环境风作用下弧形立面机场航站楼自然排烟效果的数值模拟研究

机场航站楼公共区域多采用自然排烟方式,而环境风和排烟窗控制模式会影响排烟效果,尤其在弧形立面航站楼中影响效果显著。以某弧形立面航站楼为研究对象,选取候机指廊弧形区域的自然排烟窗分3段控制启闭,进而确定6种控制模式,采用 FDS火灾动力学模拟软件研究了不同模式下的排烟效果,获得了环境风和发烟量对危险到达时间的影响规律, 提出了排烟窗控制模式流程。结果表明:对于弧形立面航站楼而言,排烟窗分段控制模式下的排烟效果优于不分段控制模式,可以有效降低环境风对自然排烟的影响。当环境风速超过10 m/s时,须关闭迎风向排烟窗,避免烟气倒灌。在法向环境风作用的不利条件下,当发烟量在0. 05~0. 20之间时,危险到达时间的最大值tmax与发烟量x满足一次函数关系式:tmax=-3018x+1030.5。     

侧向风速作用下聚氨酯泡沫双火源火蔓延研究

为探究环境风作用下逆向双点火源聚氨酯泡沫火蔓延及融合行为，开展多组对照实验并从材料传热机理角度分析侧向风速对火蔓延行为中火羽流形态、质量损失和辐射热流场等特征参数的影响。结果表明:风速与上述参数之间存在非线性关系。环境风效应使火焰被拉长且敷贴于预热区表面，增大预热区面积和热反馈；侧向风速的增加对FPU板材质量损失的影响逐渐弱化，且板材的熔滴率与风速呈正相关；无论侧向风是否存在，两侧逆向火焰融合后均达到整个蔓延过程中的峰值温度；风速的存在限制了火焰温度与辐射热通量峰值，也缩短了温度和辐射峰值出现的时间。     

隧道竖井外流场模拟及风塔外型结构优化研究

为提高长大隧道施工及运营通风过程中竖井通风效率，实现隧道节能通风。依托实际隧道工程，基于Fluent软件，选用RNG k-ε湍流模型，对无竖井风塔、圆柱风塔、矩形风塔和凸台状风塔4种结构、不同环境风速工况下的流场进行数值模拟，并对模拟结果进行交叉比对分析，提出最优化的竖井风塔结构形式。研究结果表明:采用风塔结构竖井出流量要远大于无风塔结构竖井出流量；对于不同竖井风塔外型结构，竖井出流量与环境风速关系呈多项式函数；当倾斜角为45°、圆心角为90°时，凸台结构对应竖井出流量最大，且受环境风影响波动较小，通风稳定性较好。建议竖井出口结构选用矩形或凸台状，实现对环境风的高效利用，提高竖井通风效率。     

街道峡谷内汽车排放污染物浓度分布的观测与数值模拟

在上海某典型街道峡谷内按一定的空间布点,在一定时段内同时对各布点进行采样并做一氧化碳浓度分析,同时记录车辆种类、车流量、气象条件等,分析街道峡谷内污染物浓度的分布.运用风向频率加权(WDFW)方法,结合大气流动和污染物扩散的CFD模型进行数值模拟计算.结果表明,数值模拟结果和现场观测结果较吻合,建筑物低的一侧污染物浓度远高于建筑物高的一侧污染物浓度,两侧的污染物浓度随着高度的增加而降低.     

风蚀条件下人类活动对高寒草地水分保持功能影响的定量模拟

青藏高原是我国重要的生态屏障,具有重要的水源涵养功能。高寒草地生态系统退化将严重地影响高原的水源涵养功能。为了定量研究人类活动对青藏高原高寒草地水分保持功能的影响,采用大型风洞实验,模拟人类不同的干扰方式和干扰程度对高寒草地土壤水分保持功能的影响。实验样品采自青藏高原三个草地类型:高寒草甸、草原化草甸和高寒草原。干扰方式包括破坏草地地上部分和根系,干扰程度包括轻干扰、中度干扰、重度干扰和全部破坏。实验结果表明,地上部分破坏后,土壤含水率均有下降,降幅分别为:QZ1为6.9%～9.2%、QZ2为6.8%～10.1%、QZ3为9%～10%;当根系遭到破坏后土壤含水率的降幅则分别为:QZ1为16%～30%,QZ2为17.25～32.1%,QZ3为22%～50%,显然,根系破坏后土壤含水率降幅远大于地表植被破坏后的情形。因此,植被根系是高寒草地水分保持功能的关键。随着干扰程度的加剧,土壤含水率在迅速下降。从试验模拟结果看,三个草地类型中土壤水分保持功能分别是高寒草甸>草原化草甸>高寒草原。     

