- 1757 -

Using Stone Column as a Suitable Liquefaction Remediation in Persian

Gulf Coast

Hossein Moayedi

Department of Civil Engineering, University Putra Malaysia Serdang, Selangor, Malaysia

Bujang B. K. Huat

Department of Civil Engineering, University Putra Malaysia Serdang, Selangor, Malaysia

Mehdi Mokhberi

Scientific member of board, Islamic Azad University Estahban Branch, Estahban, Iran

Arash Ansari Moghaddam

Department of Architecture, Eastern Mediterranean University, Cyprus

Shokoufeh Ansari Moghaddam Department of English Language, Faculty of Modern Communication

ABSTRACT This research investigates behavior of gravel drain piles under high-level earthquake loading beneath the structures foundation. To achieve this purpose one of the waste water septic tank project in north of Persian Gulf in Hormoz Island was selected as a case study to find suitability of gravel drain pile system to reduce excess pore water pressure. According to high susceptibility of local soil layers liquefaction and its short distance of waste water tank to the sea, the mentioned project becomes one of the most important issues regarding geo-environmentally hazards impact after tank structural collapsing. The drain piles were used to control excess pore water pressures beneath the foundation of mentioned project. Furthermore, different static and cyclic triaxial tests, Standard Penetration Test (SPT), the hydraulic conductivity and density tests were conducted to enhance the proper understanding of the dynamic behavior of soil layer under the foundation. According to the numerical modeling results, using these drain piles has focal effects on the excess pore water pressure rate and creates a liquefaction zone during the time of earthquake loading. Keyword: Stone column; Liquefaction; Numerical Modeling; Vertical drain, Loose Sand.

INTRODUCTION Generation of excess pore pressures during earthquakes in sandy soils causes significant engineering

problems; buildings may settle and tilt, slopes can slide and spread, and also subsurface sand is able to boil to the surface. Vertical drains are often installed to relieve excess pore pressures before they can

Vol. 15 [2010], Bund. P 1758 build up to dangerous levels [1]. Drains also continue dissipation rapidly after shaking is abated. Previous studies reported that liquefaction-induced failures have occurred during the prolonged period of excess pore pressures that followed an earthquake [2-5]. Since ground deformation patterns do not represent a unit cell condition, some techniques have been developed for simulating the vertical drain effect in two-dimensional finite-element analysis [6].

The main objective of the present research is to control excess pore water pressure beneath the structures which caused liquefaction in susceptible area. Several finite elements modeling were carried out and analyzed regarding vertical gravel drain to make such a reduction and/or postponed the time duration of maximum achieved excess pore water pressure during earthquake loading.

Literature review

Loosely saturated sandy soils are universally acknowledged to be susceptible to large excess pore pressure generation during earthquake shaking. If these excess pore water pressures reach a level equal to the initial effective stress, particles no longer stress one another and will be suspended that is a state of liquefaction. It is clear that denser saturated sandy soils are capable of generating excess pore pressures at this level, known as initial liquefaction [7-9]. Additionally, Remediation schemes for the treatment of liquefiable ground, such as vertical drains, compaction has evolved largely over the last 30 years as a result of consideration of uniform sand beds. Moreover, the use of vertical drains as liquefaction remediation is largely based on the original design charts of Seed and Booker, [1]. In examining the behavior of individual drains, Brennan and Madabhushi [9] re-described their operation in terms of an outwardly moving “flow front.” Within this flow front, fluid flows toward the drain, and outside the flow front, then, soil is oblivious to the drain’s presence. Benefits were seen to be more evident at the base of strata, as fluid discharging from depth has priority in the drain, installing a drainage material vertically into the ground can shorten the drainage path of alluvial deposits. In the past few decades, gravel stone column and drain pile improvement techniques have been widely applied in sandy soil engineering, against liquefaction [10-11]. By using stone column, the rate of dissipation of excess pore water pressure is accelerated so that the effective stress will not decrease very much or, even in the worst case of 100% pore pressure rise, the effective stress recovers quickly before ground fails substantially. There are some factors govern the liquefaction process for in situ soil. The intensity of earthquake and its duration, location of ground water table, soil type, the soil relative density, particle size gradation, particle shape, depositional environment of soil, soil drainage conditions, confining pressures, aging and cementation of the soil deposits, historical environment of the soil deposit, and building/additional loads on these deposits are some of the mentioned factors. In summary, the site conditions and soil type that are most susceptible to liquefaction are given in the following sections [12]. (i) Site conditions; the site that is close to the epicenter of fault rupture of a major earthquake, and has a ground water table close to ground surface. (ii) Soil type; Sand that has uniform gradation and rounded particles, very loose density state, recently deposited with no cementation between soil grains, and no prior preloading or seismic shaking. One of the key parameters regarding estimating liquefaction potential through the susceptible zone is Cyclic Shear Ratio (CSR). The cyclic strength curve is frequently normalized by the initial effective confining pressure. This normalized cyclic stress is called the Cyclic Stress Ratio, ( CSR ). CSR , is defined differently in different types of tests. For the cyclic simple shear test, CSR is defined as a ratio of the cyclic shear stress over the initial vertical effective pressure. In the cyclic triaxial test which used in this research, it is defined as a ratio of the maximum cyclic shear stress over the initial effective confining pressure.

