Recent Research on EPS Geofoam Seismic Buffers

Richard J. Bathurst and Saman Zarnani GeoEngineering Centre at Queens-RMC Canada

What is a wall (SEISMIC) buffer?

A compressible inclusion placed between a rigid wall and the retained soil Purpose: To reduce lateral earth pressure by allowing controlled yielding of backfill (soil straining) Can be used for both static and dynamic loading conditions For static case, reduction of pressure to near active case (quasi -active) For dynamic earth pressure case, the concept of earth pressure reduction is the same except that the loads are higher The product of choice is expanded polystyrene geofoam (EPS)

retained soil rigid basement wall

Geofoam blocks

buffer

First example of EPS seismic buffer

Inglis et al. 1996 Deep basement in Vancouver BC Canada

Numerical analysis (FLAC) showed that the EPS seismic buffer (1 m thick) could reduce seismic forces on the rigid basement walls by up to 50%

PROOF OF CONCEPT

Experimental study: General arrangement of shaking table tests

One control wall without buffer and 6 walls with different buffer densities were tested

(Bathurst, R.J., Zarnani, S. and Gaskin, A. 2007. Shaking table testing of geofoam seismic buffers. Soil Dynamics and Earthquake Engineering, Vol. 27, No. 4, pp. 324-332.)

View of geofoam buffer during construction

1.4 m

Experimental study: Properties of EPS geofoam buffer material

Wall # 1 2 EPS bulk density (kg/m3) EPS initial tangent Youngs modulus (MPa) EPS Thickness (m) EPS type (ASTM C 578)

Control structure (rigid wall with no seismic buffer) 16 4.7 0.15 I

3
4 5

12
14 6 (50% removed by cutting strips)

3.1
0.6 1.6

0.15
0.15 0.15

XI
Elasticized XI

6 (57% removed by coring)

1.32 (89% removed by coring)

1.3

0.15

XI

7
Note:

0.34

0.15

XI

Density of unmodified EPS geofoam = 12 kg/m3

Experimental study: Properties of backfill soil

artificial sintered synthetic olivine material (JetMag 30-60) silica-free
Property Density Peak angle of friction Residual friction angle Cohesion Relative density Dilation angle Value 1550 kg/m3 51 46 0 kPa 86% 15

Experimental study: Table excitation

1.0 0.8 0.6

Acceleration (g)

0.4 0.2 0.0 -0.2 -0.4 -0.6 -0.8 -1.0 0 10 20 30 40 50 Time (s) 60 70 80 90 100

stepped-amplitude sinusoidal base input excitation frequency = 5Hz

1.0 0.8 0.6 0.4 0.2 0.0 -0.2 -0.4 -0.6 -0.8 -1.0 39 40 Time (s) 41 42

Acceleration (g)

3-second window

Experimental study: Buffer forces

Experimental study: Total force versus (peak) acceleration

24 22 20 18 16 14 12 10 8 6 4 2 0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Wall 1 (no buffer)

Ftotal

horizontal wall force (kN)

Wall 2 3 buffer density =16 kg/m Wall 7 3 buffer density =1.32 kg/m

0.9

1.0

1.1

1.2

acceleration (g)

(Zarnani, S. and Bathurst, R.J. 2007. Experimental Investigation of EPS geofoam seismic buffers using shaking table tests, Geosynthetics International, Vol. 14, No. 3, pp. 165-177.)

Experimental study: Buffer compressive strains and stresses

Experimental study: Dynamic geofoam modulus

Experimental study: Dynamic geofoam modulus

10

initial elastic Young's modulus, E i (MPa)

modified EPS

maximum

average

range of values reported range of modulus values in the literature based on correlations (Bathurst et al. 2006a) reported by Bathurst et al. (2006)

minimum

0.1 0 2 4 6 8 10 12 14 16 18

geofoam bulk density (kg/m3)

NUMERICAL MODEL VERIFICATION

Numerical studies: Model in FLAC

A slip and separation interface with friction angle of 15

Numerical study:

actual shaking

Constitutive models
Soil modeled as a purely frictional, elastic-plastic material with Mohr-Coulomb failure criterion
Perfectly plastic

Elastic

e
Soil M-C model

Geofoam buffer material modeled as a linear elastic, purely cohesive material

Elastic 1%

Geofoam

Numerical studies: Numerical results - Forces

Ftotal

14000 Wall 2, EPS 12000 = 16 kg/m3 experimental

12000 Wall 7, EPS 10000 = 1.32 kg/m3 experimental

total wall force (N / m)

10000 8000 6000 4000 2000 0 0 10 20 30 40 50 60 70 80 90 100 numerical

total wall force (N / m)

8000 6000 4000 2000 0 0 10 20 30 40 50 60 70 80 90 100 110 numerical

time (s)

time (s)

Wall 2, EPS =16 kg/m3

Wall 7, EPS =1.3 kg/m3

(Zarnani, S. and Bathurst, R.J. 2008. Numerical modeling of EPS seismic buffer shaking table tests, Geotextiles and Geomembranes. Vol. 26, No. 5, pp. 371-383.)

