Date post: | 31-Jul-2015 |
Category: |
Software |
Upload: | deltaressoftwaredagen |
View: | 35 times |
Download: | 0 times |
Earthquake resistant design of levees and adjacent sheet piling
Marcel Visschedijk
Proven earthquakes damage to levees
24-Jun-15 2
Especially if sand in or under the levee liquefies
All pictures from: Y. Sasaki et al. / Soils and Foundations 52 (2012) 1016–1032
Damage caused by liquefaction
24-Jun-15 3
Damage caused by liquefaction
Kobe, picture from http://geot.civil.metro-u.ac.jp/archives/eq/95kobe/index.html
24-Jun-15 4
24-Jun-15 5
Proven earthquakes damage to levees
Slide from: Presentation of Ikuo Towhata, at the Second International Conference on Performance‐Based design in Earthquake Geotechnical Engineering, Taormina, Italy on May 29, 2012
Proven earthquake damage in the Netherlands
24-Jun-15 6
Roermond, 1992 (magnitude 5.8)
All pictures from: beeldbank.rws.nl
Groningen levees
Questions for the 2013 study*: • Show the current state of the primary and
regional defences • Give an indication of the required
improvement, differentiating between the situation without and with earthquakes
24-Jun-15 7
Continued studies in 2014/2015 on most critical stretches: • Eemskanaal: levees and sheet-piling • Eemshaven-Delfzijl
*http://www.rijksoverheid.nl/onderwerpen/aardbevingen-in-groningen/documenten-en-publicaties/rapporten/2014/01/17/deltaers-effecten-van-aardbevingen-op-kritische-infrastructuur.html
Important questions for EQ resistant design
• Which mechanisms to consider and which (software) models to use
• Which loading to apply
• (How to reduce liquefaction, or minimize its effect)
24-Jun-15 8
PGA
T
(Picture from: Y. Sasaki et al.)
(Picture from: Y. Sasaki et al.)
(Picture from: I. Towhata)
Further content of the presentation
• Which mechanisms to consider • Recently applied software models for the Groningen levees
24-Jun-15 9
Mechanims caused by earthquakes
• Instability or damage by earthquake force
• Crest settlement or damage by liquefaction of sand
24-Jun-15 10
Content of the presentation
• Which mechanisms to consider • Recently applied software models for the Groningen levees
24-Jun-15 11
AAD: dedicated software for embankments
24-Jun-15 12
AardbevingsAnalyse Dijken
Input: D-Geostability schematizations + cone resistances Output: Fragility curves
AAD v1: Applied Peak Ground Acceleration
Along Eemskanaal
00
01
02
03
04
05
06
07
1 10 100 1000 10000
km31.5
Along Eemshaven-Delfzijl PGA [m/s^2]
Return period [year]
• Probability distribution According to KNMI • Semi-probabilistic determination of design value
24-Jun-15 13
PGA [m/s^2]
Return period [year]
AAD v2: Nonlinear Soil Response
24-Jun-15 14
AAD: Applied Liquefaction model
24-Jun-15
EERI MNO-12*: Determines the excess pore pressure ratio 𝑟𝑢= Δ𝑢
𝜎 𝑣.0′ in clean sand as a function of
field stresses, CPT resistance (𝑞𝑐) en Peak Ground Acceleration (PGA)
Example: Sand layer of 10m, without cover layer: 𝑟𝑢 5m below surface
* Idriss, I.M. and Boulanger, R.W., 2008, Soil liquefaction during earthquakes, EERI MNO-12
𝑞c = 12 Mpa 𝑃𝑃𝑃 = 0.1𝑔
Empirical cyclic shear stress ratio resistance
15
AAD: models for the embankment
24-Jun-15 16
Displacement by sliding
Settlement by squeezing
Settlement by compaction
F
Displacement by sliding
24-Jun-15 17
Force = Mass * Acceleration Displacement above yield value by double integration
“Newmark Sliding Block” – also mentioned in EC8
Criterion: maximum sliding displacement = 0.15m (Jibson*)
For design: inverse application to determine the critical acceleration value where static stability with D-Geostability has to be preserved to keep displacement below criterion.
* Jibson, R.W., 2011, Methods for assessing the stability of slopes during earthquakes—A retrospective: Engineering Geology, v. 122, p. 43-50.
