H.S. Wong, N.R. Buenfeld

The mass ratio of water-to-cement (w/c) is the most important parameter when designing concrete mixtures because it controls the mechanical properties and durability of the hardened concrete. As such, it is desirable to have the ability to determine the original w/c ratio of a hardened concrete for compliance testing, quality assurance, forensic analysis and research.

However, once concrete sets and hardens, it is very difficult to find out the exact amounts of cement and water that were originally used. The two main existing methods are the physico-chemical method (BS1881:124:1988) and the fluorescence microscopy method (Nordisk NT361-1999).

The physico-chemical method is known to have low precision, estimated to be within 0.1 (w/c ratio) or greater, and thus has little practical value. The fluorescence method relies on the use of reference standards for comparison, which must have the same ingredients and proportions, and cured to the same hydration degree, in addition to w/c ratio, as the concrete being examined.

In this study, a new method to determine the original w/c of Portland cement concretes of unknown proportions was developed. The method is based on measuring the volumetric fractions of capillary pores, hydration products and unreacted cement using backscattered electron microscopy (Figure 1). The results are then used to calculate original cement content, free water content and free w/c ratio.

The method has the advantage that it is quantitative and does not require comparison with calibration graphs or reference samples made with the same materials and cured to the same degree of hydration as the unknown concrete. Another advantage of the method is that it can determine the degree of cement hydration independently.

We have tested the method on a range of concretes, mortars and cement pastes. The mix variables include w/c ratio (0.25-0.70), cement type, cement content (300-1750kg/m3), aggregate content (0-70% vol.) and curing age (3-90 days). Lab cured and field cured samples were tested. Figure 2 compares estimated values against the actual values for all samples. The error bars indicate 95% confidence interval.

The results show a good agreement between the measured and actual values. The percentage estimation errors for cement content, water content, w/c ratio and degree of hydration ranged from -3.2 to 10.2%, -2.3 to 5.8%, -8.6 to 8.4% and -11.3 to +7.2% respectively. The errors do not appear to be influenced by either the mix proportion or curing age.

It was observed that samples with high w/c ratios (≥ 0.5) tend to bleed after compaction and this can decrease the free w/c ratio significantly. It was also observed that aggregate absorption could have a significant effect on the free w/c ratio. The spread in local w/c ratio is higher in concretes compared to mortars and pastes due bleeding effects and presence of aggregates that increases the heterogeneity of the microstructure.

However, the results do converge when a representative number of images are analysed and averaged. Aggregate absorption and bleeding effects do not affect the proposed method because it is designed to measure free w/c ratio. Overall, a good agreement between the estimated and actual values was found. The largest error in the estimated free w/c ratio was found to be less than 0.025 for pastes, and less than 0.05 for mortars and concretes. Further research is ongoing to extend the technique to concretes containing supplementary cementitious materials.

References

• H.S. Wong, N.R. Buenfeld (2009), Determining the water-cement ratio, cement content, water content and degree of hydration of hardened cement paste: Method development and validation on paste samples, Cem. Concr. Res., 39, 957-965.
• H.S. Wong, K. Matter, N.R. Buenfeld (2013), Estimating the original cement content and water-cement ratio (w/c) of Portland cement concrete and mortar using backscattered electron microscopy, Mag. Concr. Res., 65 (11) 693-706.