Simulation Of Multiscale Systems Research Articles

Elastic crack propagation model for crystalline solids using a self-consistent coupled atomistic–continuum framework

Deformation and failure processes of crystalline materials are governed by complex phenomena at multiple scales. It is necessary to couple these scales for physics-based modeling of these phenomena, while overcoming limitations of modeling at individual scales. To address this issue, this paper develops self-consistent elastic constitutive and crack propagation relations of crystalline materials containing atomic scale cracks, from observations made in a concurrent multi-scale simulation system coupling atomistic and continuum domain models. The concurrent multi-scale model incorporates a finite temperature atomistic region containing the crack, a continuum region represented by a self-consistent crystal elasticity constitutive model, and a handshaking interphase region. Atomistic modeling is done by the molecular dynamics code LAMMPS, while continuum modeling is conducted by the finite element method. For single crystal nickel a nonlinear and nonlocal crystal elasticity constitutive relation is derived, consistent with the atomic potential function. An efficient, staggered solution scheme with parallel implementation is designed for the coupled problem. The atomistic–continuum coupling is achieved by enforcing geometric compatibility and force equilibrium in the interphase region. Quantitative analyses of the crack propagation process focuses on size dependence, strain energy release rate, crack propagation rate and degradation of the local stiffness. The self-consistent constitutive and crack propagation relations, derived from the concurrent model simulation results are validated by comparing results from the concurrent and full FE models. Excellent accuracy and enhanced efficiency are observed in comparison with pure MD and concurrent model results.

Multi-scale prediction of the geometrical deviations of the surface finished by five-axis ball-end milling

Surface quality and accuracy are the main factors which affect the performance and life cycle of the products. Due to the complexity of the machining process, it is difficult to evaluate the machined surface real time. Simulation of the machining process became the main method to predict and control the quality of the machined surface. This article developed a multi-scale simulation system to predict the overall geometrical features of the milled surface. The effects of locating errors, geometrical errors of the machine tool and tool deflections on the quality of the machined surface are included in the proposed model. Also, different strategies are employed to evaluate the macro-scale and micro-scale geometrical deviations of the machined surface to balance the time cost and accuracy. In comparison with the traditional method, both the form deviations and roughness feature of the machined surface can be predicted. Since the static and dynamic properties of the machining system were considered, both the stable and unstable cutting conditions can be analyzed by using the proposed method. At the end of this article, case studies are carried out to validate the proposed method. The effects of the locating errors, geometrical errors of the machine tool and cutting parameters on the quality of the machined surface are analyzed. The significance of their influences on the quality of the machined surface was investigated.

Development of a patient-specific multi-scale simulation system for prediction of hemodynamics in pre- and post-operation

Development of a patient-specific multi-scale simulation system for prediction of hemodynamics in pre- and post-operation

Higher Efficiency Modeling of Surface Location Errors by Using a Multi-scale Milling Simulation

The dynamic process behavior during milling often leads to vibrations or even chatter and, thereby, to surface location errors. In this paper, a multi-scale simulation system is presented, which can predict these errors efficiently by combining a dexel-based workpiece representation with a CSG-based process model and an oscillator-based system for the process dynamics. The different modeling techniques are described, and the run-time of the combined system is compared to a single-scale system in which the CSG-based model is used for every process step. Furthermore, the simulation is validated by examining the generated surface of a micromachined workpiece.

A Multiscale Simulation System for the Prediction of Drug-Induced Cardiotoxicity

The preclinical assessment of drug-induced ventricular arrhythmia, a major concern for regulators, is typically based on experimental or computational models focused on the potassium channel hERG (human ether-a-go-go-related gene, K(v)11.1). Even if the role of this ion channel in the ventricular repolarization is of critical importance, the complexity of the events involved make the cardiac safety assessment based only on hERG has a high risk of producing either false positive or negative results. We introduce a multiscale simulation system aiming to produce a better cardiotoxicity assessment. At the molecular scale, the proposed system uses a combination of docking simulations on two potassium channels, hERG and KCNQ1, plus three-dimensional quantitative structure-activity relationship modeling for predicting how the tested compound will block the potassium currents IK(r) and IK(s). The obtained results have been introduced in electrophysiological models of the cardiomyocytes and the ventricular tissue, allowing the direct prediction of the drug effects on electrocardiogram simulations. The usefulness of the whole method is illustrated by predicting the cardiotoxic effect of several compounds, including some examples in which classic hERG-based models produce false positive or negative results, yielding correct predictions for all of them. These results can be considered a proof of concept, suggesting that multiscale prediction systems can be suitable for being used for preliminary screening in lead discovery, before the compound is physically available, or in early preclinical development when they can be fed with experimentally obtained data.

Materials integrity in microsystems: a framework for a petascale predictive-science-based multiscale modeling and simulation system

Microsystems have become an integral part of our lives and can be found in homeland security, medical science, aerospace applications and beyond. Many critical microsystem applications are in harsh environments, in which long-term reliability needs to be guaranteed and repair is not feasible. For example, gyroscope microsystems on satellites need to function for over 20 years under severe radiation, thermal cycling, and shock loading. Hence a predictive-science-based, verified and validated computational models and algorithms to predict the performance and materials integrity of microsystems in these situations is needed. Confidence in these predictions is improved by quantifying uncertainties and approximation errors. With no full system testing and limited sub-system testings, petascale computing is certainly necessary to span both time and space scales and to reduce the uncertainty in the prediction of long-term reliability. This paper presents the necessary steps to develop predictive-science-based multiscale modeling and simulation system. The development of this system will be focused on the prediction of the long-term performance of a gyroscope microsystem. The environmental effects to be considered include radiation, thermo-mechanical cycling and shock. Since there will be many material performance issues, attention is restricted to creep resulting from thermal aging and radiation-enhanced mass diffusion, material instability due to radiation and thermo-mechanical cycling and damage and fracture due to shock. To meet these challenges, we aim to develop an integrated multiscale software analysis system that spans the length scales from the atomistic scale to the scale of the device. The proposed software system will include molecular mechanics, phase field evolution, micromechanics and continuum mechanics software, and the state-of-the-art model identification strategies where atomistic properties are calibrated by quantum calculations. We aim to predict the long-term (in excess of 20 years) integrity of the resonator, electrode base, multilayer metallic bonding pads, and vacuum seals in a prescribed mission. Although multiscale simulations are efficient in the sense that they focus the most computationally intensive models and methods on only the portions of the space–time domain needed, the execution of the multiscale simulations associated with evaluating materials and device integrity for aerospace microsystems will require the application of petascale computing. A component-based software strategy will be used in the development of our massively parallel multiscale simulation system. This approach will allow us to take full advantage of existing single scale modeling components. An extensive, pervasive thrust in the software system development is verification, validation, and uncertainty quantification (UQ). Each component and the integrated software system need to be carefully verified. An UQ methodology that determines the quality of predictive information available from experimental measurements and packages the information in a form suitable for UQ at various scales needs to be developed. Experiments to validate the model at the nanoscale, microscale, and macroscale are proposed. The development of a petascale predictive-science-based multiscale modeling and simulation system will advance the field of predictive multiscale science so that it can be used to reliably analyze problems of unprecedented complexity, where limited testing resources can be adequately replaced by petascale computational power, advanced verification, validation, and UQ methodologies.