Abstract
This thesis addresses the problem of better characterising pore-water interactions in cement at the nano-scale and of providing a macroscopic in-situ measurement of water distribution in concrete infrastructure. With regard to pore water interactions in cement three topics are addressed. First it is shown that the exchange observed in 1H Nuclear Magnetic resonance (NMR) 2D T2-T2 exchange experiment by Monteilhet et al. [Physical Review E, 2006. 74(6)] is due to water exchanging between pores of different size and not alternate Fe impurities density as had been suggested by subsequent discussion [Nanocem Open Meeting, April 2007]. Second, it is shown that the amplitude of the single quantum 1H NMR signal as a function of drying can be interpreted in term of a pore size and geometry that is consistent with other estimates. Two pore reservoirs are identified corresponding to sheet and gel pores. The pore thicknesses are founded to be 1.5 and 4.1 nm respectively. Third, Double Quantum Filtered 1H NMR is shown as a new method of study anisotropic water in cement. The theoretical spectrum of water in the proximity of a Fe3+ ion has been calculated. This has been used to provide an estimate of a characteristic distance that can be interpreted either as characteristic size of the nanoparticles constituting the C-S-H or the Fe3+-Fe3+ inter ion distance. The best estimate is 11 nm. With regard to the in-situ measurements, the characteristic signal to noise ratio and the portability of the Surface GARField magnet [Journal of Magnetic Resonance, 2007. 185(1): p. 1-11] has been greatly improved. The first measurements outside the laboratory have been successfully performed and are presented in this thesis.