Study of hydration products

Table 6 shows the principal hydrates determined by the XRD tests on the cement pastes without MSWI fly ash (P0-C1 and P0-C2) and with 20 % MSWI ash (P20-C1 and P20-C2). The results obtained for the two cements C1 and C2 are very close. The principle crystalline hydrates formed due to the addition of the MSWI ash are ettringite, Friedel's salt and thenardite. These results (in particular formation of ettringite and Friedel's salt) confirm a number of results obtained in other studies [12, 15]. The addition of the MSWI ash into the cement pastes results also in an increase in quartz and calcite (already present in the ash).

Figure 4 shows the results of the DSC tests obtained for cement paste P-C1 and for the cement paste containing 20 % MSWI fly ash (in place of cement) P-FA after 28 d (both without immersion).

Figure 5 shows the DSC curves obtained for pastes P-C1 and P-FA after successive immersion in ethanol, ethylene glycol and distilled water. We can see that the DSC curves of these two cement pastes (after immersion) are very similar.

The DSC tests carried out in the early stages (in particular after 1 d and 3 d) have shown that the MSWI ash has a strong influence on cement hydration. After 28 d hydration, the main feature observed is a movement of the peak corresponding to C-S-H towards negative temperatures for the P-FA paste. As already stated above, the successive immersion of the cement pastes in ethanol, ethylene glycol, and distilled water was carried out to replace the interstitial solution of the cement pastes by distilled water and to attempt to determine in this way the origin of the shift of the C-S-H peak (size of the material's pores or concentration of the interstitial solution). After the successive immersion, the two cement pastes both had interstitial solutions of a similar chemical composition (assuming that hydrate dissolution is sufficiently slow). The DSC curves for these cement pastes are practically identical. In particular, the temperature of the C-S-H peak is identical. We can therefore deduce from this result that the porous structures of the C-S-H in the two cement pastes (with or without the MSWI ash) are very similar. The differences observed between the DSC curves of the P-C 1 and P-FA pastes after 28 d of hydration must therefore be due essentially to a difference in the chemical composition of the interstitial solutions. A strong alkaline concentration in the interstitial solution of the P-FA paste could explain this result.

Figure 5: DSC curves obtained for the P-C1 and P-FA pastes after successive immersions