Skip to main navigation Skip to search Skip to main content

A comparative analysis of 2-D and 3-D simulation for savonius hydrokinetic turbine array

  • Xi'an Jiaotong University

Research output: Contribution to journalArticlepeer-review

21 Scopus citations

Abstract

Recently, Savonius hydrokinetic turbine arrays have garnered increasing research attention, with two-dimensional (2-D) simulations being the predominant method due to the substantial computational cost associated with three-dimensional (3-D) simulations. However, the 2-D results were proven to overestimate individual turbine performance and underestimate the wake flow recovery rate compared to the 3-D results. This study investigated the differences in the prediction of wake flow and local coupling effect between the two simulations of Savonius turbine arrays. The results indicate that the 3-D wake shape and distribution are similar to the experimental tests. However, the 2-D wake shape differs significantly from the 3-D wake, with a much lower wake recovery rate. Additionally, interference effects of different wake patterns in the two methods lead to discrepancies in flow direction impacts. 2-D simulations exhibit higher flow velocities and more pronounced low-pressure regions in the blocked region than 3-D results. As a result, 2-D simulations can be used in arrays where turbine wakes do not interfere with other turbines, providing insights into the analysis of parameter trends or serving as a preliminary reference for 3-D simulations. However, for layouts where turbine wakes interfere with each other, 3-D simulations or experimental tests are necessary.

Original languageEnglish
Article number116909
JournalOcean Engineering
Volume295
DOIs
StatePublished - 1 Mar 2024

Keywords

  • Coupling effect
  • Savonius hydrokinetic turbine array
  • Three-dimensional simulations
  • Two-dimensional simulations
  • Wake effect

Fingerprint

Dive into the research topics of 'A comparative analysis of 2-D and 3-D simulation for savonius hydrokinetic turbine array'. Together they form a unique fingerprint.

Cite this