提高光子能量沉积计算效率的几种技巧

MCNP程序提供了计算光子沉积能的两种方法, 分别是F6和*F8。 当次级电子射程与网格尺度比较接近时, *F8的精度比F6高, 但计算效率比F6低很多。 分析比较了几种提高计算效率的技巧, 分别是: 网格大于10倍电子射程用F6; F6与*F8的联合使用; 改变电子能量子步数; 关掉δ电子; 次级电子自适应截断等。 数值模拟表明, F6与*F8的联合使用以及次级电子自适应截断这两种技巧对精度与效率兼顾得很好。 F6 and *F8 are two methods to calculate energy deposition of photon by MCNP program. If the dimension of grid is almost the same long as the range of secondary electron, the precision of *F8 is higher than that of F6, but the efficiency of F6 is greatly higher than that of *F8. This paper presents five techniques to increase the efficiency,namely: F6 is use to the model with grid dimension bigger than 10 ranges of secondary electron; combination of F6 and *F8; change the substeps of electron; turn off knock on electron; self adaption cutoff secondary electron. The two techniques, combination of F6 and *F8,self adaption cutoff secondary electron, can provide a good precision and efficiency.     

强爆炸火球辐射流体自适应网格高精度数值模拟

提出一种自适应网格方法,应用于基于Euler方法的强爆炸辐射流体高精度数值求解。通过与Zinn数值结果对比,验证该方法的正确性。研究自适应网格对冲击波和光辐射输出模拟精度的影响,对比不同网格尺度下的计算耗时。在相同的条件下,使用自适应网格与均匀网格加密3倍得到的冲击波超压分布、光辐射输出演化接近,计算效率提升约8.5倍。由此可知,提出的自适应网格方法可用于强爆炸问题的高精度数值模拟,显著地提升模拟精度和模拟效率。     

基于自适应非结构化网格的VOF方法

本文提出了一种基于自适应非结构化网格的VOF算法,根据相界面网格的相函数值对相界面网格进行自适应细化与合并,通过基于非结构化网格的界面构造方法构造相界面,在适量增加网格单元数量的情况下提高了计算的精度。该方法随着时间及相界面的变化无需重新整体生成网格,算法效率较高。经典算例的验证结果表明,本文自适应网格方法计算得到的结...     

块结构自适应网格上任意区域和任意界面的数值积分

研究块结构自适应网格上计算任意界面上和任意区域内的数值积分方法,其中任意界面和任意区域通过一个水平集函数表示.首先介绍在一致网格上任意界面上和任意区域内的数值积分方法.然后,将该方法推广到块结构的自适应网格上.数值算例表明,自适应网格方法有二阶精度.同一致网格方法相比,自适应网格方法显著地减少了计算机存储量的需求.     

自适应间断有限元方法求解三维欧拉方程

将龙格库塔间断有限元方法(RDDG)与自适应方法相结合,求解三维欧拉方程.区域剖分采用非结构四面体网格,依据数值解的变化采用自适应技术对网格进行局部加密或粗化,减少总体网格数目,提高计算效率.给出四种自适应策略并分析不同自适应策略的优缺点.数值算例表明方法的有效性.     

光子能量沉积计算的一种新方法

光子沉积能广泛地应用于放射医疗和辐射防护领域。 MCNP程序中*F8功能统计的是进出网格的光子与电子能量差。 *F8只能采用真实网格, 计算效率较低。 由于光子的能量沉积都是通过次级电子来完成的, 对光子能量沉积的计算可转换为次级电子能量沉积的计算。 文中据此给出了一种新的光子能量沉积统计方法, 该方法只统计次级电子能量沉积, 且可以采用虚拟网格计数。 新方法若采用真实网格, 计算精度与效率与*F8完全一样; 若采用虚拟网格, 新方法计算精度略低, 但几何建模简单, 计算效率较高。 Energy deposition distribution is very important to study radiotherapy or radioprotection. The *F8 method of MCNP program counts the energy loss of photon and electron together. Only real grid is allowed to *F8, so its computation efficiency is low. This paper gives a new calculation method for energy deposition of photon. Because the energy deposition of photon is accomplished by secondary electron, only electron is counted in the new method. The real collision of the electron is counted by the new method, whereas *F8 only counts the information of particle across the interface. Moreover, virtual grid is allowed in the new method. With real grid, the new method has the same precision and efficiency as *F8. If virtual grid is adopted by the new method, although the precision is slightly down, the efficiency is greatly increased.     

