Mass Transfer Model for Predicting Water Uptake in
Coated Concrete
J. Liu and C. Vipulanandan
Center for Innovative
Grouting Materials and Technology (CIGMAT)
Department of Civil and
Environmental Department
University of Houston,
Houston , TX , 77204-4003
Tel: (713) 743-4291 E-mail Address: jliu5@mail.uh.edu
Abstract
A mass transfer model was developed to predict the
weight increase in coated concrete in a non-reactive solution. The model was
verified using epoxy coated concrete specimens. The mass transfer model
parameters for water-coating combinations were obtained from controlled
experiments.
Introduction
When cement concrete specimens are submerged in liquids for testing, the weight change of the concrete is a key factor that indicates the change in concrete. Mebarkia and Vipulanandan (1995)[1] developed a cylindrical model to predict weight increase in polymer concrete when submerged in water. For coated concrete, because of the porous nature of the substrate, the penetration of coating into the concrete is a very important factor and the interface properties may have affected the penetration of liquids; therefore, it was necessary to incorporate the interface properties related to coating-concrete in modeling the coated concrete behavior.
Modeling
When coated concrete comes into contact with water, it will
penetrate the coating film into the concrete. The Physical model for coated
concrete when immersed in water is shown in Fig.1.

Where SoCT ultimate degree of saturation in coating
film, g water/cm3 solid;
SiCO degree of saturation on the interface of
coating film and concrete substrate, g water/cm3 solid;
S(t) Degree
of saturation distribution of liquid inside corroded zone of concrete cylinder,
g water/cm3 solid.
![]()
For mass transport in cylindrical specimens, if the degree of saturation is a function of radius and time only, the second order differential equation is:
, (1)
where
DCO is the mass transfer coefficient of the substrate and r is the
radius of cylinder.
For liquid transport in a coated concrete
cylinder without chemical reaction, the degree of saturation on the concrete
surface changes with time. If the degree of saturation at the concrete surface
is
, assuming
, the sorption-time curve is given by Crank (1975)
. (2)
From Fig. 2, the solution uptake is
determined by the parameter
which represents the effects of the coating material
properties and the parameter
(x-axis) which
represents the time effect. Let
.
Approximating
Eq. (3) and considering an exponential function of the form
(3)
where
m is a constant.
The Eq. (3) is best fitted to the
standard curves in Fig. 2 for different lCO values by using the least-square method. The value of m
varied from 0.98 to 1.26 when the value of l was in the range of 0.01 to 5.
The
approximate solution for Eq. (3) is in good agreement with the solution
suggested by Crank (1975) (Fig. 2). The solid curves are the standard curves
(Eq. (3)) as given by Crank (1975) while the dotted lines are the approximate
solution from Eq. (3).

Figure 3.
Comparison of the experimental Data with Model Predictions in D. I. Water

Figure 2. Comparing Exact (Solid Line) Solution to
the Approximate Solutions
The mass transfer coefficients of
concrete (DCO) can be obtained by fitting the cylindrical model
(Mebarkia, 1995[1]) to experimental data of uncoated concrete specimens. The
values of mass transfer coefficients was 2.55 ´10-10 m2/s in D. I. water. Using Eq.
(3) and experimental data, the parameters n, l, and
can be obtained for
different coatings. The comparisons of model prediction (Eq. (4)) to experiment
data are shown in Fig. 3.
A model was developed to predict the weight increase of coated concrete in water. Parameters in the model can be obtained from controlled experiments. The model prediction was in agreement with the test results.
This project was supported by the Center for Innovative Grouting Materials and Technology (CIGMAT) under grants from the City of Houston, National Science Foundation (CMS-9526094, CMS-9634685), and various industries.
Reference: