Simulation Techniques for Cosmological Simulations

Modern cosmological observations allow us to study in great detail the evolution and history of the large scale structure hierarchy. The fundamental problem of accurate constraints on the cosmological parameters, within a given cosmological model, requires precise modelling of the observed structure. In this paper we briefly review the current most effective techniques of large scale structure simulations, emphasising both their advantages and shortcomings. Starting with basics of the direct N-body simulations appropriate to modelling cold dark matter evolution, we then discuss the direct-sum technique GRAPE, particle-mesh (PM) and hybrid methods, combining the PM and the tree algorithms. Simulations of baryonic matter in the Universe often use hydrodynamic codes based on both particle methods that discretise mass, and grid-based methods. We briefly describe Eulerian grid methods, and also some variants of Lagrangian smoothed particle hydrodynamics (SPH) methods.     

绳驱动机器人的解耦分析和位置控制仿真

为提高机器人的负载能力,基于几何分析和螺旋理论的基本原理,对绳驱动 拟人臂机器人的构型设计、钢绳传动方式和运动学控制问题进行了深入的分析.首先针对绳 驱动走线方式的多样性选择,设计了一种具有肘、腕关节耦合和肩、肘关节耦合的绳驱动传 动机构;然后以肩、肘、腕7-DOF关节为中间媒介对机器人末端和驱动钢绳间的运动学关系 进行了分析,即先考虑机器人末端位姿和各关节角度之间的关系,再研究各关节角度和其相 应的驱动钢绳绳长之间的关系,从而得到绳驱动机器人的运动学解;最后进行了仿真研究. 仿真结果显示机器人的关节角度和绳长变化曲线十分平滑,末端的实际轨迹跟踪理想轨迹的 误差非常微小,说明建议的算法正确,绳驱动机器人运行理想,可以有效克服常规机器 人自重过大带来的负面影响.      

车身逆向设计中点云多视拼合技术

针对整车车身点云空间尺寸较大,数据量庞大,还原精度要求高等特点,提出基于骨架点的点云拼合算法,算法的基本思想是构造整车模型的骨架点和分块点云的mark点,由全等三角形法则搜索骨架点与mark点的映射关系,应用加速迭代的改进ICP(Iterative Closest Point)算法拼合整车点云.某厂轻卡整车点云的拼合实例证明,该算法拼合精度高,运算速度快,是拼合整车点云行之有效的方法.      

层流平板摩擦阻力的数值计算

为提高层流摩擦阻力数值计算的精度,采用数值求解N-S方程的方法,对层流平板摩擦阻力进行了计算.计算中采用4种先进上风格式,同时配合以14种不同的壁面法向网格分布和5种常用限制器,并将计算结果与布拉修斯解进行对比.结果表明,van Leer的FVS(Flux Vector Splitting)格式数值耗散大,对网格分布和限制器敏感,不适合用于摩擦阻力计算.限制器和计算格式的粘性分辨率越高,摩擦阻力的计算精度越高.在计算量允许的情况下,可以减小壁面法向第1层网格高度,以提高计算精度,但应同时兼顾网格长宽比.对于摩擦阻力计算,边界层内应至少布置4个网格点.      

Intrinsic time scale for reconnection on the dayside magnetopause

Recently much attention has been focused on the transient behavior of the magnetopause in response to pressure pulses and southward fluctuations of the interplanetary magnetic field. We examine the motion of the magnetopause behind the foreshock and conclude that this motion is affected by foreshock pressure variations but not by fluctuations in the direction of the magnetic field. Neither magnetopause erosion nor flux transfer event occurrence is controlled by the foreshock. On the contrary, flux transfer events occur at times of steady IMF and thier quasi-periodic behavior is controlled by the magnetopause or the magnetosphere and is not driven by the external boundary conditions. Since flux transfer events are clearly due to reconnection, this observation implies that the IMF must be southward some time perhaps as long as 7 minutes before flux transfer begins.     

Response of Hainan GPS ionospheric scintillations to the different strong magnetic storm conditions

Using the GPS ionospheric scintillation data at Hainan station (19.5°N, 109.1°E) in the eastern Asia equatorial regions and relevant ionospheric and geomagnetic data from July 2003 to June 2005, we investigate the response of L-band ionospheric scintillation activity over this region to different strong magnetic storm conditions (Dst < −100 nT) during the descending phase of the solar cycle. These strong storms and corresponding scintillations mainly took place in winter and summer seasons. When the main phase developed rapidly and reached the maximum near 20–21 LT (LT = UT + 8) after sunset, scintillations might occur in the following recovery phase. When the main phase maximum occurred shortly after midnight near 01–02 LT, following the strong scintillations in the pre-midnight main phase, scintillations might also occur in the post-midnight recovery phase. When the main phase maximum took place after 03 LT to the early morning hours no any scintillation could be observed in the latter of the night. Moreover, when the main phase maximum occurred during the daytime hours, scintillations could also hardly be observed in the following nighttime recovery phase, which might last until the end of recovery phase. Occasionally, scintillations also took place in the initial phase of the storm. During those scintillations associated with the nighttime magnetic storms, the height of F layer base (h’F) was evidently increased. However, the increase of F layer base height does not always cause the occurrence of scintillations, which indicates the complex interaction of various disturbance processes in ionosphere and thermosphere systems during the storms.     

Spectroscopic analysis of the solar flare event on 2002 August 3 with the use of RHESSI and RESIK data

We use simultaneous observations from RESIK and RHESSI instruments to compare plasma properties of a major solar flare in its rise and gradual phase. This event occurred on 2002 August 3 (peak time at 19:06 UT). The flare had a very good coverage with RESIK data and well-resolved soft and hard X-ray sources were seen in RHESSI images. Spectra of X-ray radiation from RHESSI images are studied and compared with RESIK measurements in different flare phases. Result shows large differences in flare morphology and spectra between flare rise and gradual phase.     

A steady-state model for the D- to F-region of the polar cap

For obvious reasons the ionosphere of the polar cap, surrounded by the auroral zone, is only poorly investigated. Even ionosonde data are very scant from geomagnetic latitudes beyond 70°. Since 1997 the European incoherent scatter radar facility EISCAT has an additional installation on Svalbard and has been providing electron density data nearly continuously ever since. These measurements which mainly cover the E- and F-regions are supplemented by rocket data from Heiss Island at a comparable magnetic latitude; these data are more sporadic, but cover lower altitudes and densities. A provisional, steady-state, neural network-based model is presented which uses the data of both sites.     

Roe格式中不同类型熵修正性能分析

针对Roe格式中较为流行的3类熵修正:Muller型、Harten-Yee型和Harten-Hyman型熵修正,从理论分析入手, 辅以Euler方程为控制方程的激波管问题,前台阶流动和运动激波的双马赫反射3个数值实验,对不同熵修正的性能做了深入研究,得出如下结论: Muller型和Harten-Yee型熵修正方法作用于激波和非物理的膨胀激波2种情况:激波情况下引入的数值耗散,能根本改善"Carbuncle"现象,膨胀激波时其数值粘性也足以使膨胀激波得以耗散;而类Harten-Hyman型熵修正只对膨胀过程起修正作用,对激波情况无效,不能改善"Carbuncle"现象,该类熵修正的数值粘性较小,不足以使膨胀过程正确求解;直接采用 δ 值代替特征值的特征值修正方法效果好于传统的特征值修正方法.