Chemical admixtures - Concrete Mix Design, Quality Control and Specification - page 80

Air

Liquid

θ

Liquid

θ

Solid

Dicult to wet
θ is large

Easy to wet
θ is small

Figure 4.2 Effect of contact angle on the “wettability” of a surface. (From Rixom, M. R.,
and Mailvaganan, N.P., Chemical Admixtures for Concrete , E & F Spon, London,
UK, 1986.)

angle of approximately 120° based on that surfaces coated with waxes or
fatty acids would have this contact angle (Figure 4.2).
Using a contact angle of 120° and maximum capillary size of 500 nm,
Rixom and Mailvaganan (1986) calculate the theoretical pressure necessary
to penetrate the concrete would be 14 metres. In fact, wetting still does occur
due to defects in the hydrophobic film or larger voids within the matrix.
Research by Yiannos (1961) showed a monolayer of stearic acid on copper
had a contact angle of 104° and this would be a more conservative estimate.
The hydrophobic pore-blocking admixtures combine a hydrophobic material
with a polymeric material, which coalesces under pressure to form a plug.
The conflicting claims regarding different proprietary products makes it
difficult for specifiers and users to assess this group of admixtures. Figure 15.5
in ACI 212.3R highlights the problem. The coefficients of permeability for
the reference samples as reported in the three BBA certificates are 2.2 × 10 -12
m/s, 2.0 × 10 -13 m/s, and 4.3 × 10 -14 m/s indicating that the concretes were of
markedly different quality. Accordingly the relative values given in the figure
have little real meaning. The second permeability result given in the British
Board of Agreement (BBA) certificate for the hydrophobic pore-blocking
admixture, which shows an order of magnitude reduction, is not shown
in the figure or table. Clearly, data on the relative permeability of the same
concrete mixture containing different admixtures would be more helpful.
The comprehensive test program conducted by Roberts and Adderson
of the Building Research Establishment (BRE) in 1985 provides dramatic
evidence of the vast variability in short-term performance of concrete con-
taining different hydrophobic admixtures. Five out of the nine hydrophobic
admixtures tested (and the water reducing admixture) actually increased
water penetration compared with the control in 10 or more of the 15 com-
binations of test method and curing procedure. Only one admixture (a
hydrophobic pore-blocking ingredient) consistently reduced water penetra-
tion under all five test methods. Although two admixtures gave significant
reductions compared with the control in the absorption tests (namely, ISAT,