Laboratory modelling of shear behaviour of soft joints under constant normal stiffness conditions

Shear behaviour of regular sawtooth rock joints produced from casting plaster are investigated under constant normal stiffness (CNS) conditions. Test results obtained in this investigation are also compared with the constant normal load (CNL) tests. It is observed that the peak shear stress obtained under CNL conditions always underestimates the peak shear stress corresponding to the CNS condition. Plots of shear stress against normal stress show that a nonlinear (curved) strength envelope is acceptable for soft rock joints subjected to a CNS condition, in comparison with the linear or bilinear envelopes often proposed for a CNL condition. Models proposed by Patton (1966) and Barton (1973) have also been considered for the predictions of peak shear stress of soft joints under CNS conditions. Although Patton's model is appropriate for low asperity angles, it overestimates the shear strength in the low to medium normal stress range at higher asperity angles. In contrast, while Barton's model is realistic for the CNL condition, it seems to be inappropriate for modelling the shear behaviour of soft joints under CNS conditions. The effect of infill material on the shear behaviour of the model joints is also investigated, and it is found that a small thickness of bentonite infill reduces the peak stress significantly. The peak shear stress almost approached that of the shear strength of infill when the infill thickness to asperity height ratio (t/a) reached 1.40. This paper also introduces an original, empirical shear strength envelope to account for the change in normal stress and surface degradation during CNS shearing. © Rapid Science Ltd. 1998     

结构损伤初步定位的简便方法

提出将基于动态特性的检测方法与局部物理检测手段相结合的探伤思路。利用动力学方法进行实时监测及损伤区域的粗略定位,再由物理探测方法实现损伤程度和位置的精确判断,从而降低对动态检测方法的精度要求。此外,文中尝试直接由结构的动力响应信号构建能量指标识别结构损伤的方法,不需要进行模态参数识别,算法简单,有望应用于海洋平台、高层建筑等可简化为串联多自由度体系的结构的实时监测和损伤初步定位。     

Hydrodynamic simulations of merging clusters of galaxies

We present the results of high-resolution AP³M+SPH simulations of merging clusters of galaxies. We find that the compression and shocking of the core gas during a merger can lead to large increases in bolometric X-ray luminosities and emission-weighted temperatures of clusters. Cooling flows are completely disrupted during equal-mass mergers, with the mass deposition rate dropping to zero as the cores of the clusters collide. The large increase in the cooling time of the core gas strongly suggests that cooling flows will not recover from such a merger within a Hubble time. Mergers with subclumps having one eighth of the mass of the main cluster are also found to disrupt a cooling flow if the merger is head-on. However, in this case the entropy injected into the core gas is rapidly radiated away and the cooling flow restarts within a few Gyr of the merger. Mergers in which the subcluster has an impact parameter of 500 kpc do not disrupt the cooling flow, although the mass deposition rate is reduced by ∼30 per cent. Finally, we find that equal mass, off-centre mergers can effectively mix gas in the cores of clusters, while head on mergers lead to very little mixing. Gas stripped from the outer layers of subclumps results in parts of the outer layers of the main cluster being well mixed, although they have little effect on the gas in the core of the cluster. None of the mergers examined here resulted in the intracluster medium being well mixed globally.     

数字高程模型的质量检查

提出基于方差-尺度规律的新型插值方法评价体系,根据“在给定研究区域内,变量的统计方差随尺度增大而减小”的理论规律,提出三项预设的评价准则。以DEM插值应用为例,选取高精度曲面模型、样条插值、克里金插值、反距离权重插值4种方法进行对比评价。数值实验表明,方差-尺度递变趋势的评价准则值得存疑。两组DEM插值案例表明,地形的复杂特性、插值算子的光滑效应以及尺度的综合效应致使理论的渐变规律无法达到。引入采样方差值后,不同采样密度下仍可以将方差变化趋势及方差的整体大小作为精度评价准则,在对第二、第三条预设评价准则修缮后确立三项评价准则。综合所有案例可知,此评价体系原理简单、操作方便,为DEM插值方法评价提供了一套新的标准和框架,并且可针对不同采样密度条件提出完善的插值方法适用性建议。     

