Stochastic processes for line shapes and intensities

Stochastic processes provide flexible and fast calculations for modeling dynamical interactions between an atom and charged particles. We use a stochastic renewal process for the plasma microfield being the cause of Stark broadening. The accuracy and improvement possibilities of Lyman profiles calculations with a renewal process are analyzed by comparing to ab initio simulations for ion broadening only. Stochastic processes may also be applied to out of equilibrium plasmas. We present our first results for the effect of Langmuir waves on a line broadened by electrons only, and for the changes of atomic populations submitted to strong temperature fluctuations.     

平面工艺功率型半导体晶体管铝迁移实验

在平面工艺制造的功率型半导体晶体管芯片表面,发现高强度电场条件下存在跨越0.05 mm半绝缘硅带的铝迁移现象,对此开展了一系列实验研究. 通过开帽物理观察与分析、低温测试与烘干验证、迁移物质分析对比,确定了迁移现象中的迁移物质,同时分别取证环境温度、湿度、芯片表面污染物以及器件局部结构与此迁移现象的关联性,以此提出了迁移机理假说,为在不同的诱发条件下对此迁移机理进行确认和量化研究提供指导.      

基于神经网络的维修保障流程影响度分析方法

针对维修保障能力的主要影响因素分析及其影响度的量化问题,以维修保障业务流程作为分析对象,通过建立业务流程参数与装备使用能力的关系函数,建立影响因素和影响度的分析方法,并给出影响程度、关键要素、关键影响要素以及综合影响度定义.以流程数量、效率和稳定性作为分析指标,利用MATLAB软件基于神经网络实现对关系函数的逼近求解,为装备制造企业提供确定关键流程和主要影响因素的系统实现.以某航空装备制造企业维修保障业务流程为例给出具体应用,与专家评价方法相比该方法可获得更为客观的分析结果.      

焊接工艺方法对6061-T6铝合金焊接接头疲劳性能的影响

通过对6061铝合金MIG焊接头和TIG焊接头在对应的加载应力条件下疲劳寿命的测定,对比分析了两种工艺方法对铝合金焊接接头疲劳性能的影响。结果表明,在加载应力低于焊接接头静载力学性能的90%时,焊接接头的疲劳寿命均能满足需求背景的需要(100000次不破坏)。同样载荷条件下,MIG焊接头的疲劳性能优于TIG焊接头,尤其是在高应力条件下,MIG焊接头的优势更为明显。焊接接头显微组织分析表明,MIG焊接头比TIG焊接头具有更为细小的晶粒和焊接热影响区,有效地提高了接头的滑移形变抗力,抑制了循环滑移带的形成和开裂,从而提高接头的疲劳性能。疲劳断口分析显示,试件的表面缺陷(疏松、气孔、夹杂等)及机械损伤是疲劳裂纹主要的策源地。     

基于Markov Chain Monte Carlo的幂律过程的Bayesian分析

在多种合理的无信息先验分布下,基于Markov Chain Monte Carlo方法,提出了一种简单且易于抽样的幂律过程的Bayesian分析方法.所提方法将失效、时间截尾数据统一分析,能快捷地获取幂律过程模型参数的Markov Chain Monte Carlo样本,利用该样本不但能直接给出模型参数函数的后验分布,还能给出单样预测和双样预测的分析.一个经典工程数值算例说明了所提方法的可行性、合理性与有效性.该方法具有一定的优越性,可为小子样可靠性增长分析提供一种值得参考的方法.      

Modelling ignition probability with pairwise mixing-reaction model for flame particle tracking

Reduced order models for ignition analysis can offer insights into ignition processes and facilitate the combustor optimization. In this study, a Pairwise Mixing-Reaction (PMR) model is formulated to model the interaction between the flame particle and the surrounding cell mixture during Lagrangian flame particle tracking. Specifically, the model accounts for the two-way coupling of mass and energy between the flame particle and the surrounding shell layer by modelling the corresponding turbulent mixing, chemical reaction and evaporation process if present. The state of a flame particle, e.g., burnt, hot gas or extinguished, is determined based on particle temperature. This model can properly describe the ignition process with a spark kernel being initiated in a nonflammable region, which is of practical importance in certain turbine engines and has not been rigorously accounted for by the existing models based on the estimation of local Karlovitz number. The model is integrated into an ignition probability analysis platform and is demonstrated for a methane/air bluff-body flame with the flow and fuel/air mixing characteristics being extracted from a non-reacting simulation. The results show that for the spark location being at the extreme fuel-lean outer shear layer of the recirculation zone, PMR can yield ignition events with a significant number of active flame particles. The mechanisms for the survival of the initial flame particles and the entrainment of the survived flame particles into the recirculation zone are analyzed. The results also show that the ignition probability map from PMR agrees well with the experimental observation: a high ignition probability in the shear layer of the recirculation zone near the mean stoichiometric surface, and low ignition probabilities inside the recirculation zone and the top stagnation region of the recirculation zone. The parametric study shows that the predicted shape of the ignition progress factor and ignition probability is in general insensitive to the model parameters and the model is adequate for quantifying the regions with high ignition probabilities.     

流量调节器管路系统自激振荡特性研究（液体火箭推进专刊）

流量调节器管路系统在小流量大压降工况下会出现低频自激振荡现象,为了深入认识自激振荡产生机理,结合某型自稳流型流量调节器及管路系统,基于流量调节器弹簧振子动力学模型开展数值仿真研究。数值仿真得出自激振荡频率为94 Hz,与发动机试验结果一致。自激振荡产生机理是流量调节器的液动力受滑阀窗口型面变化率影响对系统形成正反馈作用,流量调节器综合刚度小于零时系统失稳。分析了流量调节器结构参数对系统稳定性的影响作用,三角形滑阀节流口流通面积能够抑制管路系统自激振荡。结合某型流量调节器负载特性试验系统进行验证,得出随着流量调节器压降升高,管路系统稳定性变差,获得的幅频特性,稳定边界与试验结果一致。     

类芬顿反应在碳化硅光学材料研抛中的作用

本文讨论了以零价铁代替亚铁盐形成类芬顿反应(Fenton-Kind process)体系应用于碳化硅光学材料抛光过程中的反应机理,通过双转子研抛机床对SiC光学材料进行了研抛实验,验证了类芬顿反应能够有效地提高SiC材料的加工效率和加工质量。通过对比不同磨料粒度研抛液加工结果,获得了SiC材料研抛过程中化学与机械作用规律。     

基于浴盆曲线故障率函数的FFOP预计方法

与传统可靠性指标中假设产品的随机失效不可避免不同,无故障工作期(FFOP)内产品不会发生任何故障(即零故障)。首先阐述了FFOP的概念内涵、与平均故障间隔时间(MTBF)的区别和联系,提出了一种FFOP的预计方法。该方法假设产品的故障率函数具有浴盆曲线特征、故障发生过程为泊松过程、产品具有固定的免维修工作期。然后以一种改进的Weibull分布函数描述具有浴盆曲线函数特征产品的故障率。基于泊松过程理论,给出了FFOP的预计算法、流程和仿真验证手段。最后以某型无人机舵机为案例对研究方法的可用性进行了验证。结果表明:FFOP与免维修工作期(MFOP)、置信度水平密切相关,及时维修的产品能够保证较长的FFOP。在工程应用时,FFOP的确定应综合考虑运行维护费用进行权衡。