九点格式中网格边上切向流的计算——节点自适应加权格式

针对网格扭曲的不同情形,直接考虑网格边上切向流的离散.基于扩散方程法向流连续的条件,给出离散法向流的构造,导出扭曲网格上九点计算格式中网格边上离散切向流的表达式,从而推导出加权系数的计算公式,适应于各种扭曲的网格.数值结果表明,与九点格式中节点量简单加权的方法相比,基于网格边离散切向流的节点自适应加权九点格式的精度有明显改进,迭代求解次数减少,计算效率明显提高.     

求解双曲型守恒律方程的一类自适应多分辨方法

针对双曲型守恒律方程问题,发展一种有效的自适应多分辨分析方法.通过对嵌套网格上的数值解构造离散多分辨分析,建立小波系数与多层嵌套网格点之间的对应关系.对于小波系数较大的网格点采用高精度WENO格式计算,其余区域则直接采用多项式插值.数值试验表明,该方法在保持原规则网格方法的精度和分辨率的同时,显著地减少计算的CPU时间.     

三维空间二阶精度流体体积界面重构方法

给出三维空间网格模板含81个单元的最小二乘流体体积界面重构方法,并和Youngs方法及网格模板含125个单元的最小二乘流体体积界面重构方法进行比较.静态和动态的测试例子均表明:该方法能精确重构任意方向的平面界面,对C²光滑曲面它能达到二阶收敛精度.和网格模板含125个单元的最小二乘流体体积界面重构方法相比,在达到同样网格精度的条件下,减少了计算量,节省了计算时间,提高了计算效率.     

面向结构网格自适应并行计算的矩形区域求差集快速算法

结构网格自适应程序需要使用矩形区域求差集算法计算网格层间数据依赖关系和网格层嵌套关系.原有的矩形区域求差集算法时间复杂度较高,成为该类应用大规模并行计算可扩展性能瓶颈.本文利用分而治之的方法,构造近似线性时间复杂度的矩形区域求差集快速算法,并利用区域分解实现该算法的并行计算.分别针对规则矩形区域和多层自适应网格的非规则矩形区域求差集问题,验证该算法的效率.结果表明,该算法具有近似线性计算复杂度,对于大规模计算问题,加速效果显著.     

低能电子在水介质中的输运过程

考虑到低能电子(能量下限为1eV)在水中输运时的电离,激发,俘获以及超激发引起的自电离等非弹性散射机制,并且考虑到OH⁺,H⁺等自由基的产生和分布,运用Monte Carlo方法模拟了电子在水中输运的径迹结构,揭示了电子在低能情况下输运时单条轨道的空间分布结构特点(云团,团点和短径迹等空间分布实体)和包含在大量轨道中的径迹结构的统计性质,并通过对比截止能量取30和1eV两种情况,分析了30eV以下的低能电子的作用.     

Extension of high-order harmonic cutoff frequency by synthesizing the waveform of a laser field via the optimization of classical electron trajectory in the laser field

We theoretically investigate high-order harmonic generation by employing strong-field approximation （SFA） and present a new approach to the extension of the high-order harmonic cutoff frequency via an exploration of the dependence of high-order harmonic generation on the waveform of laser fields. The dependence is investigated via detailed analysis of the classical trajectories of the ionized electron moving in the continuum in the velocity-position plane. The classical trajectory consists of three sections （Acceleration Away, Deceleration Away, and Acceleration Back）, and their relationship with the electron recollision energy is investigated. The analysis of classical trajectories indicates that, besides the final （Acceleration Back） section, the electron recollision energy also relies on the previous two sections. We simultaneously optimize the waveform in all three sections to increase the electron recollision energy, and an extension of the cutoff frequency up to Ip ＋ 20.26Up is presented with a theoretically synthesized waveform of the laser field.     

