Measurement of residual stresses in dissimilar friction stir-welded aluminium and copper plates using the contour method

The longitudinal residual stresses in the friction stir-welded plates of 5A06 aluminium and pure copper were determined using the contour method. The results revealed the presence of high tensile and compressive residual stresses on the aluminium and copper sides, respectively. The residual stresses were detected on the weld zone as well as the thermo-mechanically affected zone (TMAZ) of the aluminium plate. In contrast, the compressive residual stresses in the copper plate had a much narrower width along the weld line. Peak tensile stresses up to 240?MPa were found in the TMAZ of the aluminium plate.     

Kevlar织物软壁包容环抗冲击数值仿真分析研究

针对机匣包容性问题,提出Kevlar平纹织物本构模型的拟合方法,分别采用单层壳、多层壳、层合壳三种有限元建模方式对Kevlar织物软壁包容环弹道冲击过程进行模拟,发展了机匣包容性仿真分析模型。结果表明,多层壳模型和层合壳模型的计算结果更准确,并且可以通过调节摩擦因子使仿真结果与试验结果更接近。在此基础上研究了撞击点位置和弹体入射姿态对弹体剩余速度的影响。结果表明:在较小偏移距离内,撞击点的位置对剩余速度的影响可以忽略不计;滚转角和俯仰角对剩余速度影响较小,而对于偏航角,当角度大于30°时,剩余速度会随着角度的增大明显呈下降趋势,且弹体出射姿态也会发生明显变化。     

宁夏红寺堡盆地隐伏古隆起成因机制及其对区域沙漠化的影响

目前对于青藏高原东北缘活动构造研究多集中于活动断裂带的活动性,而对夹持于其间的沉积盆地构造格架及活动性研究甚少。以夹持于烟筒山断裂与牛首山—罗山断裂之间的宁夏红寺堡盆地为研究对象,采用重力资料重处理解译、音频大地电磁测深(AMT)和地震勘探相结合,揭示红寺堡盆地隐伏构造特征,并进一步采用盆山一体化思路分析隐伏构造的成因机制及其对区域沙漠化的控制作用。结果表明:青藏高原在中新世末发生强烈的NE向推挤、扩展和隆升,红寺堡盆地由坳陷盆地转变为挠曲盆地; 在青藏高原NE向扩展的影响下,烟筒山断裂发生强烈的逆冲作用,古生代—中生代基底逆冲于古近系—新近系之上,受构造变形影响的最新地层为中新统彰恩堡组; 红寺堡盆地内的隐伏古隆起呈NW—SE向,与烟筒山构造带具有相同的构造动力学背景,受青藏高原NE向扩展影响形成于中新世末,并且至今仍具有活动性; 隐伏古隆起周缘断裂的活动破坏了地表稳定性、蓄水能力和地表植被,导致区域沙漠化呈现有规律的NW—SE向带状展布。该研究成果对于宁夏红寺堡盆地区域稳定性评价及沙漠化的综合治理具有现实指导意义。     

Emergency decision support modeling for COVID-19 based on spherical fuzzy information

Significant emergency measures should be taken until an emergency event occurs. It is understood that the emergency is characterized by limited time and information, harmfulness and uncertainty, and decision-makers are always critically bound by uncertainty and risk. This paper introduces many novel approaches to addressing the emergency situation of COVID-19 under spherical fuzzy environment. Fundamentally, the paper includes six main sections to achieve appropriate and accurate measures to address the situation of emergency decision-making. As the spherical fuzzy set (FS) is a generalized framework of fuzzy structure to handle more uncertainty and ambiguity in decision-making problems (DMPs). First, we discuss basic algebraic operational laws (AOLs) under spherical FS. In addition, elaborate on the deficiency of existing AOLs and present three cases to address the validity of the proposed novel AOLs under spherical fuzzy settings. Second, we present a list of Einstein aggregation operators (AgOp) based on the Einstein norm to aggregate uncertain information in DMPs. Thirdly, we are introducing two techniques to demonstrate the unknown weight of the criteria. Fourthly, we develop extended TOPSIS and Gray relational analysis approaches based on AgOp with unknown weight information of the criteria. In fifth, we design three algorithms to address the uncertainty and ambiguity information in emergency DMPs. Finally, the numerical case study of the novel carnivorous (COVID-19) situation is provided as an application for emergency decision-making based on the proposed three algorithms. Results explore the effectiveness of our proposed methodologies and provide accurate emergency measures to address the global uncertainty of COVID-19.     

