Skip to main navigation Skip to search Skip to main content

Pressure drop and optimization meta-models for arbitrary low-height pleated filter shapes and flowrates

  • Paul Choi
  • , Christian Ariane Santos
  • , Min Kun Kim
  • , Hyunsook Jung
  • , Do Young Hong
  • , Junemo Koo

Research output: Contribution to journalArticlepeer-review

7 Citations (Scopus)

Abstract

This study delivers equations useful for low-height pleated fibrous filter design: two pressure drop equations and one set of optimum design equations applicable to arbitrary pleated filter shapes and flowrates. The pressure drop equations were derived to predict the pressure loss of the pleated filter. They were made through regression analysis with a total of 1024 CFD data. The set of optimum design equations was developed to find the optimum filter shape minimizing pressure drop. All equations were validated through the 8-fold cross-validation method and were accurate enough to replace the CFD simulations. Additionally, novel contour plots were made to describe how optimum filter geometry changes due to flowrate, height, and media permeability. The delivered equations were applied to an actual filter design problem and verified with additional CFD simulations. This study allows filter designers to predict the pressure loss and to design the optimum filter shape without any simulations.

Original languageEnglish
Pages (from-to)5007-5022
Number of pages16
JournalJournal of Mechanical Science and Technology
Volume35
Issue number11
DOIs
Publication statusPublished - Nov 2021

Bibliographical note

Publisher Copyright:
© 2021, The Korean Society of Mechanical Engineers and Springer-Verlag GmbH Germany, part of Springer Nature.

Keywords

  • Computational fluid dynamics
  • Latin-hypercube sampling
  • Optimization
  • Pleated fibrous filter
  • Pressure drop
  • Regression analysis

Fingerprint

Dive into the research topics of 'Pressure drop and optimization meta-models for arbitrary low-height pleated filter shapes and flowrates'. Together they form a unique fingerprint.

Cite this