Model test of stone columns as liquefaction countermeasure in sandy soils

Abstract

The shaking table model test was conducted to investigate earthquake resistant behavior of stone columns under the intensity of an earthquake resistance of buildings is VIII. The test results show that when acceleration is less than 0.20 g, composite foundation is not liquefied, settlement is also small and pile dislocation is not observed; when acceleration is 0.3g, ground outside embankment’s slope toe is liquefied and ground within stone column composite foundation is not. It is suggesting that reinforcement scale of stone column foundation should be widened properly. The designed stone column composite foundation meets the requirements for seismic resistance.

Keywords

stone column composite foundation / seismic liquefaction / shaking table test

Introduction

Designed as China Railway I grade electrified railway, Yuxi-Mengzi Railway is an important part of the eastern line of Yunnan International Railway line (Trans-Asian Railway). The whole line of Yuxi-Mengzi Railway (142 km) is located in VII~VIII degree earthquake zone, 120 km (about 85% of the total length) of which is in VIII degree earthquake zone. Besides, the rift basins at Tonghai, Jianshui are on the railway line, which mainly contain soft plastic to flow plastic lacustrine sediments muddy clay, silt sand and fine sand. There are five sections of sand liquefaction line, total up to 10km, Tonghai basin is the typical section. Thus, Yuxi-Mengzi Railway foundation reinforcement involves not only the settlement control, but also need to have the ability to withstand high intensity earthquake. Therefore, it is necessary to study the foundation liquefaction of Yuxi-Mengzi Railway for the selection and design of anti-liquefaction strengthening method and providing a reference for engineering design.

The construction of the embankment in soft soil area is a very challenging project, involving bearing capacity, settlement and the roadbed stability, and the subgrade stability under dynamic load. Stone column is commonly used in soft foundation treatment, which can eliminate the earthquake liquefaction of liquefiable soil, improve the bearing capacity of foundation soil and reduce the compression deformation of soil under the upper load [ 1]. Although many scholars have studied the effect of stone columns in reinforcing liquefaction foundation in recent years [ 2– 14], the effect of used in high-intensity earthquake area still needs more experience.

Compared with field test, model test has the advantages in good controlling influence factors and doing test repeatedly. In model test, the simulating material, model boundary are very important, many researches [ 15– 25] do a great effort in considering the influence factors of model test and trying their best to meet the similarity criterion and guaranteeing model test results in effect, so with model test, they achieve good results in their research fields.

Thus, regarding the DK22+ 500 part of Yuxi-Mengzi Railway as the prototype, conducting the shaking table model test to study the effects and seismic performance of stone columns reinforcing liquefaction foundation in seismic intensity of VIII degrees region, which can provide parameters for the design of stone columns reinforcing seismic liquefaction foundation of Yuxi-Mengzi Railway.

Profile of test field

The test field is subgrade, which is from DK21+ 500 to DK27+ 060, the whole length is 5.56km. Because of the special sedimentary environment of the basin, the stratum structure is very complex. In the basin, sand steam in the 20 m from the surface belongs to liquefied layer, which is prone to sand liquefaction, leading to subgrade settlement and disease. The foundation soil of section DK22+ 500 contains two layers, the 0.8 m thick soft soil(muddy clay) lies upper and the silt and fine sand lies under. In this section, stone column composite foundation was selected on the purpose of resisting high intensity earthquake. The pile length is 10 m, space is 1.2 m and diameter is 0.3 m, Fig. 1 shows a vibro stone column layout.

The roadbed width is 7.8 m, center filling height is 5.48 m and the slope rate is 1:1.5, as shown in Fig. 2.

Shaking table model test

Test equipment and test components

The test is conducted on the earthquake simulation shaking table driven by unidirectional electro-hydraulic servo drive, whose table size is 4 m × 2 m, maximum load capacity is 25 t.

In the shaking table model test, the performance of the model box directly affects the accuracy of the tests and reliability of the test results. The rigid model box is welded by channel and steel and has the size of 3.5 m × 1.5 m × 2.2 m. To observe the internal situation in the model box during the test, organic glass instead of steel was facilitated at both sides of the front and rear. After repeated comparison, in order to reduce the effects of model box, a 50 mm thick of foam board is adhered to the inner wall which is perpendicular to the horizontal vibration direction. Figure 3 is the model box.

According to the purpose and requirements of the test, the testing components used are as follows:

1) Pore pressure gauge, ranges 100 kPa, resolution≤0.12%F·S.

2) Displacement gauge: Model WD310 of displacement transducer, ranges±25mm, basic error (me)<±1.

