Numerical and Experimental Investigation of Secondary Flow and Loss Generation in Inducers

对于诱导轮内的二次流，目前还没有定量的研究，如B图。Lakshminarayana（1996）没有考虑在内。B。Lakshminarayana的图片只提供了一个定性的图片，不包含有关单个区域（3D边界层，3D分离线，奇异点，核流等）的范围的信息。不包括叶片、轮毂和机匣上的二次流图片。这一知识差距将由预期的项目填补。将对几个诱导轮进行CFD流动模拟，并描述二次流。通过测量（性能和现场测量）和油图，对计算流体动力学结果进行验证。此外，还应查明实际设计规则是否适用于确定诱导轮的轴向长度。对于轴向长度的确定，Li的奇异性方法提供了一种合适的方法，尽管从申请人的观点来看，损耗模型是不够的。其目标是：1）通过CFD模拟，确定80%、100%和120%（最佳效率点）的三维边界层的流动拓扑。申请人意识到，通过CFD只能部分实现这一目标。但是，申请人成功地将压缩机叶栅的实验和数字油图结合在一起，成功地完成了DFG项目，因此，他认为，尽管他非常清楚这种方法的缺点，但它是可行的。2)流动拓扑结构由实验油图确定。因此，油漆（公司：Leinos类型：自然白色，Nr。820，密度：1.19 g/cm³）与石油（密度约0.85 g/cm³）混合，使混合物具有接近水的密度。在主席SAM调查进行，以测试这种方法的可行性，显示令人满意的结果。这种方法也适用于布伦瑞克的Pfleiderer研究所（今天的IFAS）的轴向叶轮泵。3)要确定前缘后掠对轮毂区域流动拓扑的影响。申请人假设后掠前缘对轮毂边界层的流动拓扑结构具有可识别的影响，但对流动损失的影响很小。

Until now there are no quantitative investigations about secondary flow in inducer impellers, if the pictures of B. Lakshminarayana (1996) are not taken into account. B. Pictures of Lakshminarayana only provide a qualitative picture and contains no information about the extent of single regions (3D-boundary layer, 3D-separation lines, singular points, kernel flow etc.). Secondary flow pictures on blades, hub and casing are not included. This knowledge gap is to be closed by the intended project. CFD flow simulations are to be performed for several inducers and the secondary flow is to be depicted. By measurements (performance- and field measurements) and oil pictures the CFD results are to be validated. Furthermore it is to be found out whether the actual design rules are appropriate to determine the axial length of an inducer. For the determination of the axial length the singularity method of Li provides a suitable method, although the loss model is insufficient from the applicant point of view. The aims are: 1) By CFD-simulations the flow topology of the three-dimensional boundary layer is to be determined for 80%, 100% and 120% (Best Point of Efficiency). The applicant is aware that by CFD this aim can be reached only partly. But the applicant succeeded in bringing together experimental and numerical oil pictures of compressor cascades in a successfully closed DFG-Project, so that he is of the opinion that it will work although he knows very well the shortcomings of this method. 2) The flow topology is to be determined by experimental oil pictures. Therefore oil paint (Company: Leinos Typ: Natural-White, Nr. 820, Density: 1,19 g/cm³) is to be mixed with petroleum (density approximately 0.85g/cm³) so that the mixture possesses nearly the density of water. At the chair SAM investigations were performed in order to test the feasibility of this method which shows satisfactory results. This method was also applied for an axial impeller pump at Pfleiderer Institute of TU Braunschweig (today IFAS). 3) The influence of sweep of the leading edge on the flow topology of the hub region is to be determined. The applicant supposes that the swept leading edge has an identifiable influence on the flow topology of the hub boundary layer, but that the influence on the flow losses is small.