Home > Research > Publications & Outputs > Pore-scale numerical investigation of pressure ...

Associated organisational unit


Text available via DOI:

View graph of relations

Pore-scale numerical investigation of pressure drop behaviour across open-cell metal foams

Research output: Contribution to Journal/MagazineJournal articlepeer-review

  • T. P. de Carvalho
  • Herve Morvan
  • David Hargreaves
  • Hatem Oun
  • A. Kennedy
<mark>Journal publication date</mark>03/2017
<mark>Journal</mark>Transport in Porous Media
Issue number2
Number of pages26
Pages (from-to)311-336
Publication StatusPublished
Early online date15/03/17
<mark>Original language</mark>English


The development and validation of a grid-based pore-scale numerical modelling methodology applied to five different commercial metal foam samples is described. The 3-D digital representation of the foam geometry was obtained by the use of X-ray microcomputer tomography scans, and macroscopic properties such as porosity, specific surface and pore size distribution are directly calculated from tomographic data. Pressure drop measurements were performed on all the samples under a wide range of flow velocities, with focus on the turbulent flow regime. Airflow pore-scale simulations were carried out solving the continuity and NaviertextendashStokes equations using a commercial finite volume code. The feasibility of using Reynolds-averaged NaviertextendashStokes models to account for the turbulence within the pore space was evaluated. Macroscopic transport quantities are calculated from the pore-scale simulations by averaging. Permeability and Forchheimer coefficient values are obtained from the pressure gradient data for both experiments and simulations and used for validation. Results have shown that viscous losses are practically negligible under the conditions investigated and pressure losses are dominated by inertial effects. Simulations performed on samples with varying thickness in the flow direction showed the pressure gradient to be affected by the sample thickness. However, as the thickness increased, the pressure gradient tended towards an asymptotic value.