification, validation, and uncertainty quantification to support actionable prediction from such calculations is discussed.

Graphene is a household name, but is not yet a household product. The science of 2D materials was sparked with graphene’s isolation in 2004, theoretical predictions made since have heralded nothing short of revolution in the fields of composite materials, electronics, and energy storage. Those revolutions are yet to materialise, but the feverish interest amongst the scientific community continues, motivated by graphene’s tantalising properties and our dogged desire to exploit them. One exciting prospect for graphene is to exploit its mechanical properties as an effective reinforcing component in composite materials. For this to be possible one must overcome the difficulty in producing high quality graphene dispersions in large quantities and effectively transfer its properties to the bulk material. I developed an experimentally and theoretically verified forcefield for molecular dynamics, which replicates graphene’s non-bonded interactions. Using this, I was able resolve graphene’s unusual behaviour whilst in a low friction state (known as superlubricity) and the micromechanical exfoliation of nanoflakes of graphene. I find that graphene’s low bending energy results in a pealing mechanism requiring less work than simply shearing graphite. I give insights into the nanostructure of graphene oxide, and predict that graphene oxide’s percolation threshold will arrive at carbon oxygen ratios below 6, an important result for use in electronic devices. The work presented in this thesis is part of an ongoing effort to develop a multiscale simulation method that links finite element analysis with molecular dynamics, with the aim of predicting macroscale properties of materials from nanoscale structures. This method exploits the power of high performance computing and shows that single scale simulation of graphene nanocomposites is often insufficient.

Validation, verification, and uncertainty quantification (VVUQ) of simulation workflows are essential for building trust in simulation results, and their increased use in decision‐making processes. The EasyVVUQ Python library is designed to facilitate implementation of advanced VVUQ techniques in new or existing workflows, with a particular focus on high‐performance computing, middleware agnosticism, and multiscale modeling. Here, the application of EasyVVUQ to five very diverse application areas is demonstrated: materials properties, ocean circulation modeling, fusion reactors, forced human migration, and urban air quality prediction.

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV2) virus has rapidly spread worldwide since December 2019, and early modelling work of this pandemic has assisted in identifying effective government interventions. The UK government relied in part on the CovidSim model developed by the MRC Centre for Global Infectious Disease Analysis at Imperial College London, to model various non-pharmaceutical intervention strategies, and guide its government policy in seeking to deal with the rapid spread of the COVID-19 pandemic during March and April 2020. CovidSim is subject to different sources of uncertainty, namely parametric uncertainty in the inputs, model structure uncertainty ( i.e. , missing epidemiological processes) and scenario uncertainty, which relates to uncertainty in the set of conditions under which the model is applied. We have undertaken an extensive parametric sensitivity analysis and uncertainty quantification of the current CovidSim code. From the over 900 parameters that are provided as input to CovidSim, we identified a key subset of 19 parameters to which the code output is most sensitive. We find that the uncertainty in the code is substantial, in the sense that imperfect knowledge in these inputs will be magnified to the outputs, up to the extent of ca. 300%. Most of this uncertainty can be traced back to the sensitivity of three parameters. Compounding this, the model can display significant bias with respect to observed data, such that the output variance does not capture this validation data with high probability. We conclude that quantifying the parametric input uncertainty is not sufficient, and that the effect of model structure and scenario uncertainty cannot be ignored when validating the model in a probabilistic sense.

Controlling the structure of graphene and graphene oxide (GO) phases is vitally important for any of its widespread intended applications: highly ordered arrangements of nanoparticles are needed for thin-film or membrane applications of GO, dispersed nanoparticles for composite materials, and 3D porous arrangements for hydrogels. By combining coarse-grained molecular dynamics and newly developed accurate models of GO, the driving forces that lead to the various morphologies are resolved. Two hydrophilic polymers, poly(ethylene glycol) (PEG) and poly(vinyl alcohol) (PVA), are used to illustrate the thermodynamically stable morphologies of GO and relevant dispersion mechanisms. GO self-assembly can be controlled by changing the degree of oxidation, varying from fully aggregated over graphitic domains to intercalated assemblies with polymer bilayers between sheets. The long-term stability of a dispersion is extremely important for many commercial applications of GO composites. For any degree of oxidation, GO does not disperse in PVA as a thermodynamic equilibrium product, whereas in PEG dispersions are only thermodynamically stable for highly oxidized GO. These findings-validated against the extensive literature on GO systems in organic solvents-furnish quantitative explanations for the empirically unpredictable aggregation characteristics of GO and provide computational methods to design directed synthesis routes for diverse self-assemblies and applications.