再加工工艺对N18锆合金板残余应力的影响

采用“热轧+冷轧+退火”工艺对N18锆合金板进行再加工,通过X射线法分析板的表面残余应力,采用EBSD技术分析晶界取向差角分布和大小角度晶界。结果表明,热轧后N18锆合金板的表面残余应力呈现无规律分布状态;冷轧后板的表面残余应力均为压应力,其大小随着冷轧变形量的增加而增加;退火后板的表面残余应力值处于较低水平,当退火制度为500 ℃×2 h时,残余应力处于最低水平;退火后,板的微观结构以小角度晶界为主,且随着退火温度的升高,小角度晶界的密度先增加后趋于稳定。     

Effect of cutting process on the stress corrosion susceptibility of AISI 304L stainless steel

During manufacturing of a component, cutting, turning, grinding, and milling operations are inevitable and these operations induce surface residual stresses. In this study, it is shown that, depending on the process employed for cutting, residual stresses generated at the cut surfaces can vary widely and they can, in turn, make the cut surfaces of austenitic stainless steel (SS) prone to stress corrosion cracking (SCC). An austenitic SS 304L plate was cut using three different procesess: bandsaw cutting, cutting using the cut-off wheel, and shearing. Surface residual stress measurement using the X-ray diffraction (XRD) technique is carried out close to the cutting edges and on the cross-section. SCC susceptibility studies were carried out as per ASTM G36 in 45% boiling magnesium chloride solution. Optical microscopic examination showed the presence of cracks, and confocal microscopy was used to measure the depth of cracks. The study confirmed that high tensile residual stresses present in the cut surfaces produced by cut-off wheel and shear cutting make the surfaces susceptible to SCC while the surfaces produced by bandsaw cutting are resistant to SCC. Hence, it is shown that there is a definite risk of SCC for product forms of austenitic SS with cut surfaces produced using cutting processes that generate high tensile residual stresses stored for a long period of time in a susceptible environment.     

镓锗铜萃余液综合利用研究与应用

本文针对镓锗铜综合回收系统产出的萃余液,设计通过中和氧化除杂、锌粉置换除铜镉、有机试剂除钴镍、纯碱法生产碱式碳酸锌、高温煅烧生产活性氧化锌和一步法生产元明粉的生产工艺流程,有效的将镓锗铜萃余液分类分离富集,产出具有价值的富集渣和工业产品,进一步提升综合回收经济效益。     

改进的卷积神经网络单幅图像超分辨率重建

针对经典的基于卷积神经网络的单幅图像超分辨率重建方法网络较浅、提取的特征少、重建图像模糊等问题，提出了一种改进的卷积神经网络的单幅图像超分辨率重建方法，设计了由密集残差网络和反卷积网络组成的新型深度卷积神经网络结构。原始低分辨率图像输入网络，利用密集残差学习网络获取更丰富的有效特征并加快特征梯度流动，其次通过反卷积层将图像特征上采样到目标图像大小，再利用密集残差学习高维特征，最后融合不同卷积核提取的特征得到最终的重建图像。在Set5和Set14数据集上进行了实验，并和Bicubic、K-SVD、SelfEx、SRCNN等经典重建方法进行了对比，重建出的图像在整体清晰度和边缘锐度方面更好，另外峰值信噪比（PSNR）平均分别提高了2.69?dB、1.68?dB、0.74?dB和0.61?dB。实验结果表明，该方法能够获取更丰富的细节信息，得到更好的视觉效果，达到了图像超分辨率的增强任务。     

基于深度学习的彩色以及近红外图像去马赛克

单传感器捕获的彩色-近红外（RGB-NIR）图像存在光谱干扰，从而导致重建出的标准彩色图像（RGB）图像与近红外（NIR）图像存在色彩失真以及细节信息模糊。针对这个问题提出一种基于深度学习的去马赛克方法，通过引入跳远连接与稠密连接解决了梯度消失和梯度弥散问题，使得网络更容易训练，并且提升了网络的拟合能力。首先，用浅层特征提取层提取了马赛克图像的像素相关性以及通道相关性等低级特征；然后，将得到的浅层特征图输入到连续多个的残差稠密块以提取专门针对去马赛克的高级语义特征；其次，为充分利用低级特征与高级特征，将多个残差稠密块提取到的特征进行组合；最后，通过全局跳远连接恢复最终的RGB-NIR图像。在深度学习框架Tensorflow上使用公共的图像与视觉表示组（IVRG）数据集、有植被的户外多光谱图像（OMSIV）数据集和森林（Forest）三个公开数据集进行实验。实验结果表明，所提方法优于基于多级自适应残差插值、基于卷积卷积和神经神经网络以及基于深度残差U型网络的主流的RGB-NIR图像去马赛克方法。     

2219铝合金搅拌摩擦焊和钨极保护焊焊接件残余应力的中子衍射研究

利用中子衍射法对2219铝合金搅拌摩擦焊（FSW）和钨极保护焊（TIG）焊接件开展了三维残余应力测量，并对残余应力分布规律进行了分析。结果表明：焊接件的纵向残余应力数值较大；FSW焊接件残余应力整体较TIG焊接件的小；FSW和TIG焊接件的残余拉应力最大值分别为101 MPa和174 MPa，FSW焊接件残余拉应力最大值较TIG焊接件的小；FSW残余拉应力最大值处于轴肩边缘，且前进侧峰值大于后退侧峰值；TIG焊接件残余拉应力最大值处于焊缝边缘。通过中子衍射实验获得的焊接件残余应力分布，将可用于焊接工艺的优化与焊接件的寿命预测。