Model design and making

1) Model similarity law

To make the test results reproduce the dynamic characteristics of the prototype, a design of model similarity law was carried out [ 15]. Considering the model box size and the maximum effective load of shaking table, the model geometrical similarity ratio is 1:10. Due to the model and prototype is the same material and loaded with same acceleration of gravity, so

C_r = 1 and C_g = 1. Because of the 50mm thick of foam board, which can minimize the experimental error caused by model box effect, the damping ratio and the similarity coefficient of dynamic Poisson's ratio can take as 1. Based on the above basic similarity coefficient and the Bockingham p Theorem, the similarity coefficients of other physical quantities was derived through dimensional Analysis method, which are shown in Table 1.

2) Design of foundation model

Based on the geometry similarity coefficient, the model soil thickness is as follow: the soils from top to bottom are soft soil (8 cm), silt and fine sand reinforced (92 cm) and without reinforced (20 cm). Taking into account the experimental conditions, silt soil was used to replace the soft soil when preparing model. The density of silt soil is 1.91 g/mm³, the silt sand is 1.96 g/mm³ which are same to the soil in field. The grain composition of the soils are shown in Table 2–4.

The foundation model was prepared by dry sampling method and the compactness was controlled by mass and thickness of the filling soils. The model columns were made by the reserving PVC pipe filled with coarse sand and tamped by vibro-densification method.

3) Design of subgrade model

The embankment prototype height is 5.48 m, top width is 7.8 m, the slope of embankment slope is 1:1.5 and the subgrade bottom width is 24.3 m. According to the similarity ratio of 1:10, take the model high 0.55 m, top width 0.75 m and the bottom width 2.43 m.

This test mainly studies the liquefaction characteristics of stone column reinforced foundation soil, on which the effect of the roadbed is mainly the roadbed weight, and the type of filler is not that important, so the roadbed of the model is made by silt sand contains 3% of cement and 15% of water. Filling every 100mm and tamping and using the cast iron balls to make enough weight.

4) Adding water and saturating

Inject water into the model box through four water pipes at four corners of the box at the same time when soil filling and transducers installing is finished. Keep the water level stable in the pipes while injecting water. Observe the rising of water level through the organic glass and end the injection when the water level achieves to the ground of foundation, inject water if the water level descends until the level remains stable.

Measuring element arrangement

The arrangement of the measuring element is shown in Fig. 2. There are 16 pore pressure transducers (WP1~WP16), 7 vertical displacement transducers (DP1~DP6) and 4 lateral displacement transducers (DP13~DP16). Because of the symmetry of the model, some transducers are only arranged at one half of the model.

Loading scheme

The natural frequency of seismic waves and general foundation soil are between 1 – 2Hz and the natural period of foundation soil is around 0.6–1.4 s (natural frequency is around 0.7–1.7Hz), besides, there is a limitation on maximum acceleration of the shaking table when loading at 1 Hz. So 2 Hz is suitable for testing and data analyzing.

In the test, input a sine wave whose frequency is 2 Hz via shaking table, the duration of each loading is 10 s and the direction is along the roadbed section. The Yuxi-Mengzi Railway line has a seismic fortification intensity of VIII degree, corresponding seismic loading acceleration is 0.2g. In test, the loading acceleration must higher than the 0.2 g, so the loading acceleration is from 0.05 to 0.3g, and the step is 0.05 g.

Test results analysis

Macro-phenomena of the test

The shaking table model test well simulated macro-phenomena like site liquefaction and construction failure triggered by natural earthquake. The process of test load and model demolition has been specifically recorded as follows:

1) During 0.05 to 0.2g loading acceleration, there is no visible settlement on embankment; at 0.25 g loading acceleration, there is obvious uplift on both sides of ground surface and visible settlement can be seen on embankment itself; while at 0.3g loading acceleration, huge settlement of the embankment and subgrade failure appear.

2) The horizontal shear displacement of ground is not obvious when loading acceleration is less and gradually increases with the increase of loading acceleration. At 0.3 g loading acceleration, both sides of subgrade and the upper embankment appear cracks parallel to the strike of subgrade coming with crack transfixion and subgrade failure, as shown in Fig. 3.

3) At 0.25 g loading acceleration, sandboils and waterspouts begin to appear on ground surface. At 0.3 g loading acceleration, sandboils and waterspouts significantly increase, leaving puddle with depth up to 30 mm between subgrade and the uplift of ground surface. The water completely seep back into the sandy soil and sand holes caused by sandboils and waterspouts can be seen on ground surface in 2 h after the test.

4) Dredging the upper embankment after test loading finished, the deformation of ground surface is midst concave and both sides convex along the cross-sectional direction of subgrade. During excavation foundation, there is no obvious dislocation pile.