Graphene continues to attract considerable attention from the materials science community through its potential for improving the mechanical properties of polymer thermosets, yet there remains considerable uncertainty over the underlying mechanisms. The effect of introducing graphene sheets to a typical thermosetting polymer network on mechanical behaviour is explored here through concurrently coupling molecular dynamics with a finite element solver. In this multiscale approach, Graphene is observed to act in two ways: as passive microscopic defects, dispersing crack propagation (high deformation); and as active geometric constraints, impeding polymer conformational changes (low deformation). By contrast, single-scale atomistic simulations alone predict little measurable difference in the properties of the graphene-enhanced epoxy resins as compared with the pure polymer case. The multiscale model predicts that epoxy resins reinforced with graphene nanoparticles exhibit enhanced overall elastoplastic properties, reducing strain energy dissipation by up to 70%. Importantly, this is only observed when taking into account the complex boundary conditions, mainly invoking shear, arising from coupling physics on length scales separated by five orders of magnitude. The approach herein dearly highlights a novel role of graphene nanoparticles in actively constraining the surrounding polymer matrix, impeding local dissipative mechanisms, and resisting shear deformation.

Multiscale simulations are an essential computational method in a range of research disciplines, and provide unprecedented levels of scientific insight at a tractable cost in terms of effort and compute resources. To provide this, we need such simulations to produce results that are both robust and actionable. The VECMA toolkit (VECMAtk), which is officially released in conjunction with the present paper, establishes a platform to achieve this by exposing patterns for verification, validation and uncertainty quantification (VVUQ). These patterns can be combined to capture complex scenarios, applied to applications in disparate domains, and used to run multiscale simulations on any desktop, cluster or supercomputing platform.

Mechanical exfoliation techniques are widely used to create high quality graphene samples for analytical use. Increasingly, mechanical methods are used to create large quantities of graphene, yet there is surprisingly little molecular insight into the mechanisms involved. We study the exfoliation of graphene with sticky tape using molecular dynamics. This is made possible by using a recently developed molecular dynamics forcefield, GraFF, to represent graphene's dispersion interactions. For nano-sized flakes we observe two different mechanisms depending on the polymer-adhesive used. A peeling mechanism which mixes shearing and normal mode exfoliation promotes synthesis of graphene rather than many-layered graphite. Armed with this new chemical insight we discuss the experimental methods that could preferentially produce graphene by mechanical exfoliation. We also introduce a mathematical model describing the repeated exfoliation of graphite.

Graphene oxide (GO) is an amorphous 2D material, which has found widespread use in the fields of chemistry, physics, and materials science due to its similarity to graphene with the benefit of being far easier to synthesize and process. However, the standard of GO characterization is very poor because its structure is irregular, being sensitive to the preparation method, and it has a propensity to transform due to its reactive nature. Atomistic simulations of GO are common, but the nanostructure in these simulations is often based on little evidence or thought. We have written a computer program to generate graphene oxide nanostructures for general purpose atomistic simulation based on theoretical and experimental evidence. The structures generated offer a significant improvement to the current standard of randomly placed oxidized functional groups and successfully recreate the two-phase nature of oxidized and unoxidized graphene domains observed in microscopy experiments. Using this model, we reveal new features of GO structure and predict that a critical point in the oxidation reaction exists as the oxidized region reaches a percolation threshold. Even by a conservative estimate, we show that, if the carbon to oxygen ratio is kept above 6, a continuous aromatic network will remain, preserving many of graphene’s desirable properties, irrespective of the oxidation method or the size distribution of graphene sheets. This is an experimentally achievable degree of oxidation and should aid better GO synthesis for many applications.

Graphite's lubricating properties due to the "weak" interactions between individual layers have long been known. However, these interactions are not weak enough to allow graphite to readily exfoliate into graphene on a large scale. Separating graphite layers down to a single sheet is an intense area of research as scientists attempt to utilize graphene's superlative properties. The exfoliation and processing of layered materials is governed by the friction between layers. Friction on the macroscale can be intuitively understood, but there is little understanding of the mechanisms involved in nanolayered materials. Using molecular dynamics and a new forcefield, graphene's unusual behavior in a superlubric state is examined, and the energy dissipated between two such surfaces sliding past each other is shown. The dependence of friction on temperature and surface roughness is described, and agreement with experiment is reported. The accuracy of the simulated behavior enables the processes that drive exfoliation of graphite into individual graphene sheets to be described. Taking into account the friction between layers, a peeling mechanism of exfoliation is predicted to be of lower energy cost than shearing.