Influence of constitutive model on numerical results?

Simple M-C model

Equivalent Linear Method (ELM)

unload-reload cycles with hysteresis behavior

modulus degradation and damping ratio variation

Influence of material constitutive model, ELM

Shear modulus variation

Damping ratio variation

Resonant column testing of geofoam specimens

Cyclic load testing of geofoam specimens using PIV

EPS material properties for ELM hysteresis model

1.0

a)
0.8 EPS type confinement Athanasopoulos et al. (1999)

G / Gmax

0.6

0.4

0.2

D24 - 0 kPa D24 - 30 kPa D24 - 60 kPa Ossa & Romo (2008) D30 - 0 kPa D30 - 30 kPa D32 - 60 kPa D15 - 0 kPa D15 - 20 kPa current study D29 - 0 kPa D29 - 20 kPa Athanasopoulos et al. used in this study (2007) 0.0001 0.001 0.01 0.1 1

0.0 0.00001

10

100

cyclic shear strain (%)

30

b)
25

damping ratio (%)

20 Athanasopoulos et al. (1999) 15 10 5 0 0.00001

0.0001

0.001

0.01

0.1

10

100

cyclic shear strain (%)

Influence of material constitutive model, ELM

1.0

a)
0.8

G / Gmax

0.6

fit with FLAC default function

0.4

0.2

Sand modulus degradation & damping curves

damping ratio (%)

range of shear modulus values for sand (Seed and Idriss 1970) 0.0 0.00001 0.0001 0.001 0.01 0.1 1 10 100

cyclic shear strain (%)

70

b)
60 50 40 30 20 10 0 0.00001 range of damping ratio values for sand (Seed and Idriss 1970) fit with FLAC default function

0.0001

0.001

0.01

0.1

10

100

cyclic shear strain (%)

Numerical studies: Influence of material constitutive model Comparison of numerical results (RIGID wall)
20 18 16 experimental, Test 1, Rigid control wall numerical (ELM, with hysteresis damping) numerical (linear elastic-plastic, with constant Rayleigh damping)
rigid wall geofoam

a)

wall force (kN/m)

14 12 10 8 6 4 2 0 0

20

40

60

80

100

time (s)

(Zarnani, S. and Bathurst, R.J. 2009. Influence of constitutive model on numerical simulation of EPS seismic buffer shaking table tests. Geotextiles and Geomembranes, Vol. 27, No. 4, pp. 308-312.)

Numerical studies: Influence of material constitutive model Comparison of numerical results (EPS wall)
20 18 16 experimental, Test 2, EPS density = 16 kg/m3 numerical (ELM, with hysteresis damping) numerical (linear elastic-plastic, with constant Rayleigh damping)

b)

wall force (kN/m)

14 12 10 8 6 4 2 0 0 20 40 60 80 100

time (s)

(Zarnani, S. and Bathurst, R.J. 2009. Influence of constitutive model on numerical simulation of EPS seismic buffer shaking table tests. Geotextiles and Geomembranes, Vol. 27, No. 4, pp. 308-312.)

PARAMETRIC NUMERICAL STUDY

Parametric numerical studies: Matrix of variables

Input excitation Wall height (H) backfill width (B) Thickness of geofoam (t / H)* Type of EPS geofoam #

Peak acceleration

(f / f11) 0.3 0.5

0.85

1 (m) 5 (m) 3 (m) 15 (m)

6 (m) 30 (m)

0 0.025
0.05

EPS19 EPS22
EPS29 0.7g

9 (m) 45 (m)

0.1
0.2 0.4

1.2
1.4

t = seismic buffer thickness = 0 to 3.6 m # based on ASTM D6817-06 f = predominant frequency of the input excitation and f11 = natural frequency of the wall-backfill system

Parametric numerical studies: Model excitation

Variable amplitude sinusoidal acceleration record:

(t ) u

e t t sin(2 ft)
0.8 0.6 0.4

f = 1.25 Hz f / f11 = 0.5 for 6 m high wall

acceleration (g)

