相互作用分析中地震动输入长周期校正研究

由于长周期误差信号的影响．仅通过直流分量滤波处理地震加速度记录,积分获得的地震波位移时程曲线仍存在很严重的零线飘移现象．通过利用最小二乘法构建原始加速度均值的二次多项式拟合曲线,提出了简便的加速度记录长周期误差时域校正法,在几乎不改变地震加速度频谱主要性质的前提下,很好地改善了加速度记录的积分性质．具体的工程应用验证了其有效性．并显示在大跨度结构-地基动力相互作用时域分析中,地震波输入位移时程对结构总体响应有着明显影响。     

大跨度斜拉桥在多点随机地震激励作用下的响应分析

以南京第二长江大桥南汊桥为例,采用随机振动分析方法研究了地震动的空间变化特性,包括行波效应、部分相干效应以及局部场地效应对大跨度斜拉桥地震响应的影响.研究结果表明:与一致地震激励相比,地震动的空间变化特性可以使斜拉桥的地震响应改变30%;同时考虑行波效应、部分相干效应时的地震响应接近于仅考虑行波效应时的地震响应;行波效应对斜拉桥地震响应的影响明显大于部分相干效应的影响;在加速度均方根相同的随机地震作用下,斜拉桥在软场地条件下的地震响应明显大于在硬场地条件下的响应,中等场地条件下的地震响应介于两者之间.     

C+L波段分布式拉曼光纤放大器实验研究

采用后向泵浦方式,利用波长分别为1426,l440,1454,1472及1496nm的大功率半导体激光器作为泵源,实现了带宽为80．nm(1530～1610nm)的宽带分布式拉曼光纤放大器,对40个信道的信号进行分布式放大,获得了较好的增益平坦及噪声特性．信号经60km的普通单模光纤传输后,平均开关增益为16．5dB,增益平坦度为1.5dB,最大等效噪声指数为-0．9dB．采用Agilent 8166B测试系统的多通道信号源(C L-band)作为信号源测量增益谱．由于考虑了信号光之间的相互拉曼作用及信号光对泵浦的影响,实验结果更精确．     

基于传递矩阵方法的排气消声器计算机辅助设计

为实现消声器选型的程序化,在建立消声系统数学模型的基础上,通过对常见的抗性排气消声器的结构分析,将基本的消声元件分为7种,在考虑均匀流和线性温度梯度的前提下,推导了它们的传递矩阵;并且在此基础上建立消声系统消声量预测模型和压力损失计算模型.利用VC 和Matlab联合编制了排气消声器性能预测的通用CAD软件,将软件的计算结果与试验结果进行比较,吻合较好.该软件具有一定的实用价值及应用前景.     

多层工业厂房火灾后加固设计与抗震性能分析

为探究多层钢筋混凝土工业厂房在加固前后以及火灾前后的抗震性能变化情况,围绕火灾后多层工业厂房的加固设计和抗震性能分析展开研究.首先,建立多层钢筋混凝土框架结构火灾后抗震性能的有限元分析方法;随后,介绍某多层工业厂房结构特点和火灾后结构损伤情况,据此综合利用增大截面、粘贴钢板和粘贴纤维复合材料等方法对结构进行了加固设计;进而,建立了结构三维有限元模型,开展了结构动力特性和弹性地震响应分析;最后,进行了结构的动力弹塑性分析,探究了火灾前、后结构整体的抗震性能和加固措施对火灾后结构抗震性能的改善,并探讨了不同地震动特性对多层钢筋混凝土框架抗震性能的影响.研究结果表明:火灾后,整体结构地震响应明显大于火灾前,层间位移角最大增加了143％,且不同楼层层间位移角的增大幅度与楼层受火情况密切相关;近场地震作用下的结构响应大于远场地震,罕遇地震作用下底层层间位移角在近场地震作用下可达远场地震作用的2.71倍;所提出的加固方案有效提升了结构抗震性能,罕遇地震作用下底层层间位移角平均减少115.7％.研究成果可为多层工业厂房火灾安全性评估和加固设计提供参考.     

Noise spectra of microcavity-emitting field in the linear regime

Our work presents a theoretical study on the behavior of the fluctuation statistics of the field emitted from a semiconductor microcavity containing a quantum well. We derive an analytical expression of the noise spectra in non-resonant pumping. Dynamical behaviors of the noise spectra depending on the thermal bath and the detuning between exciton and cavity frequencies are discussed.     

Domain wall delocalization,dynamics and fluctuations in an exclusion process with two internal states

We investigate the delocalization transition appearing in an exclusion process with two internal states, respectively on two parallel lanes. At the transition, delocalized domain walls form in the density profiles of both internal states, in agreement with a mean-field approach. Remarkably, the topology of the system’s phase diagram allows for the delocalization of a (localized) domain wall when approaching the transition. We quantify the domain wall’s delocalization close to the transition by analytic results obtained within the framework of the domain wall picture. Power law dependences of the domain wall width on the distance to the delocalization transition as well as on the system size are uncovered, they agree with numerical results.     

Active Brownian motion models and applications to ratchets

We give an overview over recent studies on the model of Active Brownian Motion (ABM) coupled to reservoirs providing free energy which may be converted into kinetic energy of motion. First, we present an introduction to a general concept of active Brownian particles which are capable to take up energy from the source and transform part of it in order to perform various activities. In the second part of our presentation we consider applications of ABM to ratchet systems with different forms of differentiable potentials. Both analytical and numerical evaluations are discussed for three cases of sinusoidal, staircaselike and Mateos ratchet potentials, also with the additional loads modelled by tilted potential structure. In addition, stochastic character of the kinetics is investigated by considering perturbation by Gaussian white noise which is shown to be responsible for driving the directionality of the asymptotic flux in the ratchet. This stochastically driven directionality effect is visualized as a strong nonmonotonic dependence of the statistics of the right versus left trajectories of motion leading to a net current of particles. Possible applications of the ratchet systems to molecular motors are also briefly discussed.     

Accounting for the energies and entropies of kinesin’s catalytic cycle

When the processive motor protein kinesin walks along the biopolymer microtubule it can occasionally make a backward step. Recent single molecule experiments on moving kinesin have revealed that the forward-to-backward step ratio decreases exponentially with the load force. Carter and Cross (Nature 435, 308-312, 2005) found that this ratio tightly followed 802 × exp[−0.95F], where F is the load force in piconewtons. A straightforward analysis of a Brownian step leads to L/(2k _B T) as the factor in front of the load force, where L is the 8 nm stepsize, k _B is the Boltzmann constant, and T is the temperature. The factor L/(2k _B T) does indeed equal 0.95 pN⁻¹. The same analysis shows how the 802 prefactor derives from the power stroke energy G as exp[G/(2k _B T)]. There are indications that the power stroke derives from the entropically driven coiling of the 30 amino acid neck linker that connects the two kinesin heads. This idea is examined and consequences are deduced.     

Lattice Lotka-Volterra model with long range mixing

A spatio-temporal process in the Lattice Lotka Volterra (LLV) model, when realized on low dimensional support, is studied. It is shown that the introduction of a long-range mixing causes a drastic change in the system’s behavior, which transits from small random-like fluctuations to global oscillations when the mixing rate transcends above a critical point. The amplitude of the induced oscillations is well defined by the mixing rate and is insensitive to the initial conditions and the lattice size variations. The observed behavior essentially differs from that predicted by the Mean-Field model which is conservative. The oscillations are of limit-cycle type and appear as a stochastic analog of a Hopf bifurcation.