Concretos De Alta Resistencia

HOW THE WATER/CEMENT RATIO AFFECTS
THE STRENGTH OF CONCRETE

by
Pierre-Claude AÏTCIN
and
Adam M. NEVILLE

(submitted to Materials and Structures)


INTRODUCTION
This is not just another paper on the water-cement ratio (w/c) : it goes beyond the well-known relation between w/c and strength, and tries to get behind that relation. In other words, our intention is to try to find outwhether the influence of the w/c on strength arises simply from the fact that a part of the mixing water produces voids that weaken the mass of hydrated cement paste, or what are the other, possibly more fundamental, factors involved. To resolve the problem we have to consider not only the hydrated cement paste but also mortar or concrete containing this cement paste. In this connection, it is usefulto look into the relation between the strength of mortar and concrete, both containing the same cement paste. To start with what is well-known, we shall first consider the early studies on the w/c.

RELATION BETWEEN STRENGTH AND W/C
For the sake of brevity, the term "strength" will be used to mean "compressive strength". Probably the first formulation of the relation between strength and thenon-solid ingredients of concrete was made by Féret in France in 1892 [l]. He understood the fact that the presence of water- and air-filled space in mortar had a negative influence on strength, and, from his experimental work, established a power relation:
[1]
where c, w, and a represent the content of, respectively, cement, water, and entrapped air by volume; and is the compressivestrength of mortar determined on test specimens of specific shape and size at a specific age. Féret found that a power of 2 gave the best fit and determined the value of the coefficient k by experiment.
In order to introduce the commonly, indeed universally, used term w/c, we can rewrite Eq. [1] as follows:
[2]
We should repeat that we are deallng with volumetric proportions, which is the onlyscientifically sound approach. The volumetric proportions of the relevant ingredients in 1
cubic metre of mortar (or concrete) are typlcally: c = 0.1, a = 0.01 to 0.02, and w = 0.2. Thus, w/c = 2, and a/c = 0.1. We can, therefore, neglect the term a/c, and re-write Eq. [2] as
[3]
Given that concrete is batched not by volume but by mass (possibly with the exception of water) and thatengineers usually work in terms of mass proportions, we can write Eq. [3] as
[4]
where 3.15 is a typical value of the specific gravity of Portland cement.
A relation between strength and the w/c was developed, probably independently, by Abrams in the United States [2]:
[5]
where w/c is expressed by volume, and A and B are constants dependent on the specific conditions such as the cementused, curing, and age at test.
The work of Féret and of Abrams represents a significant contribution to the understanding and, above all, use of concrete, and for that they deserve considerable credit. However, the cements they used are very different from modern cements and, even more importantly, modern concretes usually contain additional ingredients. Their cements were much more coarselyground than modern cements, and their chemical composition was different, too. Moreover, they used no plasticizing admixtures, indeed no admixtures at all. Thus, the initial reactivity of their cement was low, so that the workability of concrete was almost solely governed by the water content in the mix. As always, the concrete used both by Féret and by Abrams had to have an adequate workability toachieve reasonably full compaction. Because cement was an expensive ingredient, the cement content in the mixes used was relatively low, so that a high water content entailed a high w/c. Typical values used by Féret and by Abrams were between 0.6 and 1.0 by mass. A visual appreciation of the difference in the spacing of cement particles at a w/c of 0.65 and of 0.25 can be obtained from Fig. 1....