- 1759 -

MATERIALS AND METHODS

Study Area

The research area was located in Hormoz Island in the Hormozgan province which stated in North of Persian Gulf in Iran. The project which was a 40×40 meter septic tank was built near the beach with position of N30.365027 ′′′ and E56.257256 ′′′

(Fig. 1).

Figure 1: Study area which was located in North of Persian gulf (Map from Wikipedia)

Due to high amount of waste water which is going to be stored in this tank, it is important to adopt suitable remedies to enhance liquefaction risk reduction during earthquake loading. The static parameters that used in numerical analysis are presented in Table 1. Different tests such as SPT, permeability, and triaxial test were conducted in order to observe the static and dynamic properties of soil layers.

Table 1: Soil layer properties in the study area

Materials Layer Depth Saturation

Unit Weight Permeability

Internal Friction Angle

Cohesion Elastic

modulus Poisson

ratio

(m) kN/m m/day Deg [o] kPa kPa --

Loose Sand 0--10 18.2 3.888 30 1 15000 0.33 Dense Sand 10--40 20.4 4.320 36 1 40000 0.33

Due to the architectural plan, a 10 meter excavation height was required for waste water septic tank. Since project location were nearby Persian gulf coast and it has a high susceptibility to geoenvironmental effect on the sea water after collapsing foundation, that is one of the key issues related to the stability of structure against the liquefaction problem to find depth of liquefiable under tank foundation. It should be mentioned that ground water level was in 2 meter depth from the ground surface and this was also an important matter since it is the main reason that excess pore water pressure will increase during any seismic loading. Different samples have been observed to obtain static and dynamic parameters from

Vol. 15 [2010], Bund. P 1760 different depths. Consequently, depths were classified in two loose and dense layers one due to their SPT value.

Initial Liquefaction Estimation

There are a number of ways to determine possible liquefaction depth beneath the structrues; however, more conveniently, using SPT test recommended [1-8-16]. Due to the SPT results, the study area is clearly susceptible to liquefaction. Figures 2 and 3 showed the liquefaction zone calculation which have estimated for two different cases. The first case belongs to the initial site condition and the second case considered additional loading caused by foundation weight.

Figure 2: liquefiable depth in study area

It is worth nothing that foundation weight significantly affect liquefying zone. Obviously, during both cases sandy soil were liquefied at different depths, however, in first case, it was liquefied up to 8 meters from the ground level (Fig. 2), however, in case 2 only 2.5 meters under the foundation sandy soil liquefied (Fig. 3). It is clear that, uplift pressure is produced under foundation actually try to decrease factor of safety against liquefaction while foundation weight prevent liquefaction under lower depths.

