混沌系统模型误差平均绝对误差增长过程研究

混沌系统的平均绝对误差增长最初是用来刻画初始值误差增长的,本文依照平均绝对误差增长的定义来研究模型误差的增长过程,获得了一些很有意义的结论．研究发现,在初期模型误差的平均绝对误差增长呈指数级增长,增长指数同模型的扰动相关,与真实系统最大Lyapunov指数没有直接关系．其后模型误差进入非线性增长过程,误差增长放缓,最终达到饱和．误差饱和值恒定,当真实系统和模型系统吸引子差别较小时,模型误差饱和值基本上同真实系统的初始值误差饱和值相等．利用上述研究结论可以求出模型的可预报期限,这在数值天气预报中具有重要的意义．进而利用模型的可预报期限可以对同一真实系统的不同模型进行评价,相对真实系统越精确的模型拥有更高的可预报期限．这对新模型的开发具有很强的指导作用．     

混沌系统可预报期限随初始误差变化规律研究

利用非线性误差增长理论计算了Logistic映射和Lorenz系统可预报期限随初始误差的变化,发现Logistic映射等简单混沌系统的可预报期限与初始误差的对数存在线性关系.在非线性误差增长理论的框架下,理论分析表明,平均误差增长达到一定值时,误差增长进入明显的非线性增长阶段,最终达到饱和;对于一个确定的混沌系统,在控制参数固定的情况下误差增长的饱和值也是固定的,因此可预报期限只依赖于初始误差. 在可预报期限与初始误差对数存在的线性函数关系式中,线性系数与最大Lyapunov指数有关,在已知混沌系统的最大 关键词： 非线性局部Lyapunov指数 可预报期限 初始误差 混沌系统     

Al2O3Hx(x=1—3)分子团簇的结构与光谱研究

用密度泛函理论(DFT)的B3lyp方法在6-311++g(d,p)水平上对Al₂O₃H_x(x=1—3)分子的几何构型, 电子结构, 振动频率等性质进行了系统研究. 并给出了它们可能基态结构的总能量(E_T), 零点能(E_z), 摩尔热容(C_v), 标准熵(S), 原子化能(ΔE_m), 垂直电离能(IP)及垂直电子亲和能(E_A). Al₂O₃H和Al₂O₃H₂分子可能的基态的几何构型都为平面结构. Al₂O₃H₃的两个可能为基态的几何构型都是在立体Al₂O₃(D_3h)的几何结构基础上加三个氢原子构成. 这三个分子的能量最低结构为Al₂O₃H(²A′)C_s, Al₂O₃H₂(¹A′) C_s, Al₂O₃H₃ (²A) C₁.     

基于混合漫射近似的空间分辨漫反射光学参量灵敏度的研究

根据空间分辨漫反射的双点源混合漫射近似模型, 推导了空间分辨漫反射率对生物组织吸收系数μ_a和有效散射系数μ'_s灵敏度的解析表示, 系统研究了在强吸收条件近光源区域吸收系数μ_a和有效散射系数μ'_s对漫反射光子分布的影响. 研究表明: 吸收系数的灵敏度随光源与探测器间距ρ的增加呈线性增长, 其斜率正比于(μ'_s/μ_a)^1/4, 比例系数约为1.4, 同时获得一个优化的探测距离ρ_opt, 距离光源约3.4个输运平均自由程, 在这个距离处有效散射系数的变化对测量吸收的影响最小. 这项研究对于生物组织的光学性质测量以及漫反射光谱技术的应用具有重要意义.     

自辐射场下PuO分子光谱研究

在相对论有效原子实势近似下, 以Pu为SDD基组、O为6-311+G^*基组, 采用优选的密度泛函 B3LYP方法, 研究了用电场摸拟钚本身产生自辐射场(-0.005—0.005 a.u.) 作用下氧化钚(PuO)基态分子的最高占据轨道(HOMO)能级E_H、最低空轨道(LUMO)能级E_L、能隙E_g和费米能级E_F. 结果表明: 在所加的电场范围内, E_H随着电场的增加均逐渐减少, E_F随着电场的增加均逐渐增大, E_g始终处于增大的趋势, 费米能级E_F上升, 占据轨道的电子难以被激发至空轨道而形成激发态, PuO分子在自辐射场中更趋于稳定, 可以阻止O₂, H₂等扩散到表面内层而腐蚀钚表面, 有利于了钚在自辐射场中抗腐蚀.     

HD+离子在两个超短脉冲激光场作用下的解离几率分布

本文报告HD⁺离子在波长λ=306.7nm的三倍频超短脉冲弱光场E₃和不同延时的二倍频超短脉冲强光场E₂作用下解离几率分布的定量计算结果。E₃和E₂的脉冲宽度均为10fs.E₃先行,E₂延时取0、4、8、12、16、20、24、28fs.计算结果用三维图表出。图中看到,沿相对动量坐标有三个峰,分别位于动量2.1×10^-18、4.4×10^-18、5.8×10^-18g·cm/s处,依次称为1号峰、2号峰、3号峰,位置各是λ=306.7nm的单光子能量、双光子能量、三光子能量消耗于解离能之后离子碎片具有相对动量。据分析,1号峰来源于E₂场引起的受激发射作用。     

LiNbO3晶体中屏蔽光伏孤子自偏转的 时空演化与可控因素

基于带输运模型理论建立了 LiNbO₃ 晶体屏蔽光伏孤子的时空演化动力学方程, 用有限差分方法求解发现, LiNbO₃ 晶体中明、暗屏蔽光伏孤子存在大的自偏转, 并且光孤子形状变得具有不对称性, 偏转方向的曲线斜率绝对值变大, 偏转反方向的曲线斜率绝对值变小. 分析研究表明影响其自偏转度和形变的因素包括受主浓度 N_A, 暗辐射强度 I_d 和外加电场 E₀ . 其他条件不变的情况下N_A 越大, 明孤子的自偏转度与形变越小, 暗孤子的自偏转度与形变反而越大; 对于 I_d , 它对明暗孤子的影响是相同的, I_d 越小, 晶体里诱导出的空间电荷场越容易达到饱和, 当信号光中心光强与暗辐射强度之比为 10^-1时无饱和现象产生; 随着 E₀ 数值的增大, 明孤子的自偏转度和形变减小, 而暗孤子的自偏转度和形变反而增大.     

Effective response in nonlinear spherical coated composites under external AC and DC electric field

By using the perturbation method, effective nonlinear direct current (DC) and alternating current (AC) responses of nonlinear composites with spherical coated inclusions randomly embedded in a host medium are studied under the action of external electric field E_a =E₀ +E₁ sin ωt+E₃ sin 3ωt with different amplitudes and frequencies. The local potentials of composites at all harmonics are given in the inclusion particles and the host regions. All effective nonlinear responses to composites and the relationship between the effective nonlinear responses at all harmonics are also deduced for the spherical coated inclusions in a dilute limit.     

氮原子在800 nm附近的同位素位移

The Doppler-limited absorption spectra of ¹⁴N and ¹⁵N atoms were measured around 800 nm using concentration modulation spectroscopy to study their isotope shifts. The nitrogen atoms were generated by discharging molecular nitrogen buffered with helium in a homemade discharge tube. The isotope shifts of four multiplets (3s⁴P_J→3p⁴D_J^o, 3s⁴P_J→3p⁴P_J^o, 3s²D_J→5s²P_J^o, and 3p²P_J^o→5s²D_J^o) were measured and their J-dependent specific mass shifts were observed and discussed.     

Probing the interactions of charmed mesons with nuclei in ¯p-induced reactions

We study the perspectives of resonant and nonresonant charmed-meson production in ˉp + A reactions within the Multiple Scattering Monte Carlo (MSMC) approach. We calculate the production of the resonances Ψ(3770),Ψ(4040) and Ψ(4160) on various nuclei, their propagation and decay to D,ˉD, D ^*,ˉD^*, D _s,ˉD_s in the medium and vacuum, respectively. The modifications of the open charm vector mesons in the nuclear medium are found to be rather moderate or even small such that dilepton spectroscopy will require an invariant mass resolution of a few MeV. Furthermore, the elastic and inelastic interactions of the open charm mesons in the medium are taken into account, which can be related to (u, d )-, s- or c-quark exchange with nucleons. It is found that by studying the D/ˉD ratio for low momenta in the laboratory ( ? 2 - 2.2 GeV/c) as a function of the target mass A stringent constraints on the c-quark exchange cross-section can be obtained. On the other hand, the ratios D ^- _s/D ⁺ _s as well as D/D ^- _s and D/D ⁺ _s at low momenta as a function of A will permit to fix independently the strength of the s-quark exchange reaction in D ^- _s N scattering. Received: 1 March 2002 / Accepted: 21 March 2002     