Displacement by sliding in case of liquefaction
24-Jun-15 18
Reduced tangent of the angle of friction tan(𝜙) by (partial) liquefaction • 50 % reduction during earthquake: tan𝜙reduced = 1 − 0.5ru ⋅ tan𝜙initial • 100 % reduction after earthquake: tan𝜙reduced = max (0.06, 1 − ru ⋅ tan𝜙initial) (analogous to draft Dutch National Application Document – NPR 9989)
24-Jun-15 19
Scaling of measured accelerogram
Derivation of (horizontal) design accelerograms
1. EC8: use at least 3 representative ground signals (4 used)
2. Scale measured signals for acceleration and time* with the Peak Ground Acceleration ratio: 𝑃𝑃𝑃design/𝑃𝑃𝑃measured
PGA
Crest settlement by squeezing
24-Jun-15 20
h_levee
d_top
d_liquefied
Finite Element results were used to fit an approximate (!) function between crest settlement and ℎlevee ⋅
𝑑liquefied𝑑top2 (inspired by Finn*)
* Finn, W. (2000). State-of-the-art of geotechnical earthquake engineering practice. Soil Dynamics and Earthquake Engineering, 20, pp 1-15.
Crest settlement by compaction
Accepted empirical function of cyclic Factor of Safety against liquefaction and Relative Density 𝐷𝑟 (based on Ishihara & Yoshimine*)
24-Jun-15 21
* Ishihara, K., & Yoshimine, M. (1992). Evaluation of settlements in sand deposits following liquefaction during earthquakes. Soils and Foundations, Vol. 32, No.1, March 1992, pp 173-188.
Factor of safety is cyclic shear ratio (CSR) divided by cyclic resistance ratio (CRR)
Example of resulting Fragility Curves
24-Jun-15 22
FoS Slope Stability versus PGA
Crest Settlement versus PGA
Models for Sheet-Piling: before and after EQ
Before earthquake: D-Sheetpiling
24-Jun-15 23
After earthquake: D-Sheetpiling with 100 % reduced strength by (partial) liquefaction
Models for Sheet-Piling: during EQ
24-Jun-15 24
Comparison of bending moments between generalized Mononobe-Okabe method (also suited for cohesive soil) and Finite Element Method (FEM)
FEM is preferred: Mononobe-Okabe method is too conservative, and not able to find solutions for larger PGA values
M-O
M-O
M-O
FEM
• Used to derive an approximate factor between static and dynamic moments and anchor forces (load and site specific)
• Liquefaction modeling by manual strength reduction (using 50 % of the final 𝑟𝑢) in combination with (damping) Hardening Soil Small Strain model
• In progress: comparison with implicit pore pressure generation by
constitutive models such as UBC-Sand or Hypoplasticity.
24-Jun-15 25
Current usage for Groningen levees:
Dynamic FEM for sheetpiling during EQ
24-Jun-15 26
Dynamic FEM for sheetpiling during EQ
Derivation of (horizontal) base acceleration
1. EC8: use at least 3 representative ground signals (4 used)
2. Translate measured surface signals to the base at the location of the measurement, using the local soil profile and for example the EERA* software
3. Scale the base signals for acceleration and time with the ratio 𝑃𝑃𝑃design/𝑃𝑃𝑃measured (see Newmark Sliding Block)
*A Computer Program for Equivalent linear Earthquake site Response Analyses of Layered Soil Deposits. Bardett et all, 2000
surface base
Finite Element simulation of liquefaction
24-Jun-15 27
• Hypoplasticity model is more generic than UBC-Sand model (initial state, different CSR values), but parameter determination is more difficult
• Models are sensitive to variations of dominant parameters (HP: blow count, UBC-Sand: initial void ratio)
• Current models can determine the undrained onset of liquefaction, but don’t supply reliable post-liquefaction deformations
Cyclic DSS test simulation on Loose sand
0.08
0.1
0.12
Green: UBCSAND Green: Hypoplastic
Excess Pore Pressure Ratio
28
Shear Stress Ratio
24-Jun-15
Expected further application of Finite Elements
• Liquefiability of Groningen sand with “single pulse” signals, compared to tectonic signals
• Nonlinear response of subsoil plus levee, incl. pore pressure
generation • Effectivity of mitigating measures
24-Jun-15 29