Teracluster LSSC-II - Its Designing Principles and Applications in Large Scale Numerical Simulations

The teracluster LSSC-II installed at the State Key Laboratory of Scientific and Engineering Computing, Chinese Academy of Sciences is one of the most powerful PC clusters in China. It has a peek performance of 2Tflops. With a Linpack performance of 1.04Tflops, it is ranked at the 43rd place in the 20th TOP500 List (November 2002), 51st place in the 21st TOP500 List (June 2003), and the 82nd place in the 22nd TOP500 List (November 2003) with a new Linpack performance of 1.3Tflops. In this paper, we present some design principles of this cluster, as well as its applications in some largescale numerical simulations.

Numerical Simulation on the Micro-Scale Bending Induced by Pulsed Laser Beam

利用自行研制的含热传导、冲击动力学大、变形有限元程序,模拟了小尺寸梁在脉冲激光加热条件下的变形过程。在此基础上,利用商用程序模拟了冷却及残余应力的产生,研究了激光参数（强度及分布）等对于微弯曲的影响。数值模拟结果与文献中的实验观察相吻合。

Equivalency and Locality in Nano-Scale Measurement

Two principal problems of equivalency and locality in nano-scale measurement are considered in this paper. The conventional measurements of force and displacement are always closely related to the equivalency problem between the measuremental results by experimental system and the real physical status of the sample, and the locality of the mechanical quantities to be measured. There are some noticeable contradictions in nano-scale measurements induced by the two problems. In this paper, by utilizing a coupled molecular-continuum method, we illustrate the important effects of the two principal problems in atomic force microscopy (AFM) measurements on nano-scale. Our calculations and analysis of these typical mechanical measurement problems suggest that in nano-meter scale measurements, the two principal problems must be carefully dealt with. The coupled molecular-continuum method used in this paper is very effective in solving these problems on nano-scale.

Microdamage Evolution, Energy Dissipation and Their Trans-Scale Effects on Macroscopic Failure

The process of damage evolution concerns various scales, from micro- to macroscopic. How to characterize the trans-scale nature of the process is on the challenging frontiers of solid mechanics. In this paper, a closed trans-scale formulation of damage evolution based on statistical microdamage mechanics is presented. As a case study, the damage evolution in spallation is analyzed with the formulation. Scaling of the formulation reveals that the following dimensionless numbers: reduced Mach number M, damage number S, stress wave Fourier number P, intrinsic Deborah number D*, and the imposed Deborah number De*, govern the whole process of deformation and damage evolution. The evaluation of P and the estimation of temperature increase show that the energy equation can be ignored as the first approximation in the case of spallation. Hence, apart from the two conventional macroscopic parameters: the reduced Mach number M and damage number S, the damage evolution in spallation is mainly governed by two microdamage-relevant parameters: the Deborah numbers D* and De*. Higher nucleation and growth rates of microdamage accelerate damage evolution, and result in higher damage in the target plate. In addition, the mere variation in nucleation rate does not change the spatial distribution of damage or form localized rupture, while the increase of microdamage growth rate localizes the damage distribution in the target plate, which can be characterized by the imposed Deborah number De*.

Critical Sensitivity and Trans-Scale Fluctuations in Catastrophic Rupture

Rupture in the heterogeneous crust appears to be a catastrophe transition. Catastrophic rupture sensitively depends on the details of heterogeneity and stress transfer on multiple scales. These are difficult to identify and deal with. As a result, the threshold of earthquake-like rupture presents uncertainty. This may be the root of the difficulty of earthquake prediction. Based on a coupled pattern mapping model, we represent critical sensitivity and trans-scale fluctuations associated with catastrophic rupture. Critical sensitivity means that a system may become significantly sensitive near catastrophe transition. Trans-scale fluctuations mean that the level of stress fluctuations increases strongly and the spatial scale of stress and damage fluctuations evolves from the mesoscopic heterogeneity scale to the macroscopic scale as the catastrophe regime is approached. The underlying mechanism behind critical sensitivity and trans-scale fluctuations is the coupling effect between heterogeneity and dynamical nonlinearity. Such features may provide clues for prediction of catastrophic rupture, like material failure and great earthquakes. Critical sensitivity may be the physical mechanism underlying a promising earthquake forecasting method, the load-unload response ratio (LURR).

