393
International RILEM Conference on Materials, Systems and Structures in Civil Engineering
Conference segment on Service Life of Cement-Based Materials and Structures
22-24 August 2016, Technical University of Denmark, Lyngby, Denmark
except for G4, the sample size and shape effects on the initial mass loss can be accurately
taken into consideration by multiplying the mass loss by the corresponding drying radius.
However, this is only true if similar concrete and surface conditions are applied to all
geometries.
4.2. Stage 2: main drying initiation
The second stage of drying consists in a linear increase of the mass loss as a function of the
square root of time. The slope of this increase is related both to the concrete properties
(porosity, permeability, diffusivity, …) and to the humidity gradient between the concrete and
the ambient air. As for the first stage, the effect of the sample size and shape can be accurately
taken into account by the drying radius. A higher drying radius induces a lower slope. By
multiplying the drying radius by the slope of the mass loss curve during this second stage, all
geometries present the same behaviour, as shown in Figure 4c and 4d. The end of this stage is
not clearly defined, and probably occurs at a point where the effective permeability is
significantly decreased in comparison with its initial value. This occurs when the overall
humidity has significantly decreased in the whole sample, which could occur close to a
specific mass loss. Further results should provide additional answers regarding this
observation.
4.3. Stage 3: advanced drying and permeability decrease
As stated previously, the third stage consists in a significant decrease of the mass loss,
resulting in a progressive decrease of the slope of the mass loss versus square root of time
curve. At a given time, this phenomenon is more clearly observed in Figure 6a for geometries
with lower drying radius. Indeed, G4 and G3 have already lost a significant amount of water,
in contrast with G1 or G2.
Figure 6: Whole mass loss curves (dots = experiments, dashed lines = simulations)
As for the first and second stage, the simulation captures this behaviour. However, as the
mass loss increases, an overestimation of the simulation can be observed, especially when it
overpasses 1%. This is due to the parameters expressed in equations 4, 5 and 8. These
equations relate the relationship between saturation degree, relative humidity, and effective
permeability and diffusivity. Indeed, all parameters from these equations were extracted from
394
International RILEM Conference on Materials, Systems and Structures in Civil Engineering
Conference segment on Service Life of Cement-Based Materials and Structures
22-24 August 2016, Technical University of Denmark, Lyngby, Denmark
another study [1]. Therefore, even if the same concrete was used, these equations should be
adapted to this study. In particular, the decrease of effective permeability when decreasing the
saturation degree should be more important. This point will be discussed further when longer
term experiments have been performed on all sample sizes and shapes.
5. Conclusion and perspectives
Mass loss experiments on various specimen size and shape are performed. The model
developed and used in [1] is improved by taking into account the
boundary layer of air at the
sample-air interface during drying. The curves of mass loss as a function the square root of
time can be divided in three stages identified in this study: a non-linear acceleration initial
stage, a linear drying initiation, and an advanced drying consisting in a decreased slope of the
curve. These three stages can be accurately reproduced by the model, independently on the
sample size and shape.
Further works are required in order to confirm the relevance of this model for longer term
experiments, and for identifying mechanisms at the origin of the switch between the second
stage and the thirst stage. Additional results regarding the relative humidity gradient inside the
specimens will allow a mode complete description of these phenomena, as well as an
improved strategy for identifying the parameters of the model. Finally, this validity of this
model should be challenged with drying/wetting cycles experiments.
References
[1] Hilaire A., Etude des déformations différées des bétons en compression et en traction, du
jeune au long terme : application aux enceintes de confinement, PhD Thesis, ENS Cachan,
2014
[2] Huang, Q.; Jiang, Z.; Gu, X.; Zhang, W. & Guo, B., Numerical simulation of moisture
transport in concrete based on a pore size distribution model, Cem Conc Res , 2015, 67, 31-43
[3] Briffaut, M., Etude de la fissuration au jeune âge des structures massives en béton :
influence de la vitesse de refroidissement, des reprises de bétonnage et des armatures, PhD
Thesis, ENS Cachan, 2010
[4] Mualem, Y. (1976). A new model for predicting the hydraulic conductivity of unsaturated
porous media. Water resources research, 12(3):513–522
[5] van Genuchten, M. A Closed-form Equation for Predicting the Hydraulic Conductivity of
Unsaturated Soils. Soil Science Society of America Journal, 44(5):892–898, 1980
[6] Millington, R., Quirk, J. (1961). Permeability of porous solids. Transactions of the
Faraday Society, 57:1200–1207
[7] Oxfall, M.; Johansson, P. & Hassanzadeh, M. Long-term hygrothermal performance of
nuclear reactor concrete containments – Laboratory evaluations of measurement setup, in situ
sampling, and moisture flux calculations, Cem and Conc Composites , 2016, 65, 128 – 138
[8] Zhang, J.; Wang, J. & Han, Y. Simulation of moisture field of concrete with pre-soaked
lightweight aggregate addition, Construction and Building Materials , 2015, 96, 599 – 614
, F. & Balabani , G. Modeling influence of hysteretic moisture
behavior on distribution of chlorides in concrete, Cement and Concrete Composites , 2016,
67, 73 – 84