Samples for SEM examination were taken out and broken into small pieces and immersed in ethanol to stop hydration. After the pore water had totally exchanged with the ethanol, the small samples were immersed in an acrylic epoxy resin for two weeks so that the resin could exchange with the ethanol. After two weeks, the resin was cured at 70 ºC. The sample was then polished and carbon coated for SEM point-counting analysis.
The point-counting procedure was modified from ASTM C 1356M. Basically, a grid was superimposed on top of the SEM image and the phases determined and counted at each grid point by an expert human operator. The detailed procedure was as follows.
The SEM magnification was set at 1000x to 1500x, depending on the size of the particles in the paste. The purpose of using different magnifications was to avoid having adjacent grid points fall on the same particle. A 25-point grid was selected using the SEM configuration setup. A simple counting device was used to record the numbers of points that fell into the following two phases in the plain portland cement pastes: (a) unhydrated cement particles and (b) all others. Three phases were recorded in blended cement pastes: (a) unhydrated cement particles, (b) unreacted fly ash or slag particles, and (c) all others. The overall number of points counted for each sample was 3000, which was chosen to be able to obtain reasonable uncertainties of the estimates. The more points counted, the smaller the uncertainties become. The phase at each grid point was identified and counted. If the phase was obscured by the area formed by the grid intersection point, then the phase on the upper right corner of the grid point was recorded as the identified phase.
Assuming that the sample had a homogeneous microstructure, the grid was moved across the specimen. There were 120 fields to be viewed in each specimen in order to get a total of 3000 points counted (3000 = 120 x 25). The total area of the field of view was 3.60 mm2 (higher magnification) to 8.55 mm2 (lower magnification). After recording all the 25 points on one field, the field of view was moved systematically so as to examine a large sample field.
Figure 3 shows typical fields of view that were encountered during the point-counting procedure. Figure 3 (a) shows a plain portland cement paste. At some of the grid points, a number "1" has been marked to show that this grid point has been identified as a cement clinker particle. The grid points that have not been marked have been identified as hydration products and voids, which are counted as "all others" in this technique. Figures 3 (b) and (c) show a fly ash blended cement paste and a slag blended cement paste, respectively. In both microstructures, the gray levels of some fly ash particles and some slag particles were very close to those of hydration products, while the gray levels of others were very close to cement clinker particles. To help distinguish between these phases, the examiner can use particle morphology and, if necessary, use the x-ray signal to show the elemental abundance at a questionable point. Using these indicators, as well as gray level, one can readily discriminate among the fly ash particles, slag particles, hydration products, and cement clinker particles at each grid point. In Fig. 3 (b), there is one fly ash particle falling at grid points that is marked "2", two cement clinker particles that are marked "1", and the unmarked intersections are counted as "all others." In Fig.3 (c), there are five cement clinker particles marked "1", one slag particle marked "3", and the unmarked grid points have been counted as "all others."
| (a) T-1D w/c=0.4, 28d, X1000 | (b) T-L w/s=0.4, 7d, X1000 | (c) T-S w/s=0.4, 28d, X1000 |
Figure 3. Typical fields of view used for point counting on plain cement paste and blended cement pastes.