Vous êtes sur la page 1sur 9

15th IACMAG www. 15iacmag.

org
19-23 October 2017, Wuhan, China

Validation of a Fully Coupled THM Finite Element Code: Simulation of


CIEMAT Mock-Up Test
Ayman A. Abeda* and Wojciech T. Soowskia
a
Aalto University, Department of Civil Engineering, P.O.Box 12100, FI-00076 Aalto, Finland
* ayman.abed@aalto.fi (corresponding authors E-mail)

Abstract

This contribution presents simulation of the Mock-Up test performed by CIEMAT (Madrid, Spain) [1].
The numerical model employs enhanced Barcelona Basic Model [2] coupled with water retention behaviour
and water transport as given in [3,4]. In addition, the used theoretical framework takes full consideration of
the thermal effects on the behaviour of bentonite. All the parameters in the simulation have been chosen
based on experimental data [3,5,6].
The calculation is performed in custom Finite Element code build at Aalto University [4]. The calculated
results are in good agreement with the experimental measurements validating the correctness of the
implementation of the code. This study also confirms that a realistic estimation of the thermo-hydro-
mechanical parameters is essential for a better prediction of the system behaviour in the physical reality.

Keywords: Finite Element Method, Unsaturated soil, nuclear waste repositories, THM coupling

1. Introduction

In the framework of THEBES project for modelling bentonite behaviour, Aalto University takes the role
of developing a new numerical code to simulate the fully coupled thermo-hydro-mechanical behaviour of
bentonite. The simulation of CIEMAT mock-up test is an excellent test for the validation of the code and
general modelling framework due to the well-defined boundary conditions, measured data and the very
good characterization of the selected bentonite. This experiment was also modelled by many research
groups [6,7] giving the opportunity to create clear idea about the expected results and possible modelling
problems. After a brief introduction on the governing equations, the physical test is described together with
the properties of the FEBEX bentonite. The finite element model and the boundary conditions are then
presented. Finally, the paper discusses the numerical results and compares them to the measurements.

2. Governing Equations and Solution Strategy

In this section, a summary of the adopted balance equations is given. The full derivation of these
equations is out of the scope of this contribution. Fully detailed description and derivation of the used
framework can be found in [4].
2.1 Mass balance equations
The soil is modelled as a porous medium consisting of the solid phase (s), the gas phase (g) and the
liquid phase (l). Each phase has its components: (i) the solid phase has the solid soil particles as the only
component, (ii) the gas phase is decomposed into the dry air (a) and the water vapour (w) and (iii) the liquid
phase consists of the liquid water (w) and the dissolved dry air (a). The compositional method [8] is used
to derive the mass balance equations for the soil components.
15th IACMAG www. 15iacmag.org
19-23 October 2017, Wuhan, China

Table 1. Summary of constitutive assumptions for fluid and thermal flow


Fluid flow Vapour diffusion
Darcys law: Philip and De Vries model [9]:

= ( + 1) =
+


Water retention curve (van Genuchten model [10]):
Note: and are diffusion coefficients that are dependent on soil
= (
)[1
+ ( ||) ] +

tortuosity , thermal enhancement factor , degree of saturation, soil


, : degree of saturation at full and residual state porosity and molecular diffusivity of water vapour in air .

Relative permeability [11]: Heat flow


3 Fouriers law:


=

( )

=
+ = 1
=
1
= ( / ) ; = ; = 1
1 ( )

= + ( ); = 0.118 1.54 104 , and : average, saturated and dry thermal conductivities
, and : water pressure head, gas pressure head and suction head
and : permeability and degree of saturation of fluid = , = .
, , , , and

: van Genuchten parameters with thermal : water surface tension.
effects [4,12].

