Testing of Permeability Models through Tow-scale Experiments and Simulations
Mentor 1
Krishna M. Pillai
Location
Union Wisconsin Room
Start Date
27-4-2018 1:00 PM
Description
A rigorous test of important theoretical models for tow permeability, backed by numerical simulations, is conducted using a novel tow-scale experiment. The models include those for flow along and across aligned fibers as well as for flow through random fibers. The domains for numerical simulations were created by randomly distributing parallel fibers in a cube-like unit cell using the software Geodict. Two separate simulations were considered: 1) Stokes-flow solution using Geodict, 2) Whitaker’s closure-formulation solution using COMSOL. The falling-head parameter was adapted to measure the permeability along and across the fibers. Multiple measurements were conducted for each of the five bundle types to establish repeatability and estimate scatter. The permeabilities obtained through experiments matched with those from the theoretical and numerical methods. But numerical permeabilities for the longitudinal flow were exceptionally accurate. Also, the specialized permeability models for longitudinal and transverse flows were more accurate than the random-fiber models.
Testing of Permeability Models through Tow-scale Experiments and Simulations
Union Wisconsin Room
A rigorous test of important theoretical models for tow permeability, backed by numerical simulations, is conducted using a novel tow-scale experiment. The models include those for flow along and across aligned fibers as well as for flow through random fibers. The domains for numerical simulations were created by randomly distributing parallel fibers in a cube-like unit cell using the software Geodict. Two separate simulations were considered: 1) Stokes-flow solution using Geodict, 2) Whitaker’s closure-formulation solution using COMSOL. The falling-head parameter was adapted to measure the permeability along and across the fibers. Multiple measurements were conducted for each of the five bundle types to establish repeatability and estimate scatter. The permeabilities obtained through experiments matched with those from the theoretical and numerical methods. But numerical permeabilities for the longitudinal flow were exceptionally accurate. Also, the specialized permeability models for longitudinal and transverse flows were more accurate than the random-fiber models.