北京市与香河县大气臭氧及氮氧化合物的变化特征

于2005年春、夏之交,在北京市及河北省香河县两地同时开展了臭氧和氮氧化合物的连续监测.结果表明,香河县和北京市的臭氧统计日变化形式基本一致,北京市大气臭氧平均浓度比香河低,分别为28×10-9V/V和43×10-9V/V;大气中0x浓度北京市高于香河县,分别为72×10-9V/V和60×10-9V/V.观测期间,两地臭氧的月最高值出现时间一致,其中6月21日北京市的臭氧浓度最高达到200×10-9V/V,香河县也达到149×10-9V/V.在三次臭氧严重超标事件发生的前一天,两地都伴随有氮氧化合物的高浓度积累;南风时北京市和香河县的Ox浓度都比平均值高25%.     

A comparative study of prognostic MM5 meteorological modeling with aircraft, wind profiler, lidar, tethered balloon and RASS data over Philadelphia during a 1999 summer episode

This study presents a comparative evaluation of the prognostic meteorological Fifth Generation NCAR Pennsylvania State University Mesoscale Model (MM5) using data from the Northeast Oxidant and Particle Study (NE-OPS) research program collected over Philadelphia, PA during a summer episode in 1999. A set of model simulations utilizing a nested grid of 36 km, 12 km and 4 km horizontal resolutions with 21 layers in the vertical direction was performed for a period of 101 h from July 15, 1999; 12 UTC to July 19, 1999; 17 UTC. The model predictions obtained with 4 km horizontal grid resolution were compared with the NE-OPS observations. Comparisons of model temperature with aircraft data revealed that the model exhibited slight underestimation as noted by previous investigators. Comparisons of model temperature with aircraft and tethered balloon data indicate that the mean absolute error varied up to 1.5 °C. The comparisons of model relative humidity with aircraft and tethered balloon indicate that the mean relative error varied from –11% to –22% for the tethered balloon and from –5% to –30% for the aircraft data. The mean relative error for water vapor mixing ratio with respect to the lidar data exhibited a negative bias consistent with the humidity bias corresponding to aircraft and tethered balloon data. The tendency of MM5 to produce estimates of very low wind speeds, especially in the early-mid afternoon hours, as noted by earlier investigators, is seen in this study also. It is indeed true that the initial fields as well as the fields utilized in the data assimilation also contribute to some of the differences between the model and observations. Studies such as these which compare the grid averaged mean state variables with observations have inherent difficulties. Despite the above limitations, the results of the present study broadly conform to the general traits of MM5 as noted by earlier investigators.     

WIND GENERATED WAVES IN LAKE OKEECHOBEE1

ABSTRACT: The study of wind generated waves is important because waves affect sediment resuspension in lakes. Measurements of wind velocity and wave elevation were made at three different stations in Lake Okeechobee. Significant wave heights were computed using a direct count from the recorded data, and verified by the root-mean-square value approach. The correlation between wind stress and significant wave height also was analyzed. The data revealed a strong correlation. In addition to field measurements, a Boussinesq-type wind-wave model was developed to simulate wind-generated, long-propagating waves. This model included the effects of wind stress and bottom viscous dissipation. Wave elevation and velocity field were evaluated numerically. A six-day simulation using 1996 wind data was conducted. Simulated significant wave heights were found to agree reasonably well with measured values. A predictive wind-wave model provides information about wind generated waves, which is used to compute bottom shear stresses required for sediment resuspension studies.     

非饱和渗流作用下土坡的数值计算分析

讨论了考虑非饱和流渗作用下的土坡的数值计算分析,采用了不同含水量对抗剪强度影响的模型,求解带有渗透力作用项的土体弹塑性平衡方程得到土坡的应力应变分布,并应用于降雨滑坡的简单算例分析。由算例分析得知,利用非饱和渗流耦合应力的弹塑性来分析非饱和渗流而引起的土体滑坡的变形和发展是可行的,并值得完善和发展。     

基于机器学习的电热防除冰表面温度变化趋势预测

针对飞机部件周期控制律电加热防除冰的应用,提出基于机器学习以预测电加热防除冰表面温度的变化趋势.依靠大型结冰风洞及其电加热防除冰控制系统完成防除冰试验,获得有效的试验数据,以通、断电周期为分割单元,将试验数据划分成验证集和训练集.根据电热防除冰过程的换热情况,构建样本的特征参数,利用机器学习的有监督学习方式,选择KNN近邻回归算法和局部线性加权回归算法预测温度变化率,再转换为温度,得到的温度变化与测量数据的线性相关性达到80％以上的高相似度结果,表明使用机器学习可快速预测电热防除冰部件的表面温度变化趋势,且不同的回归算法针对模型不同测温点位置的预测效果存在差异.