0.2 0.0 -0.2 -0.4 -0.6 -0.8 0 2 4 6 8 10 12 14 16 18

time (s)

Parametric numerical studies: Material properties of backfill soil

loose to medium dense sand
modeled as frictional material with elastic-perfectly plastic MohrCoulomb failure criterion small cohesion to ensure numerical stability at the unconfined soil surface when models were excited at high frequencies
Property Unit weight Value 18.4 kN/m3

Friction angle Cohesion

Shear modulus Bulk modulus

38 3 kPa
6.25 MPa 8.33 MPa

Parametric numerical studies: Material properties of EPS geofoam

Modeled as purely cohesive material with elastic-perfectly plastic Mohr-Coulomb failure criterion
Type
EPS19 EPS22 19 81.4 22 102 EPS29 29 150

Property Density (kg/m3)

Yield (compressive) strength (kPa) Shear strength (kPa) Youngs modulus (MPa)
Poissons ratio

40.7 5.69
0.1

51 6.9
0.12

75 9.75
0.16

Parametric numerical studies: Example wall force-time response

3 m-high wall with EPS22 excited at 0.3 f11
300
H=3m EPS22 f = 0.3f11 Control Control wall Control wall wall maximum maximum maximum wall wall force force wall with with force geofoam geofoam with t geofoam =t 0.05H = 0.05H t = 0.05H maximum maximum wall force wall with force geofoam with geofoam t = 0.1H t = 0.1H maximum wall force with geofoam t = 0.2H

250

wall force (kN/m)

200
maximum wall force-control case

150

100

50

0 0 2 4 6 8 10 12 14 16 18

time (s)

Parametric numerical studies: New design and performance parameters

Buffer stiffness K (MN/m3 )

E t

Elastic modulus of geofoam geofoam thickness

Isolation efficiency

peak force (rigid wall) peak force (seismic buffer) peak force (rigid wall)

100%

(Zarnani, S. and Bathurst, R.J. 2009. Numerical parametric study of EPS geofoam seismic buffers, Canadian Geotechnical Journal Vol. 46, No. 3, pp. 318-338.)

Design charts
70

a) H = 1 m
60

0.3f11 EPS19 EPS22 EPS29

1.4f11 EPS19 EPS22 EPS29

70

b) H = 3 m
60

0.3f11 EPS19 EPS22 EPS29

1.4f11 EPS19 EPS22 EPS29

isolation efficiency (%)

50 40 30 20 10 0 0 50

isolation efficiency (%)

50 40 30 20 10 0

100

150
3

200

50

100
3

150

K = E/t (MN/m )
70

K = E/t (MN/m )
1.4f11 EPS19 EPS22 EPS29 70

c) H = 6 m
60

0.3f11 EPS19 EPS22 EPS29

d) H = 9 m
60

0.3f11 EPS19 EPS22 EPS29

1.4f11 EPS19 EPS22 EPS29

isolation efficiency (%)

50 40 30 20 10 0 0 20 40

isolation efficiency (%)

50 40 30 20 10 0

60
3

80

100

10

20

30
3

40

50

K = E/t (MN/m )

K = E/t (MN/m )

Influence of earthquake record

0.8 0.6 0.4 0.2 0.0 -0.2 -0.4 -0.6 -0.8 0 10 20 30 time (s)
Kobe earthquake (1995)

acceleration (g)

17

40

50

60

Conclusions
Experimental shaking table test results and numerical simulations demonstrated proof of concept for using EPS geofoam material as a seismic buffer to attenuate dynamic earth pressures against rigid retaining walls. The magnitude of seismic load reduction in shaking table models was as high as 40% for the softest geofoam.
The numerical simulations of the experiments showed similar reductions in seismic-induced lateral earth force observed in physical tests.

A verified FLAC numerical model was used to carryout a parametric study to investigate the influence of different parameters on buffer performance and isolation efficiency:
Significant load attenuation occurs by introducing a thin layer of geofoam (> 0.05H) at the back of the wall and the attenuation increases as the thickness of the buffer increases. The least stiff EPS geofoam in this study resulted in the largest load attenuation.

Conclusions
The practical quantity of interest to attenuate dynamic loads using a seismic buffer is the buffer stiffness defined as:
K=E/t

For the range of parameters investigated in this study, K < 50 MN/m3 was observed to be the practical range for the design of these systems to attenuate earthquake loads.

Recent example of EPS application as seismic buffer

Queen Elizabeth Water Reservoir - Vancouver - Sandwell Engineering Protected with EPS geofoam from Beaver Plastics

Recent Research on EPS Geofoam Seismic Buffers

Tusen Takk