- 1761 -

Figure 3: liquefiable depth under foundation- beneath the 10 meter depth

Soil Dynamic Properties Measurement

Cyclic triaxial tests were carried out to find dynamic properties of insitu soil such as damping ratio D, maximum shear modulus, maxG and excess pore water pressure built up based on cyclic loading. Density

of soils has changed based on SPT results. Consequently, samples for cyclic triaxial test were prepared in different density. Increasing initial sandy soil densities caused significant increase in shear modulus of soils. However, more or less the same values of shear modulus occur beyond 0.5% shear strain level irrespective of their initial density. An increase in the density results in an increase in the cyclic strength of the soil thereby making it less susceptible to liquefaction. The amplitude of cyclic shear strain governs the liquefaction resistance of a soil characterized by the cyclic strain approach [16].

RESULTS AND DISCUSSIONS

Soil Layer Properties

Results from cyclic triaxial test applied in frequencies 0.1 HZ under 100 kPa confining pressure. In such case the cyclic stress ratio (CSR) was 0.23. From cycle 16th until the end, specimen starts to show large amount of axial strains and Excess pore water pressure was built up significantly from 60 kPa in 16th cycle to 100kPa in time when liquefaction occurred in cyclic number 23. Also, pore pressure ratio

ur (here in cyclic triaxial test equal to 32/ ceu σ ′ ) becomes equal to 1.0 in this cycle loading which caused

liquefaction (Fig. 4).

Vol. 15 [2010], Bund. P 1762

Figure 4: Result of cyclic triaxial test for dense sandy silt soil (12% silt) with Dr, = 60%, kPac 1003 =′σ .

(a) Deviatoric stress q , (b) Pore water pressure increasing,

(c) Axial strain increasing, (d) Stress path pq ′−

It is obvious that the pore water pressure builds up steadily as the cyclic shear strain is applied and eventually approaches a value equal to the initially applied confining pressure of kPa100 (cyclic pore

pressure ratio=100%) in the th23 cycles of loading. Increasing in pore water pressure, consequently, effective stress will be decreased and eventually it will be reduced to zero when the pore water pressure ratio rises to 100%. Such a state of the specimen is recognized as 'liquefaction' which is a state of softening produced suddenly with the complete loss of shear strength or stiffness. Moreover, dynamic properties of soil beneath the foundation showed in Table 2.

Table 2: Dynamic properties of soils

Soil Type: Depth From

Ground Level

Shear Modulus

(G)

Maximum. Shear modulus (Gmax)

Damping ratio

(Silty Sand) (m) MPa MPa -- Loose 0--10 30 107.1 0.1 Dense 10--30 70 194.4 0.08

Table 3 shows the cyclic stress ratio, CSR and number of cycles causing liquefaction for both loose and densely silty sands. Higher densities were carried out regarding deeper depths which samples were taken.

- 1763 -

Table 3: Cyclic stress ratio and number of cycles caused liquefaction

Soil Type Identity Silt (%) R.D. (%) C.S.R. eqN

SM Loose 18 30 0.15 10 SM Dense 12 60 0.23 10

Numerical Modeling

A plane strain modeling was carried out using equivalent linear analysis. The actual nonlinear behavior of the shear modulus and damping ratio under dynamic loading conditions can be simulated approximately by an equivalent linear analysis. In an equivalent linear analysis, the constant G and damping ratio are used during a dynamic loading analysis. The new modulus G and damping ratio are computed from the cyclic or an equivalent cyclic shear strain obtained in the dynamic analysis. Then, a new dynamic loading analysis starts with the new modulus and damping ratio.

The size of gravel drain piles is based on field conditions, however, regularly the intervals among holes are 1.5 meter and its diameter considered 0.5 meter. According to setup procedure of gravel drain piles which do not compact in most cases, the stiffness of gravel material did not take into consideration and considered equal to loose sand. This was done because the main objective of the study is to consider comparison of excess pore water pressure induced through the drain piles and not to consider the strength absorption of gravel material or dissipation of excess pore pressure to use drain piles. It is noteworthy that stiffer materials give rise to more stress absorption through the drain pile materials. Figure 7 shows the schematic view of septic tank modeling and gravel drain pile placement under its foundation.

