The arrangement of electrolyte inlet in the copper electro-refining(ER)cell has a great influence on the local flow field,which affects the distribution of electrical current density in consequence.In order to understand the complicated phenomena ofelectrolyte flow behavior in vertical counter electrodes in full-scale copper ER cell,the three-dimensional computational fluiddynamics(CFD)models with four different arrangements of electrolyte inlets,i.e.,single inlet(SI),central bottom inlets(CBI),topside interlaced inlets(TII),and bottom side interlaced inlets(BII),were established to simulate the flow behavior.Simulation resultshave revealed that the parallel injection devices help to improve the electrolyte velocity between electrodes,and while the relativerange of electrolyte velocity in CBI exceeds that of TII and BII,which is more than4times,indicating its severer unequal flowdistribution.Meanwhile,the average velocity of electrolyte in BII is4times larger than that of SI due to its higher turbulenceintensity.Generally,one of the efficient ways to supply fresh copper solution rapidly and uniformly into the inter-electrode space is toadapt the arrangement of BII.By utilizing such an arrangement,the electro-refining under high electrical current density is possible,and the productivity can be increased in sequence.
To investigate the differences and the development trends of the 400 kA aluminum reduction cell, four representative cells were deeply analyzed. By using numerical simulation methods in ANSYS software, the structure parameters were firstly compared, and then three-dimensional models of electric-magnetic-flow field were built and solved with finite element method(FEM). The comparison of the structures reveals that the cell bodies are similar while the current flow path and distribution ratio of bus bars are different. It appears that most of the current(70%-80%) in side A are used as the magnetic field compensation current and flow through two ends. The numerical simulation results indicate that the distributions of magnetic fields are different but all satisfy with the magnetohydrodynamics(MHD) stabilization, and the flow patterns are all two or multi vortexes with appropriate velocities. The comparison shows that all studied cells can satisfy with the physical field requirement, and the commercial applications also verify that the 400 kA cells have become the product of the mature and world's leading technology.
针对传统的ε-不敏感支持向量回归机(ε-insensitive support vector regression,ε-SVR)未充分考虑局部支持向量对回归预测结果的影响,不利于提高回归预测精度的问题,提出了一种ε-SVR预测误差校正方法。该方法以期望预测值与ε-SVR回归预测值及局部支持向量间的欧氏距离和最小为目标函数,以ε不敏感损失带(ε-tube)宽度为约束条件,通过利用高维特征空间中ε-tube边界上和边界外的局部支持向量对ε-SVR的回归预测值进行误差校正。利用人工产生的不同分布数据集和UCI数据集进行的仿真结果表明,与传统的ε-SVR相比,该文方法具有更高的预测精度和更强的泛化能力。
This paper focuses on the distributed parameter modeling of the zinc electrowinning process(ZEWP)to reveal the spatiotemporal distribution of concentration of zinc ions(CZI)and sulfuric acid(CSA)in the electrolyte.Considering the inverse diffusion of such ions in the electrolyte,the dynamic distribution of ions is described by the axial dispersion model.A parameter estimation strategy based on orthogonal approximation has been proposed to estimate the unknown parameters in the process model.Different industrial data sets are used to test the effectiveness of the spatiotemporal distribution model and the proposed parameter estimation approach.The results demonstrate that the analytical model can effectively capture the trends of the electrolysis reaction in time and thus has the potential to implement further optimization and control in the ZEWP.
Reagents are optimized for the simultaneous determination of trace amounts of Cu^(2+), Cd^(2+) and Co^(2+) in zinc sulfate solution, which contains an extremely large excess of Zn^(2+). First, the reagents and their doses for the experiment are selected according to the characteristics of the zinc sulfate solution. Then, the reagent doses are optimized by analyzing the influence of reagent dose on the polarographic parameters(i.e. half-wave potential E_(1/2) and limiting diffusion current I_p). Finally, the optimization results are verified by simultaneously determining trace amounts of Cu^(2+), Cd^(2+) and Co^(2+) in the presence of an extremely large excess of Zn^(2+). The determination results indicate that the optimized reagents exhibit wide linearity, low detection limits, high accuracy and good precision for the simultaneous determination of trace amounts of Cu^(2+), Cd^(2+) and Co^(2+) in the presence of an extremely large excess of Zn^(2+).
WANG Guo-weiYANG Chun-huaZHU Hong-qiuLI Yong-gangGUI Wei-hua
Region partition(RP) is the key technique to the finite element parallel computing(FEPC),and its performance has a decisive influence on the entire process of analysis and computation.The performance evaluation index of RP method for the three-dimensional finite element model(FEM) has been given.By taking the electric field of aluminum reduction cell(ARC) as the research object,the performance of two classical RP methods,which are Al-NASRA and NGUYEN partition(ANP) algorithm and the multi-level partition(MLP) method,has been analyzed and compared.The comparison results indicate a sound performance of ANP algorithm,but to large-scale models,the computing time of ANP algorithm increases notably.This is because the ANP algorithm determines only one node based on the minimum weight and just adds the elements connected to the node into the sub-region during each iteration.To obtain the satisfied speed and the precision,an improved dynamic self-adaptive ANP(DSA-ANP) algorithm has been proposed.With consideration of model scale,complexity and sub-RP stage,the improved algorithm adaptively determines the number of nodes and selects those nodes with small enough weight,and then dynamically adds these connected elements.The proposed algorithm has been applied to the finite element analysis(FEA) of the electric field simulation of ARC.Compared with the traditional ANP algorithm,the computational efficiency of the proposed algorithm has been shortened approximately from 260 s to 13 s.This proves the superiority of the improved algorithm on computing time performance.
