Skip to main navigation Skip to search Skip to main content

Thermal design and optimization of plate-fin heat exchangers based global sensitivity analysis and NSGA-II

  • Huizhu Yang
  • , Jian Wen
  • , Simin Wang
  • , Yanzhong Li

Research output: Contribution to journalArticlepeer-review

37 Scopus citations

Abstract

Morris and Sobol global sensitivity analyses are firstly studied to evaluate the effects of the design parameters on the performance of plate-fin heat exchanger with serrated fins. Five design parameters such as the fin height, fin space, fin thickness and interrupted length and channel Reynolds number are selected as optimization parameters. Heat flow rate, total annual cost and entropy generation number are considered as objective functions. Furthermore, the five design parameters are optimized by NSGA-II, in which the effectiveness of the proposed algorithm is validated by comparing a previous approach. The results show that Reynolds number and fin space s play a dominant role on the performance of plate-fin heat exchanger with serrated fin. Meanwhile, the interaction effects among these parameters are significant on total annual cost and entropy generation number. Compared with one common design, the thermal performance factor of another common design and two optimal parameters increases 8.4%, 41.2% and 58.4%, respectively. Combining global sensitivity analysis and NSGA-II has the characteristics of easy to implement and high accuracy. Therefore, it is universal and can be used for various complex engineering applications to provide guidance for better selecting the suitable parameters.

Original languageEnglish
Pages (from-to)444-453
Number of pages10
JournalApplied Thermal Engineering
Volume136
DOIs
StatePublished - 25 May 2018

Keywords

  • Genetic algorithm
  • Morris method
  • Plate-fin heat exchanger
  • Sensitivity analysis
  • Serrated fin
  • Sobol method

Fingerprint

Dive into the research topics of 'Thermal design and optimization of plate-fin heat exchangers based global sensitivity analysis and NSGA-II'. Together they form a unique fingerprint.

Cite this