Pore water pressure

Figures 4(a) and (b) respectively show the relationship between the max pore water pressure and loading acceleration of the same depth at different horizontal position and the same horizontal position at different depth. It shows that pore water pressure is almost zero at 0.2 g loading acceleration; while during 0.25 to 0.3g loading acceleration it rapidly increases; in the horizontal direction, the pore water pressure near the center of the subgrade is greater than the outer side of the subgrade; in the vertical direction, the pore water of the deeper depth is greater than that of more shallow depth. It indicates that the deeper the depth, the longer the drainage time, so the more obvious the cumulative effect of pore pressure.

Characteristics of pore pressure ratio

Pore pressure ratio as a basis for judgment is intuitive and easy to judge. Commonly, it can be considered as foundation liquidation when pore pressure ratio reaches 1.0. At loading acceleration between 0.1 to 0.2 g, the increase of excess pore water pressure measured by 15 sensors (WP14 is broken) is modest, even the maximal value is less than 0.5kPa. Concluding that excess pore pressure ratio is smaller and the regularity is not so obvious, this article just has a list of the maximum pore pressure ratio under loading acceleration of 0.25 and 0.3g (Table 5).

At 0.25 g loading acceleration, although the rise of pore pressure ratio is of different degree, the foundation is not liquefied since the maximal value is less than 0.5. At 0.3g loading acceleration, pore pressure ratio of column 1 except WP13 all exceeds 1.0 which means foundation nearby is liquefied, as well as pore pressure ratio of column 2, while foundation of column 3 and 4 is few liquefied. It follows that self weight of embankment improves the effective stress within the scope of embankment and bears to the seismic dynamic load, thus reducing opportunity of liquefaction. Subgrade is supposed to properly broaden in that ground outside embankment’s slope toe is liquefied and ground within stone column composite foundation is not.

Settlement and deformation of subgrade

Vertical deformation

Settlement of subgrade is detected by sensors numbered DP1~DP6, DP1 and DP2 of which are broken thus being replaced by a steel rule. Here suppose the positive direction downward.

Figure 5 shows the relation between the accumulated settlement of ground and loading acceleration. It shows that the accumulated settlement of ground increases with the increase of loading acceleration

Figure 6 shows the accumulated settlement distribution of ground surface along subgrade transverse section direction during the whole process of loading (positive value means settlement, negative value means uplift). The accumulated settlement distribution of ground surface decreases gradually with the increase of the distance away from the ground center line. The foundation begins to uplift nearby the toe of embankment and the uplift increases with the increase of the distance away from toe of embankment. This is sufficiently consistent with the situation where ground surface of the embankment is dredged after test finished.

Horizontal deformation

Horizontal deformation of subgrade is detected by four contact displacement sensors numbered DP13–DP16. Because subgrade is set as bilateral symmetry, here suppose the positive direction vertical outward the center line.

Figure 7 shows the accumulated horizontal displacement distribution of ground surface. At 0.05 to 0.2g loading acceleration, there is no visible horizontal displacement at toe and center line of the subgrade; at loading acceleration 0.25 g, there are some movements on both sides of toe; at 0.3 g loading acceleration, the huge settlement at the middle of subgrade appears, corresponding to macro-phenomena is embankment slope failure.

Conclusions

1) From model test results, ground outside embankment’s slope toe is liquefied at the loading acceleration of 0.3g and ground within stone column composite foundation is not. Subgrade is supposed to properly broaden in that at 0.3g loading acceleration.

2) Yu Meng railway is designed to resist earthquake of VIII seismic intensity (the max seismic loading acceleration is 0.20 g). The experimental results show that composite foundation designed is not liquefied, settlement is also small and pile dislocation has not occurred yet at loading acceleration of 0.20 g, the design meets requirements of overall stability of the foundation and subgrade.

References

[1] Yang G Q. Construction technique for treating vibrating liquefied foundation of high-speed railway line by dry vibrating crushed stone pile. Journal of railway engineering society, 2004, (2): 31–34

[2] Englehardt K, Golding H C. Field testing to evaluate stone column performance in a seismic area. Geotechnique, 1975, 25(1): 61–69

[3] Sasaki Y, Taniguchi E. Shaking table tests on gravel drains to prevent liquefaction of sand deposits. Soil and Foundation, 1982, 22(3): 1–14

[4] Baez J I, Martin G R. Permeability and shear wave velocity of vibro-replacement stone columns. Soil Improvement for Earthquake Hazard Mitigation. ASCE Geotechnical Special Publication, 1995, 49: 66–81