Figure 7: schematic view of drain pile modeling through Quake/W

Figure 8 shows variation of excess pore water pressure for both models in two different times after starting of earthquake. As depicted excess pore pressure values in non drain system model sharply increased to its high amount which caused liquefaction in initial time steps after earthquake. Amount of pore pressure ratio reaches to 1.0 in two seconds after the earthquake shaking (Fig. 8). This means that liquefaction will appear shortly after the earthquake starts without using gravel drain pile system. Also, as for model using gravel drain piles system, excess pore water pressure will increase more slightly and up to 9 seconds did not show any significant liquefaction zone during earthquake. Since, the drain piles have

Vol. 15 [2010], Bund. P 1764 greater permeability, may be by a factor of 100 as a results, less excess pore pressure will be produced during earthquake loading. High capability of excess pore pressure dissipation of drain piles rises from their intent to faster drain in comparison with the surrounding soil. Thus fluid is given the additional incentive of a radial hydraulic gradient towards the nearest drain. The effect is not immediate and does not affect all soil simultaneously.

Figure 8: Excess pore pressure in two different times

(a) without draining system, (b) with gravel drain pile system,

Figure 8 shows non-draining system in center line of the foundation which took 4 seconds to liquefy and it started from the surface and continued to depth. It is worth noting that in these depths the pore pressure ratio reaches its highest level which is equal to 1.0. It is well understood that increasing in depth did actually lead to increase safety against liquefaction. It is obvious that for depth more than 1.5 meter, liquefaction will not occur since the amount of ur for the mentioned depth will not reach 1.0. As a

Geotechnical point of view, using vertical drain leads to an increase safety against liquefaction. The amount of excess pore pressure in different depths for drained system (Fig. 9 and 11) is also much less than what observed in non-drained system. For instance, two different locations observed in this model regarding showing the differences between excess pore pressure through the vertical drain and the farthest place from the draining positions. The only place which was susceptible to liquefaction is at the first 0.5 meter depth near the surface which was located in farthest place from the drainage system. As for better construction condition of stone column, Geogrid are well advised to be used as a container due to its high tension capacity [17-19].

- 1765 -

Figure 9: Excess pore water pressure versus depths in centerline without drain pile

Figure 10: Excess pore water pressure versus depths in adjust to drain pile system

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

0 5 10 15 20 25 30 35 40 45 50Excess Pore Water Pressure [kPa]

Dep

th [

m]

T=0 sec

T=1 sec

T=2 sec

T=3 sec

T=4 sec

T=5 sec

T=6 sec

T=7 sec

T=8 sec

T=9 sec

T=10 sec

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

0 5 10 15 20 25 30 35 40Excess Pore Water Pressure [kPa]

Dep

th [

m]

T=0 sec

T=1 sec

T=2 sec

T=3 sec

T=4 sec

T=5 sec

T=6 sec

T=7 sec

T=8 sec

T=9 sec

T=10 sec

Vol. 15 [2010], Bund. P 1766

Figure 11: Excess pore water pressure versus depths in 0.5 m far from drain system

CONCLUSION Utilized gravel drain pile technique is in association with a two-dimensional plane strain finite

element analysis which enhanced behavior of vertical gravel drain pile systems. In present study, such technique were used under foundation of the waste water septic tank project in Hormoz Island in Persian Gulf coast located in south of Iran. The result clearly showed that using the draining system beneath the foundation restraining against liquefaction generation during the earthquake loading. Furthermore, it was observed that pore water pressure ratio remains less than 1.0 at different depths before 10 seconds earthquake loading when gravel drain pile used. Therefore, these depths were not liquefied during the earthquake loading even with more equivalent cyclic numbers. Pore pressure ratio also is a key parameter to design granular materials through the drain pile systems and should be retain as less as possible. The installation of the drain pile mitigates the potential for liquefaction by increasing the density of surrounding soil, allowing drainage for the control of pore pressures, introducing stiff elements (compacted stone or gravel columns) which potentially able to carry higher stress levels. It causes reduction in stress levels in the surrounding soil and provides a deformation restricting effect. If higher compaction considered for drain piles, stiffer material will be achieved and this lead to more resistant structure against liquefaction.

