diapositive 1 sym… · title: diapositive 1 author: fcu created date: 2/24/2020 1:46:32 pm
TRANSCRIPT
Page 15th International Itasca Symposium
Accounting for long term effects for the structural design of deep tunnels in
claystones
Victor Augusto RIBEIRO LIMA, Jean-François BRUCHON and Sébastien BURLON
Setec Terrasol
Page 2
1. Introduction
2. Accounting for long term effects
3. Calibration with measurements in terms of displacements
4. Comparison in terms of structural forces
5. Conclusion
Summary
Page 3
Introduction
- Complex behaviour of claystones or shales: multiscale material, thermo-hydro-mechanical couplings,brittle/ductile, anisotropy, short and long term behaviour …
- Large bibliography with complex models: hard to choice and certainly to use
Page 4
Introduction
- Complex behaviour of claystones or shales: multiscale material, thermo-hydro-mechanical couplings,brittle/ductile, anisotropy, short and long term behaviour …
- Large bibliography with complex models: hard to choice and certainly to use
Seyedi et al. (2017)
Different models tested
In situ measurements
Page 5
Introduction
From an engineer point of view, where structures have to be designed:
- “simple” models are preferred but able to catch evolution of forces in structure
- especially when long term effects play an important part in final loadings = extrapolation
Page 6
Introduction
Example of “simple” model in the context of the Meuse/Haute-Marne laboratory - Saitta et al. (2017) :
- Mohr-Coulomb’s criterion and Norton’s law
- 2 domains define from in-situ observations (intact and disturbed zones) -> anisotropy forced
- In consistency with principal results in terms of displacement of the
soil in a quasi-isotropic stress state and with isotropic laws
Saitta et al. (2017)
Saitta et al. (2017)
Intact zone
Fractured zone
Page 7
Accounting for long term effects
Norton’s law, initially developed for steels at high temperatures (1929)
- visco-plastic strain rate:
ሶ𝜖𝑣𝑝 depends only of deviatoric stress q
independent of mean stress p
ሶ𝜖𝑖𝑗𝑣𝑝=
3
2( ሶ𝜖𝑣𝑝)
sij
𝑞where ሶ𝜖𝑣𝑝 = 𝛾 < 𝑞 − 𝑞0 >
𝑛
Page 8
Accounting for long term effects
Modified Norton’s law
- ሶ𝜖𝑣𝑝 depends on the strength mobilization (p,q and Lode angle θ):
with
𝑓 𝑋 =1
1+𝑎
𝑋
𝛼 𝑓 𝑋 ∈ 0,1
𝑋 =𝑑1
𝑑2𝑋 ∈ 0,+∞
ሶ𝜖𝑖𝑗𝑣𝑝=
3
2( ሶ𝜖𝑣𝑝)
sij
𝑞where ሶ𝜖𝑣𝑝 = 𝐴 × 𝑓 𝑋
Page 9
Accounting for long term effects
Modified Norton’s law
- ሶ𝜖𝑣𝑝 depends on the strength mobilization (p,q and Lode angle θ):
with
𝑓 𝑋 =1
1+𝑎
𝑋
𝛼 𝑓 𝑋 ∈ 0,1
𝑋 =𝑑1
𝑑2𝑋 ∈ 0,+∞
ሶ𝜖𝑖𝑗𝑣𝑝=
3
2( ሶ𝜖𝑣𝑝)
sij
𝑞where ሶ𝜖𝑣𝑝 = 𝐴 × 𝑓 𝑋
cos 3𝜃 =3 3
2×
𝐽3
𝐽2Τ3 2
𝑑1 𝑞 =< 𝑞 − 𝑞0 >
𝑑2 𝑝, 𝑞, 𝜃 =𝑀 𝜃 × 𝑝 + 𝑞 − 𝑁 𝜃
𝑀 𝜃 2 + 1 2
Page 10
Accounting for long term effects
Modified Norton’s law
- q= 𝑞0 : 𝑋 =𝑑1
𝑑2= 0→ 𝑓 𝑋 = 0 → ሶ𝜖𝑣𝑝 = 0
- q= 𝑞𝑚𝑎𝑥 : 𝑋 =𝑑1
𝑑2= ∞→ 𝑓 𝑋 = 1 → ሶ𝜖𝑣𝑝 = 𝐴
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Titr
e d
e l'
axe
q (MPa)
Calcul du paramètre (d1^n) / (d1^n + d2^n) p = 13.8 MPa
α = 2 α = 3 α = 4 α = 8 α = 30
𝑓 𝑋 =1
1+1𝑋
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Titr
e d
e l'
axe
q (MPa)
Calcul du paramètre (d1^n) / (d1^n + d2^n) p = 13.8 MPa
a=1 a=2 a=3 a=4 a=5
𝑓 𝑋 =1
1+ 𝑋
2
θ = π/3, p = 13.8 MPa, q0 = 0, c = 6 MPa and ϕ = 20
≠ α ≠ a
Page 11
Calibration with measurements in terms of displacements
Application to the Meuse/Haute-Marne laboratory
Parameters: 0
40
80
120
0 30 60
Clo
sure
(mm
)
Time (day)
GRD4Horizontal closure
Vertical closure
Semi-empirical law
𝜖ሶ𝑣𝑝 = 𝐴 × < 𝑞 − >𝑛 n [-] A [ 1 . ]
[Pa]
GCS – GRD4 6.8 2. × 10 𝑞0 = 3 𝐽2,0 (initial invariant deviatoric stress)
ሶ = ×
+
Τ
A [ ] a [-] [-]
GCS – GRD4 × 10 11 2 3
Page 12
Comparison in terms of structural forces
Forces in rigid structure
- Normal force : quite similar
- Bending moment : quite different and more logical with in-situ observations (to be confirm)
- Bending moment more sensitive with variations of normal and shear stresses.
-12
-8
-4
0
-0.4 -0.2 0 0.2 0.4
No
rma
l fo
rce
(MN
/m)
Bending moment (MN.m/m)
Norton's law
New creep law
Page 13
Conclusion
- Models (simple and complex) have to be compared to measurements in terms of displacements andforces (only displacement not sufficient for structure design)
- Modified Norton’s law proposed here:
- is still a simple law based on a simple approach
- seems more reliable
- Others comparisons have to be done to confirm these results
- Test new idea is quite simple with Flac thanks to fish functions !
Acknowledgements
Andra for experimental data used in this study