Page:4SIGHT manual- a computer program for modelling degradation of underground low level waste concrete vaults (IA 4sightmanualcomp5612snyd).pdf/53

9.3Joint Failure

The failure criteria for joints is based upon the service life of the joint compound. Since joint widths will be centimeters wide, as previously stated, upon joint failure the permeability of the roof slab will be overwhelmed by the flow through the joints. Therefore, the calculations terminate upon when the internal time variable has reached the limit of the joint service life.

10Calculated Material Properties

All of the concrete physical parameters (e.g., diffusivity, permeability, etc.) are user specified inputs to 4SIGHT. However, in cases where not all properties are available, missing quantities must be approximated using existing correlations. The physical properties must be established due to both hydration and leaching. The physical properties due to hydration are the initial conditions. However, as the porosity changes due to leaching, corrected values of the physical parameters are needed.

10.1Hydration

The hydration of cement can be approximated by a reaction between tri-calcium silicate (C3S) and water, forming a calcium silicate hydrate (CSH). A more elaborate model incorporating multiple mineral phases would require chemical analysis of the cement and yield relatively little additional information concerning degree of hydration. The weight ratio of water to cement, ${\frac {w}{c}}$, is the oft reported quantity to characterize the concrete mix. After some period of hydration, the fraction of the initial C3S which has hydrated is the degree of hydration, $\alpha$. The relation between these two properties and porosity can be determined stoichiometrically [7],

$\phi =1-{\frac {1+1.31\alpha }{1+3.2{\frac {w}{c}}}}$

(25)

and is valid for ${\frac {w}{c}}$ values used in practice.

The diffusivity can be related to either ${\frac {w}{c}}$ or $\phi$. After the first 100 days of hydration the transport properties of most cement pastes are near their asymptotic values. Although the values are still changing after 100 days, these changes are small compared to the accuracy with which these transport measurements can be made. Due to this apparently steady state, an empirical relation between $D_{o}$, the chloride diffusivity in $m^{2}/s$, and ${\frac {w}{c}}$ was developed for cement paste by Atkinson, Nickerson, and Valentine[8] and Walton, Plansky, and Smith[9]

$\log _{10}D_{o}=6.0{\frac {w}{c}}-9.84$

(26)

for values of ${\frac {w}{c}}$ in the range (0.2-0.6).

21