[5] Boulanger R W, Idriss I M, Stewart D P, Hashash Y, Schmidt B. Draigage capacity of stone columns or gravel drains for mitigating liquefaction. ASCE Geotechnical Special Publication, 1998, 1:678–690

[6] Adalier K, Elgamal A, Meneses J, Baez J I. Stone columns as liquefaction countermeasure in non-plastic silty soils. Soil Dynamics and Earthquake Engineering, 2003, 23(7): 571–584

[7] Huang C X, Zhang H Y, Sui Z L, Jin J J. Shaking table test on liquefaction properties of saturated sand ground. Chinese Journal of Geotechnical Engineering, 2006, 28(12): 2098–2103

[8] Jiang G L, Liu X F, Zhang J W, Zhao R Y. Shaking table test of composite foundation reinforcement of saturated silty soil ground for high speed railway. Journal of Southwest Jiaotong University, 2006, 41(2): 190–196

[9] Bhattacharya S, Hyodo M, Goda K, Tazoh T, Taylor C A. Liquefaction of soil in the Tokyo Bay area from the 2011 Tohoku (Japan) earthquake. Soil Dynamics and Earthquake Engineering, 2011, 31(11): 1618–1628

[10] Li J Y. Application of foundation treatment with vibro-replacement stone piles to Yinping hydropower station. Chinese Journal of Rock Mechanics and Engineering, 2013, 32(S): 2968–2976

[11] Asgari A, Oliaei M, Bagheri M. Numerical simulation of improvement of a liquefiable soil layer using stone column and pile-pinning techniques. Soil Dynamics and Earthquake Engineering, 2013, 51: 77–96

[12] Zhan Y X, Jiang G L, Yao H L. Dynamic characteristics of saturated silty soil ground treated by stone column composite foundation. Advances in Materials Science and Engineering, 2014, 3: 1–7

[13] Forcellini D, Tarantino A M. Assessment of stone columns as a mitigation technique of liquefaction-induced effects during Italian earthquakes (May 2012). The Scientific World Journal, 2014, 2014(1): 1–8

[14] Liang T, Cong S Y, Ling X Z, Lu J C, Elgamal A. Numerical study on ground improvement for liquefaction mitigation using stone columns encased with geosynthetics. Geotextiles and Geomembranes, 2015, 43(2): 190–195

[15] Finn W D L, Fujita N. Piles in liquefiable soils: seismic analysis and design issues. Soil Dynamics and Earthquake Engineering, 2002, 22(9-12): 731–742

[16] Zhuang X, Augarde C, Mathisen K. Fracture modelling using meshless methods and level sets in 3D: framework and modelling. International Journal for Numerical Methods in Engineering, 2012, 92(11): 969–998

[17] Zheng W B, Zhuang X Y, Dwayne D. Tannant, Cai Y C, Samuel Nunoo. A continuous/discontinuous deformation analysis (CDDA) method based on deformable blocks for fracture modeling. Frontiers of Structural & Civil Engineering, 2013, 7(4)

[18] Budarapu P R, Javvaji B, Sutrakar V K, Roy Mahapatra D, Zi G, Rabczuk T. Crack propagation in graphene. Journal of Applied Physics, 2015, 118(6): 064307

[19] Sudhir Sastry Y B, Budarapu P R, Madhavi N, Krishna Y. Buckling analysis of thin wall stiffened composite panels. Computational Materials Science, 2015, 96B: 459–471

[20] Budarapu P R, Narayana T S S, Rammohan B, Rabczuk T. Directionality of sound radiation from rectangular panels. Applied Acoustics, 2015, 89: 128–140

[21] Yang S W, Budarapu P R, Mahapatra D R, Bordas S P A, Zi G, Rabczuk T. A meshless adaptive multiscale method for fracture. Computational Materials Science, 2015, 96(B): 382–395

[22] Sudhir Sastry Y B, Budarapu P R, Krishna Y, Devaraj S. Studies on ballistic impact of the composite panels. Theoretical and Applied Fracture Mechanics, 2014, 72: 2–12

[23] Budarapu P R, Sudhir Sastry Y B, Javvaji B, Mahapatra D R. Vibration analysis of multi-walled carbon nanotubes embedded in elastic medium. Frontiers of Structural and Civil Engineering, 2014, 8(2): 151–159

[24] Budarapu P R, Gracie R, Yang S W, Zhaung X, Rabczuk T. Efficient coarse graining in multiscale modeling of fracture. Theoretical and Applied Fracture Mechanics, 2014, 69: 126–143

[25] Budarapu P R, Gracie R, Bordas S P A, Rabczuk T. An adaptive multiscale method for quasi-static crack growth. Computational Mechanics, 2014, 53(6): 1129–1148

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg