ICDS12 International Conference DURABLE STRUCTURES: from construction to rehabilitation LNEC Lisbon Portugal 31 May - 1 June 2012 DURABLE STRUCTURES

Transcription

1 Numerical tool for durability assessment of concrete Structures subjected to aggressive environment Jonathan MAI NHU 1 Alain SELLIER 2 Patrick ROUGEAU 1 Frédéric DUPRAT 2 Manuela SALTA 3 1 CERIB Centre d Études et de Recherches de l Industrie du Béton, Epernon, FRANCE 2 LMDC Laboratoire Matériaux et Durabilité des Constructions, Toulouse, FRANCE 3 LNEC Laboratorio Nacional de Engenharia Civil, Lisboa, PORTUGAL Table of content Fundamental role of water in the study of concrete durability Simulation of all the parameters (water, hydrates consumption, carbonation, chlorides, combination of carbonation chlorides) Some results of the simulation Perspectives LNEC

2 Carbonation Fundamental role of water transport in the study of durability mechanism The saturation degree is a fundamental parameter when considering the carbonation mechanism The carbonation rate is optimum for intermediate Relative Humidity (h) High enough to allow the Low enough to not slow CO 2 solubilisation down the CO 2 diffusion At low h, free water quantity is not enough Limiting factor : CO 2 3 Chlorides At high h, CO 2 diffusion is 10 4 times lower in liquid phase than in gaseous phase Fundamental role of water transport in the study of durability mechanism Transport mode of chlorides according to the saturation degree of the material Concrete saturated in liquid water Diffusion (concentration gradient) Tiding zone Capillary absorption (pressure gradient) Wetting/drying cycle Capillary absorption Convection LNEC

3 Experimental test Original new experimental durability test Simulation: simulated specimen LNEC

4 Basic hypothesis Three phases : solid phase (cement paste matrix), liquid phase (pore solution) and gaseous phase Solid matrix is not deformable Liquid phase is not compressible Gaseous phase comprising two ideal gas: Dry air (partial pressure p a ) Vapour water (partial pressure p v ) Gravity is not taking into account Water transport are isotherm (20 C) Governing equations Mass balance equation of the three constituents: liquid, solid and gas m i is the total mass of the constituent i per unit of material volume W i is the mass flux of the constituent i μ represents the rate of mass formation per unit of volume The div() operations represent the mass exchanges with the environment LNEC

5 Governing equations Expression of the mass flux of each constituent Expression of the total mass of each constituent m i represents the total mass of the constituent i per unit of material volume v i represents the velocity of the constituent i Ø represents the total connected porosity S r represents the liquid water saturation ρ i represents the density of the constituent i M j represents the molar mass j=voua Drying equation Water in liquid and gaseous phases is taken into account m l m v Darcy s flux of liquid water Diffusive flux of water vapour LNEC

6 Drying parameters Resistance to the diffusion: the relative permeabilities to both liquid and gaseous phases (1) et (2) (1) Van Genucheten MV, «A closed form equation for predicting the hydraulic conductivity of unsaturated soils», Soil Science of America Journal, vol. 44, pp (2) Mualem Y, «A new model for predicting the hydraulic conductivity of unsaturated porous media», Water Ressources Researchl, vol. 12,n 3, pp Drying parameters Resistance of the porous media to the gas diffusion Diffusion coefficient of the gas constituents Diffusion coefficient out of the porous media Resistance of the porous media to the gas diffusion (1) De Vries et Kruger, «On the value of diffusion coefficient of water vapour air», Phénomènes de transport avec changement de phase dans les milieux poreux ou colloïdaux, ed. CNRS, pp (1) Millington et al Tiery et al LNEC

7 Drying equation: application m l m v Main variable: relative humidity h Darcy s flux of the liquid water Convective flux Diffusive flux of the vapour water Diffusive flux Variation of water quantity in the carbonation mechanisms (dissolution of hydrates) CH 1 CSH 0 AFt 32 AFm 12 LNEC

8 Initial conditions and boundaries conditions Hydric cycles (variation of relative humidity h) h=50% Boundaries conditions 0, Days LNEC

9 Simulation of carbon dioxide transport Transport of carbon dioxide Millington et al Tiery et al Simulation of carbon dioxide transport Initial conditions and boundaries conditions CO 2 pressure (4 053 Pa) After one day of simulation 0Pa LNEC

10 Simulation of CH consumption CH consumption ( expressed as equivalent Ca) Hypothesis: reactions are delayed. This hypothesis allow the numerical resolution. The CH consumption depends on the CO 2 pressure reaction does not occur anymore when all CH is consumed Simulation of AFt consumption AFt consumption ( expressed as equivalent Ca) Initial conditions LNEC

