• Journal of Advanced Marine Engineering and Technology
• /
• 제29권1호
• /
• pp.60-67
• /
• 2005

Generally. in the molecular dynamics simulations. the Verlet neighboring list algorithm is used for the reduction of a simulation time On the other hand. the application of the Verlet neighboring list forces the time evolution of a simulation system to follow an unrealistic path in a phase space. In equilibrium state, it does not matter with the simulation results because the individual molecule's motion is originally random and any effect due to a small deviation from a real time evolution can be completely ignored. However, if an unsteady state is involved. such a deviation may significantly affect to the results. That is, there is a Possibility that the simulation results Provide ones with any misleading data In this study we evaluated the effect due to the Verlet neighboring list in performing the simulation of a non-equilibrium state and suggested the method to avoid it.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2000년도 제15차 학술회의논문집
• /
• pp.494-494
• /
• 2000

This paper demonstrates that the largest Lyapunov exponent $\lambda$ of recurrent neural networks can be controlled by a gradient method. The method minimizes a square error $e_{\lambda}=(\lambda-\lambda^{obj})^2$ where $\lambda^{obj}$ is desired exponent. The $\lambda$ can be given as a function of the network parameters P such as connection weights and thresholds of neurons' activation. Then changes of parameters to minimize the error are given by calculating their gradients $\partial\lambda/\partialP$. In a previous paper, we derived a control method of $\lambda$via a direct calculation of $\partial\lambda/\partialP$ with a gradient collection through time. This method however is computationally expensive for large-scale recurrent networks and the control is unstable for recurrent networks with chaotic dynamics. Our new method proposed in this paper is based on a stochastic relation between the complexity $\lambda$ and parameters P of the networks configuration under a restriction. Then the new method allows us to approximate the gradient collection in a fashion without time evolution. This approximation requires only $O(N^2)$ run time while our previous method needs $O(N^{5}T)$ run time for networks with N neurons and T evolution. Simulation results show that the new method can realize a "stable" control for larege-scale networks with chaotic dynamics.

• Coupled systems mechanics
• /
• 제6권1호
• /
• pp.41-54
• /
• 2017

Shape-memory alloys (SMA) have interesting behaviors and important mechanical properties due to the solid-solid phase transformation. These phenomena are dominated by the evolution of microstructures. In recent years, the microstructures in SMAs have been studied extensively and modeled using molecular dynamics (MD) simulations. However, it remains difficult to identify the crystal variants in the simulation results, which consist of large numbers of atoms. In the present work, a method is developed to identify the austenite phase and the monoclinic martensite crystal variants in MD results. The transformation matrix of each lattice is calculated to determine the corresponding crystal variant. Evolution of the volume fraction of the crystal variants and the microstructure in Ni-Ti SMAs under thermal and mechanical boundary conditions are examined. The method is validated by comparing MD-simulated interface normals with theoretical solutions. In addition, the results show that, in certain cases, the interatomic potential used in the current study leads to inconsistent monoclinic lattices compared with crystallographic theory. Thus, a specific modification is applied and the applicability of the potential is discussed.

• 한국전산유체공학회:학술대회논문집
• /
• 한국전산유체공학회 2010년 춘계학술대회논문집
• /
• pp.517-523
• /
• 2010

Recently, the rapid evolution of computational fluid dynamics (CFD) has enabled its key role in industries and predictive sciences. From diverse research disciplines, however, are there strong needs for integrated analytical tools for multi-phenomena beyond simple flow simulation. Based on the concurrent simulation of multi-dynamics, multi-phenomena beyond simple flow simulation. Based on the concurrent simulation of multi-dynamics, multi-physics and multi-scale phenomena, the multi-phenomena CFD technology enables us to perform the flow simulation for integrated and complex systems. From the multi-phenomena CFD analysis, the high-precision analytical and predictive capacity can enhance the fast development of industrial technologies. It is also expected to further enhance the applicability of the simulation technique to medical and bio technology, new and renewable energy, nanotechnology, and scientific computing, among others.

• Journal of Astronomy and Space Sciences
• /
• 제33권4호
• /
• pp.257-264
• /
• 2016

A planet revolving around binary star system is a familiar system. Studies of these systems are important because they provide precise knowledge of planet formation and orbit evolution. In this study, a method to determine the evolution of an exoplanet revolving around a binary star system using different rates of stellar mass loss will be introduced. Using a hierarchical triple body system, in which the outer body can be moved with the center of mass of the inner binary star as a two-body problem, the long period evolution of the exoplanet orbit is determined depending on a Hamiltonian formulation. The model is simulated by numerical integrations of the Hamiltonian equations for the system over a long time. As a conclusion, the behavior of the planet orbital elements is quite affected by the rate of the mass loss from the accompanying binary star.