Accordingly, the general mass balance equation for any one component is written as:

( )
( ) +
+
=


(1)
/

where the symbols and denote volume fraction and density of the phase ( = , , ).
The mass fraction of component in the -phase is expressed by = . The phase-velocity vector
is represented by whereas is the non-advective flux vector of the component . The sink or source

term of the component obeys the constraint + + = 0 in absence of any external
sink/source. Based on Equation (1) the following balance equation for each component can be derived:

1- Mass balance of the water component




(1 )(

+ ) + (

+ )
+ ( ) +
(2)

)
+ + ( + ( ) + = 0

2- Mass Balance of the dry air component


[ + ]
[ + ] + + [ + ] + [ 1] +
1 (3)

+ ( ) + ( ) = 0

where and denote liquid and gas degree of saturation, respectively. The symbols , , , and
stand for temperature, porosity, volumetric strain, coefficient of solid thermal expansion and Henrys
volumetric coefficient of solubility, respectively. The flow velocity of the liquid phase and the gas phase
are assumed to obey Darcys law [13]. Furthermore, the above equations show that constitutive
15th IACMAG www. 15iacmag.org
19-23 October 2017, Wuhan, China


assumptions are required to quantify
,
, and . The used constitutive assumptions are
summarized in Table 1.

2.2 Energy balance equation

The enthalpy balance equation is given as follows [14-17]:



+
+ =0
(4)

where is the soil heat capacity and is the heat flux including conduction and convection. Symbol

represents the latent heat of water vaporization and is the rate of water vapour production which can
be derived based on Equation (1). The constitutive assumptions related to the heat flow and the soil thermal
conductivity are given in Table 1. The soil heat capacity is derived by adopting a suitable thermodynamic
state function that requires the specific heat capacities of each bentonite components to be used.

2.3 Mechanical Balance Equation

The local static mechanical balance equation is given as:

+ = 0 (5)

where and stands for the total stresses and the body forces at a certain point of the domain,
respectively. The body forces are usually due to the self-weight which is related to the density. To calculate
the developed stresses as a consequence of the applied loading increment (thermal, hydraulic or mechanical
load), a suitable constitutive model for unsaturated soil is needed. In this study, an extended version of the
Barcelona Basic Model (BBM) that accounts for thermal effects on mechanical behaviour [2,18,19] is
employed. The full description of the model and the numerical implementation is available elsewhere [20-
25]. The basic ingredients in BBM framework are listed in Appendix A.

Fig. 1 Mock-Up test details (modified after [1])


15th IACMAG www. 15iacmag.org
19-23 October 2017, Wuhan, China

2.4 Solution strategy

The finite element method is employed to discretise the balance Equations (2), (3), (4) and (5) while a
fully implicit finite differences scheme is used for the time discretization. The resulted coupled system of
nonlinear algebraic equations is solved using Newton-Raphson iterations.
This formulation is implemented into a computer code [4] using NUMERRIN numerical solver [26]. In
the remainder of this paper, the implemented code is referred to as Aalto Code. The following section
illustrates numerical results of a validation example that are generated using Aalto Code.
3. CIEMAT experiment

Within the framework of FEBEX research project [27], CIEMAT (Madrid, Spain) performed a full scale
Mock-Up test with controlled boundary conditions that resembles the real in-situ situation of nuclear waste
barrier [1]. As can be seen in Fig. 1, the test used two electrical heaters to simulate the temperature produced
by the nuclear waste canisters. Those were surrounded by a clay barrier made from compacted FEBEX
bentonite blocks. The test was fully instrumented with automatic control of the heater temperature, sensors
to measure the evolution of the main variables in the system (temperature, relative humidity, stresses etc.)
and a data acquisition system to record and process the collected data by the sensors. Due to symmetry only
the indicated part in Fig. 1 is numerically modelled in axisymmetric conditions.
The bentonite blocks have an average initial water content of 14% which corresponds to an initial suction
of about 120MPa. The initial dry density is 1770.0 Kg/m3 and the total mass of bentonite used in the test is
22.5t. The dry density would decrease after saturation to an average targeted value of 1650 Kg/m3 after the
closure of the gaps between the blocks due to bentonite swelling.