基于深度学习的太阳F_10.7辐射通量的短期预报研究

F_10.7太阳辐射通量作为输入参数被广泛运用于大气经验模型、电离层模型等空间环境模型,其预报精度直接影响航天器轨道预报精度.采用时间序列法统计了太阳辐射通量F10.7指数和太阳黑子数(SSN)的关系,给出了两者之间的线性关系,在此基础上提出了一种基于长短时记忆神经网络(Long and Short Term Memory,LSTM)的预报方法,方法结合了54 d太阳辐射通量指数和SSN历史数据来对F_10.7进行未来7 d短期预报,并与其他预报方法的预报结果进行了比较,结果表明:(1)所建短期预报7 d方法模型的性能优于美国空间天气预报中心(Space Weather Prediction Center, SWPC)的方法,预测值和观测值的相关系数(CC)达到0.96,同时其均方根误差约为11.62个太阳辐射通量单位(sfu),预报结果的均方根误差(RMSE)低于SWPC,下降约11%;(2)对预测的23、24周太阳活动年结果统计表明,太阳活动高年的第7 d F10.7指数预报平均绝对百分比误差(MAPE)最优可达12.9%以内,低年最优可达2...     

时间域航空电磁法一维正演研究

基于电磁勘探理论，导出了层状大地条件下时间域航空电磁法（偶极-偶极装置）的正演计算公式和算法，编制了相应的计算机程序，对若干典型地电断面作了正演计算. 计算结果说明时间域航空电磁法的探测能力和探测条件，进而为时间域直升机航空电磁系统的设计方案提供了依据.     

基于NPSVM模型的超新星识别方法

巡天观测与高能物理、黑洞天文等领域均有密切的联系.基于星系-超新星二分类问题,研究光谱数据预处理,结合余弦相似度改善PCA(Principal Component Analysis)光谱分解特征提取方法,用SDSS(the Sloan Digital Sky Survey)、WISeREP(the Weizmann Interactive Supernova data REPository)组成的5620条光谱数据集训练支持向量机,可以得到0.498%泛化误差的识别模型和新样本分类概率.使用Neyman-Pearson决策方法建立NPSVM(Neyman-Pearson Support Vector Machine)模型可进一步降低超新星的漏判率.     

基于斩波轮技术的K波段接收机噪声校准研究

接收机是射电天文中用于探测微弱射电信号的重要接收设备.接收机的强度校准就是将接收机对射电源的响应转换为天文意义上的流量密度.常规方法就是使用经典的冷热负载法,将接收机自身的强度响应转换为一个等效的温度值,之后再据此对射电源做进一步标定.通过搭建基于斩波轮技术的K波段接收机强度校准平台,使用斩波轮法测试K波段常温接收机的噪声温度,并与传统冷热负载法的测试结果进行比对.结果显示,在晴好天气条件下,斩波轮法在30°、90°仰角下噪声温度的最大测试误差为7.5%和8.4%,可以很好地应用于实际噪声温度测试中;但在5°仰角测试中,由于过低仰角引入了地面噪声,使得斩波轮法的测试误差上升至20%–30%之间而无法使用.希望在此基础上进一步开展K波段天空亮温度的理论计算与实测,从而完善斩波轮技术的应用,使之可以满足在不同气象条件下的噪声校准测试需求.     

Bootstrapped discrete scale invariance analysis of geomagnetic dipole intensity

The technique of bootstrapped discrete scale invariance allows multiple time-series of different observables to be normalized in terms of observed and predicted characteristic timescales. A case study is presented using the SINT2000 time-series of virtual axial dipole moment, which spans the past 2 Myr. It is shown that this sequence not only bears a clear signature of a preferred timescale of about 55.6 Ka, but additionally predicts similar features (of shorter and longer duration) that are actually observed on the timescales of historical secular variation and dipole reversals, respectively. In turn, the latter two empirical sources both predict the characteristic timescale found in the dipole intensity sequence. These communal scaling characteristics suggest that a single underlying process could be driving dynamo fluctuations across all three observed timescales, from years to millions of years.