Transverse energy measurement in Au+Au collisions by the STAR experiment

Transverse energy (E _T ) has been measured with both of its components, namely hadronic (E _T ^had ) and electromagnetic (E _T ^em ) in a common phase space at mid-rapidity for 62.4 GeV Au+Au collisions by the STAR experiment. E _T production with centrality and ?{s_NN }\sqrt {s_{NN} } is studied with similar measurements from SPS to RHIC and is compared with a final state gluon saturation model (EKRT). The most striking feature is the observation of a nearly constant value of E _T /N _ch ∼ 0.8 GeV from AGS, SPS to RHIC. The initial energy density estimated by the boost-invariant Bjorken hydrodynamic model, is well above the critical density for a deconfined matter of quarks and gluons predicted by lattice QCD calculations.     

On extending the applicability of the initial rise method for thermoluminescence glow peak analysis

We discuss an approach by which the initial rise (IR) method can be extended to evaluate the frequency factor (s) of a glow peak, in addition to activation energy (E) that is usually obtainable from it. The approach requires that the phosphor be irradiated to a dose near or equal to the saturation dose of the glow peak to be analyzed. By plotting ln(I) against 1/T for the thermoluminescence glow peak, one obtains a straight line whose gradient is ? E/k and intercept is ln(n ₀ s/β) from which E and s are determined respectively. The approach was used to analyze two numerically generated glow peaks and reproduce with reasonably good accuracy the E and s values used to generate the glow peaks. It is also shown that when it becomes necessary to include intensities above 10% of the maximum glow peak intensity, the error introduced into the calculated frequency factor by the assumption n? n ₀ is always negligible when the analysis is done by plotting ln(I/n) against 1/T. In this case the frequency factor is obtained from ln(s/β) which is now the intercept. Two experimental glow peaks were also analyzed using the new IR method approach and peak shape method. Good agreement, which again shows the validity and reliability of the present approach was observed between the values of E and s obtained by the two methods.     

多晶薄膜表面粗化与生长方式转变

采用原子力显微镜研究了磁控溅射多晶薄膜表面粗化行为对归一化沉积温度T_s/T_m(T_s是沉积温度，T_m是材料熔点)的依赖性与薄膜生长方式转变行为.随着T_s/T_m增加，薄膜表面粗糙度增加，而表征粗糙度随时间演化特征的生长指数β历经了先减小再增加的过程.β对T_s/T_m的依赖关系反映了薄膜生长方式的转变行为，即薄膜生长依次由随机生长方式向表面扩散驱动生长方式与异常标度行为生长方式转变.在低于体扩散控制薄膜生长的温度时，晶界扩散机理导致多晶薄膜的表面粗化的异常标度行为. 关键词： 多晶薄膜 表面粗化 温度 生长     

Dependences of small-signal gain and saturation-energy intensity on laser active length, buffer-gas type, and pressure variations in gold-vapor lasers

We measure the small-signal gain coefficient g ₀ and the saturation-energy intensity E _s of a gold-vapor laser by implementing an oscillator-amplifier on the base of the gold-vapor laser. The dependence of the gain and saturation properties of the laser on the length of the laser active medium, the buffer-gas pressure, and the buffer-gas type are studied. The measurements of g ₀ for the amplifier for different lengths of the laser active medium (60 and 75 cm) show that with increase in the active-medium length, g ₀ increases while E _s, vice versa, decreases. Such behavior of g ₀ and E _s is observed in the whole range of changing the total pressure of buffer gas in the He-Ne mixture.     