Fig. 2 Finite element model of the Mock-Up test: (a) mesh and dimensions; (b) mechanical boundary
conditions; (c) hydraulic boundary conditions; (d) thermal boundary conditions.
15th IACMAG www. 15iacmag.org
19-23 October 2017, Wuhan, China

After an initial saturation phase, the heaters are put to work in three stages: (i) in the first 6 days they
are subjected to an electrical power of 250W per heater (ii) in the following 4 days the power is increased
to 500W/heater (iii) after these initial phases, the temperature is automatically controlled as such it stays at
100oC throughout the test operational life. A special system (see Fig. 1) is employed to ensure continuous
hydration of the bentonite through the outer boundaries by applying water under 0.55MPa pressure. Due to
limited data, the numerical simulation is only performed for the first 2500 days of the test duration.

4. Finite Element Model

Fig. 2 shows the finite element model as used in this validation example. The mesh is first subjected to
a convergence study where an enough accurate mesh is found to consist of 1396 4-noded quadrilateral
elements with four stress integration points per element, see Fig. 2 (a). To simulate a constant volume
conditions the model is constrained in the directions normal to its boundaries as it clear in Fig. 2(b). The
hydraulic and thermal boundary conditions are shown in Fig. 2(c) and (d). The thermal load is applied as
in the real test where the boundary near the heater is first subjected to total of 250W followed by 500W of
energy flux during the first ten days. After that, the thermal boundary condition is changed to the prescribed
temperature of 100oC. The outer boundary of the model is kept at the prescribed temperature of 24oC which
simulates the recorded average room temperature. As no gas pressure measurements are provided, the gas
pressure is assumed to stay atmospheric everywhere in the solved domain. An initial stress of 10kPa is
assigned to the bentonite which simulates the average bentonite self-weight in the middle container.

4.1 FEBEX bentonite properties

The FEBEX bentonite is exposed to an intensive thermal, hydraulic, mechanical and chemical testing
to characterize its relevant properties [3,5,6]. These data are used to calibrate and derive the required
modelling parameters which are listed in Table 2.

Table 2. FEBEX bentonite properties as used in the analysis (for a better understanding of mechanical
parameters see Appendix A)

(a) Mechanical properties


[1/K] 2 1 2
0.4 0.05 0.3 0.35 1.5E-4 0.25 0.0 -3.0E-6 -0.147 0.0

[kPa]
[1/kPa] [kPa] [kPa]
1.0 0.1 12.0 0.15 1.0E-4 0.925 500.0 0.2 1.2E+4

(b) Hydraulic properties


[m/s]
[1/m] [1/K] [1/K]
3.5E-04 1.22 0.0 1.0 -1.0E-4 -1.5E-3 1.9E-14

(c) Thermal properties


[W/m/K] [W/m/K]
1.15 0.47 0.87 1.0

5. Discussion of numerical predictions

The numerical results at the control points 1, 2, 3 and 4 in Fig. 2(a) are depicted in Fig. 3. These numerical
results are compared to the measured data given by [1]. In Fig. 3(a) the code predictions are in a good
agreement with the relative humidity measurements, especially during the first 500 days. After that the
15th IACMAG www. 15iacmag.org
19-23 October 2017, Wuhan, China

numerical predictions deviate from measurements showing faster hydration. In contrary, the code
predictions are in very good agreement with the temperature measurements as can be seen in Fig. 3(b). The
same comment applies for the predicted water intake as shown in Fig. 3(d). The swelling pressure
predictions in Fig. 3(c) are the less satisfactory, even though the swelling pressure value at 2500 days is
predicted fairly well, the code predicts lower values during the initial stages of the test. That can be partially
related to micro-structure effect upon hydration which is not yet taken into account in the adopted hydro-
mechanical framework.

6. Conclusions

This article illustrates the performance of a newly developed fully coupled THM code through the
simulation of the well-documented CIEMAT Mock-Up test. The numerical results show that Aalto Code is
capable of providing reliable predictions for the bentonite behaviour under complex environmental
conditions as in the case of nuclear waste barriers. However, the adopted framework is still needs
improvement to consider properly the role and evolution of bentonite micro-structure. That is the aim of
the ongoing research, which will improve the constitutive relationships for mechanical behaviour of
bentonite, water and heat transport. On top of that, the framework should also incorporate the effects of
saline water on the behaviour of swelling clay barriers.

