Study on unsteady tip leakage vortex cavitation in an axial-flow pump using an improved filter-based model

Desheng Zhang, Lei Shi, Ruijie Zhao, Weidong Shi, Qiang Pan, B. P.M.Bart van Esch

Research output: Contribution to journalArticleAcademicpeer-review

33 Citations (Scopus)


The aim of the present investigation is to simulate and analyze the tip leakage flow structure and instantaneous evolution of tip vortex cavitation in a scaled axial-flow pump model. The improved filter-based turbulence model based on the density correction and a homogeneous cavitation model were used for implementing this work. The results show that when entering into the tip clearance, the backward flow separates from the blade tip near the pressure side, resulting in the generation of a corner vortex with high magnitude of turbulence kinetic energy. Then, at the exit of the tip clearance, the leakage jets would re-attach on the blade tip wall. Moreover, the maximum swirling strength method was employed in identifying the TLV core and a counter-rotating induced vortex near the end-wall successfully. The three-dimensional cavitation patterns and in-plain cavitation structures obtained by the improved numerical method agree well with the experimental results. At the sheet cavitation trailing edge in the tip region, the perpendicular cavitation cloud induced by TLV sheds and migrates toward the pressure side of the neighboring blade. During its migration, it breaks down abruptly and generates a large number of small-scale cavities, leading to severe degradation of the pump performance, which is similar with the phenomenon observed by Tan et al. [35].

Original languageEnglish
Pages (from-to)659-667
Number of pages9
JournalJournal of Mechanical Science and Technology
Issue number2
Publication statusPublished - 1 Feb 2017


  • Axial-flow pump
  • Density correction method of filter-based model
  • High-speed photography
  • Tip leakage flow
  • TLV cavitation


Dive into the research topics of 'Study on unsteady tip leakage vortex cavitation in an axial-flow pump using an improved filter-based model'. Together they form a unique fingerprint.

Cite this