11 Simulation of C S H consumption C S H consumption ( expressed as equivalent Ca) n=0,8 From Hyvert s thesis work At t time, the rate of hydrate consumption is not the same for all hydration products. Simulation of the calcite formation («sink of CO 2») Formation of calcite Hypothesis (for all reactions): The rates of convection mechanisms are lower than the rates of all the chemical reaction and all the reaction occur without any movement. LNEC

12 Simulation of the AFm consumtion AFm consumption (expressed as equivalent Ca) Initial conditions Carbonation Chemical fixation of chlorides Simulation of the AFm consumtion SF eq SF eq represents the total chlorides able to be chemically fixed (by formation of Friedel salts) for infinite time and for the AFm still available (that means not carbonated) 2 1 DEBY F., «Approche probabiliste de la durabilité des bétons en environnement marin». Thèse de doctorat. Spécialité Génie Civil. Université Toulouse III Paul Sabatier LNEC 1 WANG X. «Modélisation du transport multi espèces dans les matériaux cimentaires saturés ou non saturés et éventuellement carbonatés». Thèse de doctorat soutenue le 27 Avril 2012, Université de Paris Est, IFSTTAR. 1

13 Simulation of chemical fixation of chlorides SF Formation of Friedel salts When SF = SF eq the chemical fixation of chlorides and the formation of Friedel salts do not occur anymore. Simulation of physical fixation of chlorides YCSH Chlorides fixations on CSH 1&2 When YCSH = YCSH eq the chloride fixation on C S H does not take place anymore. YCSH eq represents the total chlorides able to be fixed after infinite time 1 HIRAO et al «Chloride binding of cement estimated by binding isotherms of hydrates». Journal of Advanced Concrete Technology, 3: DEBY F., «Approche probabiliste de la durabilité des bétons en environnement marin». Thèse de doctorat. Spécialité Génie Civil. Université Toulouse III Paul Sabatier LNEC

14 Chlorides transport Simulation of chlorides transport Physical fixation of chlorides Chemical fixation of chlorides Simulation of chlorides transport Influence of the saturation state on the chlorides diffusion 1 We consider that below 60 % of saturation the liquid phase is not continuous anymore and the diffusion is impossible By using this law on Francy s data, Nguyen 2 demonstrated that λ =6 1 FRANCY O., FRANÇOIS R. «Modélisation du transfert couplé ion chlore humidité dans les matériaux cimentaires». Revue française de génie civil, 5, 2 3, 2001, p LNEC 2 NGUYEN TQ, «Modélisation physico chimique de la pénétration des ions chlorures dans les matériaux cimentaires». Thèse de doctorat, Spécialité structures et matériaux, LCPC

15 Simulation parameters All the parameters input in the model Parameters Values Nature Ø Experimental T [K] 293 Environmental conditions of the test ρ l [kg/m 3 ] 10 3 Water density R [J/mol/K] 8.32 Constant of ideal gas M v [kg/mol] 18 x 10 3 Molecular mass of the water vapour p vsat [Pa] Saturated vapour pressure f(h) Sorption/desorption K l [m²] 4 x Experimental ηl [Pa.s] x 10 3 Dynamic viscosity k rl Ranaivomanana s thesis D 0 [m²/s] 2.48 x 10 5 Diffusion coefficient out of porous media Simulation results (CPU time of simulation: 20 minutes) LNEC

16 Simulation Results mass gain of water in samples Decalcification Perspectives LNEC

17 Perspectives Simulation of the oxygen consumption at the steel/concrete interface 0,25 mol/m 3 (hypothesis: the maximal concentration in the environment at the atmospheric pressure) Perspectives 0,25 mol/m 3 Simulation of the oxygen consumption at the steel/concrete interface The oxidation rate of iron is very fast compared to the rate of oxygen reduction. The corrosion rate is governed by the rate of reduction mechanisms and the oxygen diffusion. Drying mechanisms are taken into account. The oxygen diffusion is much faster in the air than in water. LNEC

18 Perspectives Simulation of the oxygen consumption at the steel/concrete interface in mol.m 2.s 1 The oxygen flux gives 4/3 of the iron flux according to the corrosion equation. The ratio M(Fe)/ρ(Fe) represents the quantity of corrosion products which is produced (for one mol of iron) to express the iron flux into corrosion rate in m.s 1 1 HUET B. «Comportement des armatures dans un béton carbonaté influence de la chimie de la solution interstitielle et d une barrière de transport». Thèse de doctorat, INSA de Lyon CHITTY W.J. «Etude d analogues archéologiques pour la prévision de corrosion pluriséculaire des armatures de béton armé : caractérisation, mécanismes et modélisation». Thèse de doctorat, Université de Technologie de Compiègne & 2 Perspectives Simulation of the oxygen consumption at the steel/concrete interface H : fonction which takes into account the depassivation conditions of the reinforcement Carbonation Decrease of ph 1 & 2 Destruction of the passive layer Chlorides LNEC

19 Thank you for your attention LNEC