(a) (b)

(c) (d)

Fig. 3 Numerical results versus measurements: (a) relative humidity; (b) temperature; (c) swelling
pressure; (d) water intake.
15th IACMAG www. 15iacmag.org
19-23 October 2017, Wuhan, China

Acknowledgements

The authors would like to gratefully acknowledge that the presented research has been funded by
KYT2018 Finnish Research Programme on Nuclear Waste Management via THEBES project.

References

[1] Martin P, Barcala J, Huertas F. Large-scale and long-term coupled thermo-hydro-mechanic experiments
with bentonite: the FEBEX mock-up test. Journal of Iberian geology 2006; 32: 259-282.
[2] Alonso EE, Gens A, Josa A. A constitutive model for partially saturated soils. Gotechnique 1990; 40:
405-430.
[3] Gens A, Snchez M, Guimaraes LDN, Alonso E, Lloret A, Olivella S, Villar M, Huertas F. A full-scale
in situ heating test for high-level nuclear waste disposal: observations, analysis and interpretation.
Gotechnique 2009; 59: 377.
[4] Abed AA, Soowski WT. A study on how to couple thermo-hydro-mechanical behaviour of unsaturated
soils: Physical equations, numerical implementation and examples." Computers and Geotechnics 92
(2017): 132-155.
[5] Villar M, Gmez-Espina R. Report on thermo-hydro-mechanical laboratory tests performed by
CIEMAT on FEBEX bentonite 2004-2008. Informes Tcnicos CIEMAT 2009; 1178.
[6] Snchez M, Gens A. Modelling and interpretation of the FEBEX mock up test and of the long-term
THM tests 2014; DELIVERABLE (D-N: D3.3-3).
[7] CODE-BRIGHT U. a 3D program for thermo-hydro-mechanical analysis in geological media, users
guide. UPC,[S.l.] 2002;.
[8] Panday S, Corapcioglu MY. Reservoir transport equations by compositional approach. Transport in
Porous Media 1989; 4: 369-393.
[9] Philip J, De Vries D. Moisture movement in porous materials under temperature gradients. Eos,
Transactions American Geophysical Union 1957; 38: 222-232.
[10] Van Genuchten MT. A closed-form equation for predicting the hydraulic conductivity of unsaturated
soils. Soil Science Society of America Journal 1980; 44: 892-898.
[11] Brgesson L, Hernelind J. Preliminary modelling of the water saturation phase of the buffer and
backfill materials. International Progress Report IPR-00-11 1999;.
[12] Jacinto AC, Villar MV, Gmez-Espina R, Ledesma A. Adaptation of the van Genuchten expression to
the effects of temperature and density for compacted bentonites. Applied Clay Science 2009; 42: 575-582.
[13] Darcy H. Les fontaines publiques de la ville de Dijon: exposition et application... Victor Dalmont,
1856.
[14] Rutqvist J, Brgesson L, Chijimatsu M, Kobayashi A, Jing L, Nguyen T, Noorishad J, Tsang C.
Thermohydromechanics of partially saturated geological media: governing equations and formulation of
four finite element models. International Journal of Rock Mechanics and Mining Sciences 2001; 38: 105-
127.
[15] Thomas H, He Y. A coupled heatmoisture transfer theory for deformable unsaturated soil and its
algorithmic implementation. International Journal for Numerical Methods in Engineering 1997; 40: 3421-
3441.
[16] Collin F, Li X, Radu J, Charlier R. Thermo-hydro-mechanical coupling in clay barriers. Engineering
Geology 2002; 64: 179-193.
[17] Wang W, Rutqvist J, Grke U, Birkholzer JT, Kolditz O. Non-isothermal flow in low permeable porous
media: a comparison of Richards and two-phase flow approaches. Environmental Earth Sciences 2011;
62: 1197-1207.
[18] Laloui L, Cekerevac C. Thermo-plasticity of clays: an isotropic yield mechanism. Computers and
Geotechnics 2003; 30: 649-660.
[19] Gens A. Constitutive laws. Springer, 1995.
15th IACMAG www. 15iacmag.org
19-23 October 2017, Wuhan, China