Magnetic properties of 36 wt% Ni-Fe alloys

This paper reports the effect of adding molybdenum, chromium or copper, in amounts up to 5 wt%, on the initial permeability μ_i, magnetostriction constant λ_ps and saturation magnetisation M_s of a 36 wt% Ni-Fe alloy. Mo and Cr decrease λ_ps and M_s, but Cu has the opposite effect. The initial permeability is increased by Mo and Cr additions, except at high concentrations, whereas Cu produces little change in μ_i. Changes in the magnetic moment per atom are discussed in terms of the effect of alloying on the mean number of (d+s) outer electrons. It is suggested that the change in magnetic moment per atom may influence the value of μ_i indirectly due to changes in the spin-orbit and exchange interactions, which alter the anisotropy and magnetostriction constants and domain wall energy.     

The strange quark mass from QCD sum rules

The strange quark mass is calculated from QCD sum rules for the divergence of the vector as well as axialvector current in the next-next-to-leading logarithmic approximation. The determination for the divergence of the axial-vector current is found to be unreliable due to large uncertainties in the hadronic parametrisation of the two-point function.From the sum rule for the divergence of the vector current, we obtain a value of (1 GeV)=189±32 MeV, where the error is dominated by the unknown perturbativeO( _s ³ ) correction. Assuming a continued geometric growth of the perturbation series, we findm _s =178±18 MeV. Using both determinations ofm _s , together with quark-mass ratios from chiral perturbation theory, we also give estimates of the light quark massesm _u andm _d.     

Electronic band structures of the alkali metals and of the noble metals and their α-phase alloys

In the monovalent metals the electronic band structure is strongly affected by the size of the band gap E _s-E _p at the Brillouin zone faces, a large gap implying a large distortion of the Fermi surface. Here E _s and E _p are the energies of the purely s-like and p-like states on the zone faces. We have made crude estimates of E _s-E _p for the alkali and noble metals, in terms of the s-p excitation energy Δ_sp of the free atoms. These suggest a single model which correlates most of the experimental information about the band structures of these metals. In particular the Fermi surface of lithium appears to make considerable contact with the zone faces. In the α-phase alloys of the noble metals, the solute always has a larger value of Δ_sp than the solvent, which raises the energy E _p relative to E _s. The Fermi surface becomes more nearly spherical in copper alloys than in copper, since E _p<E _s, whereas it distorts further in the gold alloys (E _p>E _s). This accounts for many Knight shift, electronic specific heat, magnetic susceptibility and other data on these alloys. Furthermore it provides the extension of Jones' explanation of the Hume-Rothery rule demanded by the non-spherical Fermi surface in pure copper and gold.     

A refined estimate of the bond length of methane

The atmospheric reaction of H₂S with Cl was investigated using high level ab initio calculations and Canonical Variational Transition State Theory (CVTST). The adduct formation step is the dynamical bottleneck, and the rate constant was calculated to be 1.2 × 10^?9 cm³ molecule^?1 s^?1, which is around ten times greater than the upper experimental value. Additional ab initio classical trajectory calculations show that the adduct formed in the initial collision can easily dissociate, recrossing the variational transition state. The stabilization of this species depends on the vibrational excitation of H₂S molecule, which requires an almost collinear SH-Cl collision. These dynamical effects provide an explanation for the substantial error in the rate constant obtained using CVTST.     

Supersonic Jets with Periodic Structure due to Arc plasma Expansion into a Low pressure ambient

A supersonic plasma jet was produced by a d.c. arc plasma generator operated at normal pressure and connected to a low-pressure (p∞ = 0.2-50 kPa) chamber via cylindrical nozzle with diameter of 2.5 mm. The argon gas flow rate was G = 0.025 to 0.35 g.s^?1. In some experiments current I_E ≦ 30 A passed coaxially through the initial part of the jet. Photographic records of the jet and pressure measurements are in agreement with theoretical predictions by a simple one-dimensional, gasdynamical model capable of self-consistent calculations throughout the plasma source/jet system. Periodic jet structure is observed over a wide range of experimental conditions, incl. in highly under-expanded flow. The jet expansion angle and Mach disc position vary with p∞, G and I_E, but are nearly constant at different arcing currents.