失效双氧水催化剂中钯的浸出-沉淀富集效率研究

研究了从失效双氧水催化剂中富集钯的工艺,采用“焙烧浸出-硫化沉钯”的工艺实现了钯的高效富集。重点考察在浸出中硫酸浓度、浸出温度、浸出时间、盐酸浓度、NaClO3用量对富集倍数的影响。研究了富集钯的最佳工艺条件,结果表明,钯平均富集12.09倍,钯的平均回收率为99.55%。     

颗粒浓度及粒度的光散射在线测量

介绍一种在线监测颗粒浓度及粒度变化的光散射方法。着重提出消除观察窗玻璃污染物对测量结果影响的办法。给出了部分现场实测结果。     

公路隧道CO设计浓度限值的分析研究

现行规范中给出的CO设计浓度对隧道长度、人员不同通行方式的考虑过于粗略,且未考虑海拔高度的影响。本文分析了CO对人体危害的机理,通过Coburn-Forster-Kane模型分析了CO浓度与人体血液中产生的COHb浓度之间的关系,从而得出了公路隧道内经历时间与CO浓度限值之间的关系,并分析了海拔高度、通风方式以及人员不同通行方式对CO浓度限值的影响。通过分析结果可以看出,规范中规定的CO设计浓度在经历时间较长的情况下将不再适用,取值过高,应进行修正。本文分别计算出COHb浓度控制为2％和1.5％时,在不同海拔高度的公路隧道CO设计浓度限值,并整理成表格,可在公路隧道在通风设计中参考使用。     

运用神经网络预测炸药爆速

本文计算了30个炸药分子的氧平衡(OB)、装药密度(ρ)、碳原子浓度nC/Vm、氢原子浓度nH/ Vm、氧原子浓度nO/ Vm、氮原子浓度nN/ Vm 6种参数,以这6种参数作为炸药分子结构描述符,构建人工神经网络模型,预测炸药爆速.预测结果表明,预测值和文献值的最大相对误差为5.35%,最小为0.53%,说明神经网络所确定的定量关系可以很好地用来预测炸药爆速.     

粗糙集傅立叶神经网络及在故障诊断中应用

提出了粗糙集和傅立叶神经网络相结合的方法,进行粗糙集布尔逻辑离散化,并在此基础上求取初始隶属函数,以提高隶属函数准确性;再使用傅立叶神经网络进行诊断网络训练。以连续搅拌反应釜故障诊断为实例,研究结果表明,此方法可以减少网络训练时间并提高诊断精度,有效进行故障诊断。     

Application of Mellin Transform in Wideband Underwater Acoustic Signal Processing

According to the features of the wideband underwater acoustic signals, an algorithm for the wideband ambiguity function is put forward based on Mellin transform. The wideband acoustic signal processing using the fast Mellin transform is also explored. The theoretical analysis and simulation results show that the algorithm has not only high computation efficiency but also good concentration in wideband ambiguity domain. It suits for the wideband underwater acoustic signal processing.     

泥沙输移的理论研探

本文通过对泥沙颗粒在水中的受力情况分析，从理论上导出了描述推移质运动的数学表达式．在此基础上，又导得了挟沙水流的底部含沙量，从而使泥沙运动的扩散方程能够方便地运用到工程实际问题之中．     

选冶联合法从玻璃纤维工业浇注料中回收铂铑

采用选矿和冶金联合工艺从玻纤浇注料中回收铂和铑。首先采用重力选矿法使玻纤浇注料中90%⁹3%的铂铑得到初步富集分离,得到品位为60%93%的铂铑得到初步富集分离,得到品位为60%⁸0%的铂铑精矿和尾矿,低品位尾矿采用Na2O-Al2O3-SiO2低熔点渣系,通过铅熔炼捕集、吹炼除铅工艺产出铂铑富集物料,然后通过精炼得到合格的铂铑金属产品。结果表明,全流程铂和铑的回收率分别大于99%和98.5%。     

Surface activity of quaternary ammonium salts derived from jojoba oil

A number ofcis andtrans quatenary ammonium salts were synthesized from jojoba oil. All derivatives were found to be surface active agents, i.e., they reduced the surface tension to 35 dynes/cm⁻¹, at very low concentrations. The relationship between the surface activity and the molecular structure is discussed.     

A correlation for predicting diffusion coefficients in alkanes

A correlation was developed for predicting limiting mutual diffusion coefficients of gaseous and liquid solutes in normal alkane solvents with carbon numbers ranging from 7 to 28 at temperatures to 570 K. The functional form of the correlation is based on the rough hard sphere theory. The parameters needed in the correlation are the molecular weights and hard sphere diameters of the solute and solvent molecules and molar volume of the solvent. The data used in the development of the correlation included 143 measurements with solute/solvent molecular mass ratios ranging from 0.005 to 2.26 and size ratios ranging from 0.3 to 1.3. The average absolute error in predictions was only 6.3%. When the correlation was used to predict literature values for binary n-alkane systems, the average absolute error was 16.7%.