• Bulletin of the Korean Chemical Society
• /
• 제27권11호
• /
• pp.1825-1831
• /
• 2006

The influence of solvent viscosity on conformational dynamics of the heme-pocket, a small vacant site near the binding site of myoglobin (Mb) and hemoglobin (Hb), and playing a functionally important role by serving as a station in ligand binding and escape, was studied by probing time-resolved vibrational spectra of CO photodissociated from MbCO and HbCO in $D_2O$, 75 wt% glycerol/$D_2O$, and trehalose at 283 K. Two absorption bands ($B_1$ and $B_2$) of the sample in viscous solvents, arising from CO in the heme pocket, are very similar to those in $D_2O$. Two bands in Mb and Hb under all three solvents exhibit very similar nonexponential spectral evolution ($B_1$ band; blue shifting and broadening, $B_2$ band; narrowing with a negligible shifting), suggesting that in the present experimental time window of 100 ps, the extents of the spectral shift and narrowing is much influenced neither by the viscosity of solvent nor by the quaternary contact of Hb. Spectral evolution can be described by a biexponential function with a fast universal time constant of 0.52 ps and a slow time constant ranging from 13 to 32 ps. For both proteins in all three solvents majority of spectral evolution occurs with the fast universal time constant. The magnitude of the slow rate in the spectral shift of B1 band decreases with increasing solvent viscosity, indicating that it is influenced by global conformational change which is retarded in viscous solvent, thereby serve as a reporter of global conformational change of heme proteins after deligation.

• Journal of Photoscience
• /
• 제9권2호
• /
• pp.359-361
• /
• 2002

A green mutant of Rhodobacter sphaeroides 60I was acquired by chemical induction. The blue-shifted of the carotenoid absorption was found in the Light-harvesting complex II (LH2) of the mutant. With the excitation at different wavelength, we observed that the evolution of excited-state dynamics in LH2 of Rhodobacter sphaeroides 60I. The dynamical traces demonstrate a dominant absorption followed concomitantly by an ultrafast transmission increase and decay with 818nm excitation.

• Bulletin of the Korean Chemical Society
• /
• 제21권1호
• /
• pp.93-96
• /
• 2000

A diffusion-influenced pseudo-first order reversible reaction A + B ⇔C + B is investigated by the molecular dynamics (MD) simulation method. Theoretical finding that the temporal evolution of reactants [conditional probabilities] in the reversible system can be expressed by the irreversible survival probability with an effective rate parameter is confirmed even in the presence of solvent particles. We carry out molecular dynamics simulations for both the irreversible and the reversible cases to evaluate the survival and the conditional probabilities for each cases. When the resultant irreversible survival probability is inserted into the proposed relation, the conditional probabilities given by the simulation are exactly reproduced.

• 천문학회보
• /
• 제46권2호
• /
• pp.47.2-47.2
• /
• 2021

The present study is focused on origins of the flow and magnetic structure involved in the formation and eruption of a solar prominence. To clarify them, we performed an MHD simulation based on the 3-dimensional emerging flux tube (3DEFT) model, in which self-consistent evolution of a flow and magnetic field passing freely through the solar surface was obtained by seamlessly connecting subsurface dynamics with surface dynamics. By analyzing Lagrangian displacements of magnetized plasma elements, we demonstrate the flow structure which is naturally incorporated to the magnetic structure of the prominence formed via dynamic interaction between the flow and magnetic field.

• 대한기계학회:학술대회논문집
• /
• 대한기계학회 2008년도 추계학술대회A
• /
• pp.1922-1927
• /
• 2008

In the present study, deposition of discrete and small particles, which diameter is less than $1{\mu}m$, on a filter element was simulated by stochastic method. Trajectory of each particle was numerically solved by Langevin equation and Brownian random motion was treated by Brownian dynamics. Lattice Boltzmann method (LBM) was used to solve flow field around the filter collector and deposit layer. Interaction between flow field and deposit layer was obtained from a converged solution from an inner-loop calculation. Simulation method is properly validated and collection efficiency due to different filtration parameters are examined and discussed. Morphology of deposit layer and its evolution was visualized in terms of the particle size. The particle loaded effect on collection efficiency was also discussed.