[20] Abed AA, Laitinen M, Lms J, Harjupatana T, Soowski WT, Kataja M. Hydro-mechanical modelling
of MX-80 bentonite: one dimensional study 2016; 9: 18005.
[21] Abed AA. Numerical modeling of expansive soil behavior 2008;.
[22] Soowski WT, Hofmann M, Hofstetter G, Sheng D, Sloan SW. A comparative study of stress
integration methods for the Barcelona Basic Model. Computers and Geotechnics 2012; 44: 22-33.
[23] Soowski W, Sloan S. Equivalent stress approach in creation of elastoplastic constitutive models for
unsaturated soils. International Journal of Geomechanics 2013; 15: 04014041.
[24] Soowski WT, Gallipoli D. Explicit stress integration with error control for the Barcelona Basic Model:
Part I: Algorithms formulations. Computers and Geotechnics 2010; 37: 59-67.
[25] Abed AA, Vermeer PA. Numerical simulation of unsaturated soil behaviour. International Journal of
Computer Applications in Technology 2009; 34: 2-12.
[26] Laitinen M. Numerrin 4.0 Manual. Numerola Oy: Jyvskyl, 2013.
[27] Huertas F, Fuentes-Cantillana J, Jullien F, Rivas P, Linares J, Faria P, Ghoreychi M, Jockwer N,
Kickmaier W, Martnez M. Full-scale engineered barriers experiment for a deep geological repository for
high-level radioactive waste in crystalline host rock(FEBEX project). EUR(Luxembourg) 2000;.

Appendix A

Basic components in Barcelona Basic model is summarized in what follows [2,18,19]

The net stress rate is determined as


( +2 )
= e e + e e
3(+ ) 3

where e is the elasticity matrix which is dependent on the soil swelling index , isotropic net pressure
= tr()/3, the specific volume and Poissons ratio . The symbols , , , and stands for
total strain rate, unity vector, suction, atmospheric pressure and a plastic multiplier, respectively. The
parameters , and 2 are the suction swelling index and thermal elastic expansion parameters,
respectively. The temperature increment is denoted by . In order to account for the swelling nature of
some soils, the elastic stiffness is made pressure dependent as follows:
= (1 + )

2
= (1 + 1 ( ))

where and are reference values of and , respectively. Extra material parameters , 1 , 2

and are needed in this case.

The yield surface F is described by

= 2 2 ( + )( ) = 0

where is the slope of critical state line and stands for the deviatoric stress:
15th IACMAG www. 15iacmag.org
19-23 October 2017, Wuhan, China

1
= (1 2 )2 + (2 3 )2 + (3 1 )2 ,
2

where 1 , 2 and 3 are the principal stresses. The plastic strain direction is determined using a plastic
potential function Q:
= 2 2 ( + )( ) = 0

The factor which allows for the recovery of the Jakys approximation of coefficient of at rest soil pressure
0 in 1D compression, is defined as
( 9)( 3)
= ,
9(6 )

where is the slope of normal consolidation line. Extended thermal BBM assumes, that the soil shear
strength is affected by suction and temperature as follows

= ,

where and are material constants. The soil preconsolidation pressure is also considered to be suction
and temperature dependent through the equation



= ( )

with
= [(1 ) + ]

and



= (1 log ( ))

where the preconsolidation pressure at full saturation and at the reference temperature = 20 oC is
indicated by . The parameters , , r, are material constants that have to be experimentally
determined. Finally, the preconsolidation pressure is updated as:



= ,

where is the initial preconsolidation pressure and = tr( ) is the plastic volumetric strain rate.

View publication stats

Vous aimerez peut-être aussi