Laser phase control of high-order harmonic generation at large internuclear distance: the H+ -H2+ system

Exact (Born-Oppenheimer) 3-D numerical solutions of the time-dependent Schr?dinger equation are obtained for the one electron linear H+-H2+ atom-molecule system at large internuclear distance R in interaction with two-cycles intense (I>10(14) W cm(-2)) 800 nm laser pulses. High-order harmonic generation (HHG) spectra are obtained with an energy cutoff larger than the atomic maximum of I(p)+3U(p), where I(p) is the ionization potential and U(p) is the ponderomotive energy. At large R, this extended cutoff is shown to be related to the nature of electron transfer, whose direction is shown to depend critically on the carrier-envelope phase (CEP) of the ultrashort pulse. Constructive and destructive interferences in the HHG spectrum resulting from coherent superpositions of electronic states in the H+-H2+ system are interpreted in terms of multiple electron trajectories extracted from a time profile analysis.     

Infrared catastrophe in a two-quasiparticle collision integral

Relaxation of a nonequilibrium state in a disordered metal with a spin-dependent electron energy distribution is considered. The collision integral due to the electron-electron interaction is computed within the approximation of a two-quasiparticle scattering. It is shown that the spin-flip scattering processes with a small energy transfer may lead to the divergence of the collision integral for a quasi one-dimensional wire. This divergence is present only for a spin-dependent electron energy distribution that corresponds to the total electron spin magnetization M = 0 and only for nonzero interaction in the triplet channel. In this case, a nonperturbative treatment of the electron-electron interaction is needed to provide an effective infrared cutoff.     

Kinetics of the collisionless expansion of spherical nanoplasmas

The collisionless expansion of spherical plasmas composed of cold ions and hot electrons is analyzed using a novel kinetic model, with special emphasis on the influence of the electron dynamics. Simple, general laws are found, relating the relevant expansion features to the initial conditions of the plasma, determined from a single dimensionless parameter. A transition is identified in the behavior of the ion energy spectrum, which is monotonic only for high electron temperatures, otherwise exhibiting a local peak far from the cutoff energy.     

High-order above-threshold ionization in a laser field: Influence of the ionization potential on the high-energy cutoff

High-energy above-threshold ionization by a strong linearly polarized laser field is analyzed using the strong-field approximation and the quantum-orbit formalism. An analytical formula for the cutoff is derived. The maximal electron kinetic energy is 10.007U _P + 0.538I _P. The first term with the ponderomotive energy U _P is equivalent to the known classical cutoff law, while the second term is a new result and depends on the ionization potential I _P. The validity of this formula is confirmed by comparison with the numerical results for the high-energy electron spectra.     

Systematic approach to cutoff frequency selection in continuous-wave electron paramagnetic resonance imaging

This article describes a systematic method for determining the cutoff frequency of the low-pass window function that is used for deconvolution in two-dimensional continuous-wave electron paramagnetic resonance (EPR) imaging. An evaluation function for the criterion used to select the cutoff frequency is proposed, and is the product of the effective width of the point spread function for a localized point signal and the noise amplitude of a resultant EPR image. The present method was applied to EPR imaging for a phantom, and the result of cutoff frequency selection was compared with that based on a previously reported method for the same projection data set. The evaluation function has a global minimum point that gives the appropriate cutoff frequency. Images with reasonably good resolution and noise suppression can be obtained from projections with an automatically selected cutoff frequency based on the present method.     

超短脉冲谐波辐射过程中的量子路径分析

本文数值计算了脉宽为半个光学周期的超短脉冲在不同载波包络相位下的谐波波谱,通过比较谐波波谱截止频率与电子通过主要路径可获得的最大动能的差异,研究了超短脉冲谐波辐射过程中的量子路径。结果显示,超短脉冲谐波辐射过程中电子获得动能的量子路径偏离主要量子路径,而且该偏离效应受载波包络相位的影响。     

超短脉冲谐波辐射过程中的量子路径分析

本文数值计算了脉宽为半个光学周期的超短脉冲在不同载波包络相位下的谐波波谱,通过比较谐波波谱截止频率与电子通过主要路径可获得的最大动能的差异,研究了超短脉冲谐波辐射过程中的量子路径。结果显示,超短脉冲谐波辐射过程中电子获得动能的量子路径偏离主要量子路径,而且该偏离效应受载波包络相位的影响。