Reference: D.P. Bentz, E.J. Garboczi, and D.A. Quenard, Mod. and Sim. in Mater. Sci. and Eng. 6, 211-236 (1998).

(PDF Version of Original paper)

Go back to Part II Chapter 7 Sec. (2) Go back to Table of Contents

Next: Introduction

Modelling Drying Shrinkage in Reconstructed Porous Materials:
Application to Porous Vycor Glass

Dale P. Bentz and Edward J. Garboczi
National Institute of Standards and Technology
Gaithersburg, MD 20899 USA

Daniel A. Quenard
Centre Scientifique et Technique du Batiment
38400 Saint-Martin d'Heres, FRANCE


A three-dimensional representation of the microstructure of porous Vycor glass was generated from a transmission electron micrograph, and was analyzed to compute the locations of all capillary-condensed water as a function of relative humidity. On solid surfaces where capillary-condensed water was not present, an adsorbed water layer, whose thickness is a function of relative humidity, was placed. As a function of relative humidity, fixed pressures were specified in all capillary-condensed water, and the change in specific surface free energy with relative humidity was taken into account for the adsorbed water layers. New finite element codes were developed to determine the drying shrinkage, in response to the changes in the specific surface free energy of the adsorbed water layers and to the fixed pressures in the capillary condensed water. Existing finite element and finite difference codes were used to evaluate the elastic moduli, the electrical and thermal conductivity, and the fluid permeability of the material. Bulk properties like fluid permeability and electrical and thermal conductivity agreed well with experiment. By adjusting the elastic moduli of the solid backbone, which are not experimentally determined quantities, the computed porous glass elastic moduli, and computed low and high relative humidity shrinkage all agreed well (< 10%) with experimental values. At intermediate relative humidities, the agreement for shrinkage was worse, partly due to inaccuracies in the simulated water desorption curve, and partly due to the fact that water-induced swelling of the solid backbone, an effect that is probably present in the real material, was not taken into account in the model computations.

Go back to Part II Chapter 7 Sec. (2) Go back to Table of Contents