A pore network study of evaporation from the surface of a drying non‐hygroscopic porous medium

Abstract

The phenomena occurring at the surface of a porous medium during drying in the capillary regime are investigated by pore network simulations. The impact of the formation of wet and dry patches at the surface on the drying rate is studied. The simulations indicate an edge effect characterized by a noticeable variation of saturation in a thin layer adjacent to the porous surface. Also, the results indicate a significant nonlocal equilibrium effect at the surface. The simulation results are exploited to test Schlünder's classical model which offers a simple closure relationship between the evaporation rate and the degree of occupancy of the surface by the liquid. In addition to new insights into the surface phenomena, the results open up new prospects for improving the continuum models of the drying process. © 2017 American Institute of Chemical Engineers AIChE J, 64: 1435–1447, 2018

About Schlünder’s model: A numerical study of evaporation from partially wet surfaces

“…Also, we have considered a surface where the vapor concentration is the same at the surface of each pore. As illustrated for instance in Moghaddam et al [9] from pore network simulations, the vapor concentration at the surface pore entrances actually varies during drying because of the invasion of an increasing fraction of pores by the gas phase. Both the wet pores and the dry pores at the surface contribute to the evaporation rate since vapor transport can also occur through the dry pores.…”

“…One can thus wonder whether Schl€ under's formula or a variant of Schl€ under's formula could be useful in relation with the PNM of drying. In this respect, it should be noted that the results presented in Moghaddam et al [9] as well as in previous works using a pore network model, for example, Yiotis et al, [20] were based on simulations using a quite coarse grid in the external mass transfer boundary layer. Clearly, this aspect of PNM needs improvement.…”

“…As presented for example in Moghaddam et al [9] or Metzger et al, [25] a method for computing the diffusive external mass transfer in the pore network models of drying consists in setting computational nodes in the external layer of size d. However, as illustrated in Figure 10, the distance between two computational nodes is generally taken as the distance between two pores in the network (also referred to as the lattice spacing a). Obviously, this is a quite coarse discretization, sufficient to get qualitative results but questionable if the objective is to quantitatively compute the transfer by diffusion in this layer.…”

“…Schl€ under's relationship thus analyzes the CRP from a pure external mass transfer standpoint. However, as discussed in Moghaddam et al and Lehmann and Or, [9,10] drying is less simple than considered in Schl€ under's analysis. For instance, the pore sizes are not uniform at the surface and vary from one pore to another.…”

The Brooks and Corey Capillary Pressure Model Revisited from Pore Network Simulations of Capillarity-Controlled Invasion Percolation Process

“…In pore network modeling, the void space of a porous medium is approximated by a network of connected throats and pores (herein called throat-pore model, TPM) and transport phenomena are described by unidirectional approximations at the level of individual throats and pores. Uses of PNMs for the computation of macroscopic parameters (such as permeability, relative permeability and capillary pressure curve) have already been elucidated in numerous previous studies (e.g., [20][21][22][23]) as well as in recent literature (e.g., [4,[24][25][26][27][28]). Nevertheless, none of those works have systematically investigated the impact of pore structure on the Brooks and Corey capillary pressure model parameters.…”