Micro-scale Mechanics of the Surface-Nanocrystalline Al-Alloy Material

Based on the microscopic observations and measurements, the mechanical behavior of the surface-nanocrystallized Al-alloy material at microscale is investigated experimentally and theoretically. In the experimental research, the compressive stress-strain curves and the hardness depth curves are measured. In the theoretical simulation, based on the material microstructure characteristics and the experimental features of the compression and indentation, the microstructure cell models are developed and the strain gradient plasticity theory is adopted. The material compressive stress-strain curves and the hardness depth curves-are predicted and simulated. Through comparison of the experimental results with the simulation results, the material and model parameters are determined.

Preliminary Investigation of Fluid Flow in Meso-Scale Rectangular Tube

The fluid flow associated with micro and meso scale devices is currently of interest. Experiments were performed to study the fluid flow in meso-scale channels. A straight flow tube was fabricated with 1.0x4.0mm^2 in rectangular cross section and 200mm in length, which was made of quartz for flow visualization and PIV measurements. Reynolds numbers were ranged from 311 to over 3105. The corresponding pressure drop was from 0.65KPa to over 16.58KPa between the inlet and outlet of the tube. The micro PIV was developed to measure the velocity distribution in the tube. A set of microscope object lens was mounted ahead of CCD camera to obtain optimized optical magnification on the CCD chip. The velocity distributions near the outlet of the tube were measured to obtain full-developed flow. A CW laser beam was focused directly on the test section by a cylinder lens to form a small light sheet. Thus, high power density of light was formed on the view region. It is very important to the experiment while the velocity of the flow reaches to a few meters per second within millimeter scale. In this case, it is necessary to reduce exposure time to microseconds for PIV measurements. In the present paper, the experimental results are compared with the classical theories.

Trans-Scale Mechanics: Looking For The Missing Links Between Continuum And Micro/Nanoscopic Reality

Problems involving coupled multiple space and time scales offer a real challenge for conventional frameworks of either particle or continuum mechanics. In this paper, four cases studies (shear band formation in bulk metallic glasses, spallation resulting from stress wave, interaction between a probe tip and sample, the simulation of nanoindentation with molecular statistical thermodynamics) are provided to illustrate the three levels of trans-scale problems (problems due to various physical mechanisms at macro-level, problems due to micro-structural evolution at macro/micro-level, problems due to the coupling of atoms/molecules and a finite size body at micro/nano-level) and their formulations. Accordingly, non-equilibrium statistical mechanics, coupled trans-scale equations and simultaneous solutions, and trans-scale algorithms based on atomic/molecular interaction are suggested as the three possible modes of trans-scale mechanics.

Toughness Of Ni/Al2O3 Interfaces As Dependent On Micron-Scale Plasticity And Atomistic-Scale Separation

Ceramic/metal interfaces were studied that fail by atomistic separation accompanied by plastic dissipation in the metal. The macroscopic toughness of the specific Ni alloy/Al2O3 interface considered is typically on the order of ten times the atomistic work of separation in mode I and even higher if combinations of mode I and mode II act on the interface. Inputs to the computational model of interface toughness are: (i) strain gradient plasticity applied to the Ni alloy with a length parameter determined by an indentation test, and (ii) a potential characterizing mixed mode separation of the interface fit to atomistic results. The roles of the several length parameters in the strain gradient plasticity are determined for indentation and crack growth. One of the parameters is shown to be of dominant importance, thus establishing that indentation can be used to measure the relevant length parameter. Recent results for separation of Ni/Al2O3 interfaces computed by atomistic methods are reviewed, including a set of results computed for mixed mode separation. An approximate potential fit to these results is characterized by the work of separation, the peak separation stress for normal separation and the traction-displacement relation in pure shearing of the interface. With these inputs, the model for steady-state crack growth is used to compute the toughness of the interface under mode I and under the full range of mode mix. The effect of interface strength and the work of separation on macroscopic toughness is computed. Fundamental implications for plasticity-enhanced toughness emerge.

Large Scale Fabrication Of Periodical Bowl-Like Micropatterns Of Single Crystal Zno

Nanostructured ZnO materials are of great significance for their potential applications in photoelectronic devices, light-emitting displays, catalysis and gas sensors. In this paper, we report a new method to produce large area periodical bowl-like micropatterns of single crystal ZnO through aqueous-phase epitaxial growth on a ZnO single crystal substrate. A self-assembled monolayer of polystyrene microspheres was used as a template to confine the epitaxial growth of single crystal ZnO from the substrate, while the growth morphology was well controlled by citrate anions. Moreover, it was found that the self-assembled monolayer of colloidal spheres plays an important role in reduction of the defect density in the epitaxial ZnO layer. Though the mechanism is still open for further investigation, the present result indicates a new route to suppress the dislocations in the fabrication of single crystal ZnO film. A predicable application of this new method is for the fabrication of two-dimensional photonic crystal structures on light emitting diode surfaces.

Large Scale Numerical Simulations For Multi-Phase Fluid Dynamics With Moving Interfaces

This short communication presents our recent studies to implement numerical simulations for multi-phase flows on top-ranked supercomputer systems with distributed memory architecture. The numerical model is designed so as to make full use of the capacity of the hardware. Satisfactory scalability in terms of both the parallel speed-up rate and the size of the problem has been obtained on two high rank systems with massively parallel processors, the Earth Simulator (Earth simulator research center, Yokohama Kanagawa, Japan) and the TSUBAME (Tokyo Institute of Technology, Tokyo, Japan) supercomputers.

Effects Of Subgrid-Scale Modeling On Lagrangian Statistics In Large-Eddy Simulation

The application of large-eddy simulation (LES) to turbulent transport processes requires accurate prediction of the Lagrangian statistics of flow fields. However, in most existing SGS models, no explicit consideration is given to Lagrangian statistics. In this paper, we focus on the effects of SGS modeling on Lagrangian statistics in LES ranging from statistics determining single-particle dispersion to those of pair dispersion and multiparticle dispersion. Lagrangian statistics in homogeneous isotropic turbulence are extracted from direct numerical simulation (DNS) and the LES with a spectral eddy-viscosity model. For the case of longtime single-particle dispersion, it is shown that, compared to DNS, LES overpredicts the time scale of the Lagrangian velocity correlation but underpredicts the Lagrangian velocity fluctuation. These two effects tend to cancel one another leading to an accurate prediction of the longtime turbulent dispersion coefficient. Unlike the single-particle dispersion, LES tends to underestimate significantly the rate of relative dispersion of particle pairs and multiple-particles, when initial separation distances are less than the minimum resolved scale due to the lack of subgrid fluctuations. The overprediction of LES on the time scale of the Lagrangian velocity correlation is further confirmed by a theoretical analysis using a turbulence closure theory.

Derivation Of A Nonlinear Reynolds Stress Model Using Renormalization Group Analysis And Two-Scale Expansion Technique

Adopting Yoshizawa's two-scale expansion technique, the fluctuating field is expanded around the isotropic field. The renormalization group method is applied for calculating the covariance of the fluctuating field at the lower order expansion. A nonlinear Reynolds stress model is derived and the turbulent constants inside are evaluated analytically. Compared with the two-scale direct interaction approximation analysis for turbulent shear flows proposed by Yoshizawa, the calculation is much more simple. The analytical model presented here is close to the Speziale model, which is widely applied in the numerical simulations for the complex turbulent flows.

A second-order dynamic subgrid-scale stress model

A second-order dynamic model based on the general relation between the subgrid-scale stress and the velocity gradient tensors was proposed. A priori test of the second-order model was made using moderate resolution direct numerical simulation date at high Reynolds number ( Taylor microscale Reynolds number R-lambda = 102 similar to 216) for homogeneous, isotropic forced flow, decaying flow, and homogeneous rotating flow. Numerical testing shows that the second-order dynamic model significantly improves the correlation coefficient when compared to the first-order dynamic models.