REFERENCES 1. Seed H B, Booker J R. Stabilization of potentially liquefiable sand deposits using gravel drains [J].

ASCE, Journal of Geotechnical Engineering Division, 1977, 103, (7): 757–768.

2. Inagaki K, Iai S, Sugano T, Yamakaki H, Inamoto T. Performance of caisson type quay walls at Kobe port [J]. Soils and foundations special edition, 1996: 119–36.

3. Barron R A. Consolidation of fine-grained soils by drain wells.’’ Trans. ASCE, 1948, (113): 718–742.

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

0 5 10 15 20 25 30 35 40Excess Pore Water Pressure [kPa]

Dep

th [

m]

T=0 sec

T=1 sec

T=2 sec

T=3 sec

T=4 sec

T=5 sec

T=6 sec

T=7 sec

T=8 sec

T=9 sec

T=10 sec

- 1767 -

4. Yoshikuni H. Design and construction control of vertical drain methods [M]. Gihoto, Tokyo, 1979, (in Japanese).

5. Hansbo S. Consolidation of fine-grained soils by prefabricated drains [C]. Proc. 10th Int. Conf. Soil Mech. and Found. Engrg., International Society of Soil Mechanics and Foundation Engineering, Stockholm, 1981 (3): 677–682.

6. Chai J C, Miura N, Sakajo S, Bergado D T. Behavior of vertical drain improved subsoil under embankment loading [J]. Soils and Found., Tokyo, 1995, 35(4): 49–61.

7. Lee, K. L., and H. B. Seed. cyclic stress conditions causing liquefaction of sand.” J. Soil Mech. Found. Div., 1967, 93 (1): 47–70.

8. Huang Y, Yashima A, Sawada K, Zhang A. A case study of seismic response of earth embankment foundation on liquefiable soils [J]. Journal of central south university of technology, 2009, 16(6), 994-1000.

9. Brennan A J, Madabhushi S P G. Liquefaction and Drainage in Stratified Soil [J]. Journal of Geotechnical and Geoenvironmental Engineering, American Society of Civil Engineers, 2004, 131 (7): 876-885.

10. Liu H I, Chen Y M, Zhao N. Development technology of rigidity-drain pile and numerical analysis of its anti-liquefaction characteristics [J]. Journal of central south university of technology, 2008, 15: 101-107.

11. Towhata I. Geotechnical Earthquake Engineering [M], Springer-Verlag Berlin Heidelberg 2008.

12. Sitharam T, Govindaraju G L, Sridharan A. Dynamic properties and liquefaction potential of soils [J], Journal of CURRENT SCIENCE, 2004, 87(10): 1370-1378.

13. Millea, M T. Liquefaction mitigation technology [M]. Technical Note, National Civil Engineering Laboratory, Port Hueneme, CA, 1990.

14. Baez J I, Martin G R, Advances in the design of vibro-systems for the improvement of liquefaction resistance [C]. Proc., Symposium of Ground Improvement, Vancouver Geotechnical Society, Vancouver, B.C. 1993.

15. Baez J I. A design model for the reduction of soil liquefaction by vibro-stone columns [M]. PhD thesis, University of Southern California, Los Angeles, CA, 1995, (3): 207-215.

16. Ishihara K. Soil Behaviour in Earthquake Geotechnics, Oxford University Press, New York, 1996.

17. Moayedi, H., Huat, B.B.K., Kazemian, S. and Asadi, A., 2010, "Optimization of Shear Behavior of Reinforcement through the Reinforced Slope" Electronic Journal of Geotechnical Engineering, USA.

18. Moayedi, H., Kazemian, S. Parasad, A. and Huat, B.B.K., 2009, "Effect of Geogrid Reinforcement Location in Paved Road Improvement" Electronic Journal of Geotechnical Engineering, USA.

19. Moayedi, H., Huat, B.B.K., and Asadi, A., 2010., " Strain Absorption Optimization of Reinforcement in Geosynthetic Reinforced Slope-Experimental and FEM Modeling" Electronic Journal of Geotechnical Engineering, USA. 15N.

© 2010 ejge