Liquefaction Analysis of Sandy Ground due to Tsunami

Transcription

1 Liqueaction Analysis o Sandy Ground due to Tsunami Y. Okumura Kyoto University, Japan R. Kato Nikken Sekkei Civil Engineering Ltd., Japan F. Oka Kyoto University, Japan SUMMARY The 2011 o the Paciic coast o Tohoku earthquake causes large scale o Tunami and the huge areas have been damaged. This is probably because o the huge orce due to Tunami(JSCE investigation report, 2011). In addition, it is known that the high tidal wave causes the liqueaction in the sea bed(oka et al. 1994). These studies are, however, dealing with the short period wave such as tidal one. For the tsunami, we need to study the eect o the wave with the longer period, i.e., minutes. In the present study, we have studied the liqueaction potential and the change o the eective stress due to the change o the water pressure caused by tsunami by using a liqueaction simulation method. Keywords: Tsunami, Liqueaction, Degradation o Ground, liqueaction analysis 1. INTRODUCTION The 2011 o the Paciic coast Tohoku earthquake generated large scale o Tunami and the huge areas have been damaged. In Onagawa town o Miyagi preecture, the reinorced concrete building has been damaged by pulling up the oundation pile. This is probably because o the huge orce due to Tunami. In addition, it is known that the high tidal wave causes the liqueaction in the sea bed (Oka et al. 1994) (Figure 1.1). Sassa et al.(2001) studied the liqueaction o the sea bed and its propagation considering the ully liqueied soil as a viscous luid. Then Liu et al.(2008) extended their theory considering the two-layer luid model o liqueied luid and the surrounding region. These studies are, however, dealing with the short period wave such as tidal one. For the tsunami, we need to study the eect o the wave with the longer period, i.e., minutes. In the present study, we have studied the liqueaction potential and the change o the eective stress due to the change o the water pressure caused by tsunami by using a liqueaction simulation method. Figure 1.1. Submarine ground aected by waves

2 2. ANALYSIS METHOD There are two reasons or the liqueaction o the sea bed. One is the principle stress rotation and the other is the compressibility o the pore water. The compressibility o the pore water leads to the retardation o the dissipation o the pore water pressure against the change o the applied water pressure on the sea bed surace. In the numerical analysis, we used the numerical code based on the elasto-plastic inite element method called LIQCA2D. This method has been developed based on the u-p ormulation by the research group including the authors o this paper and has been successully applied to the liqueaction problems(oka et al., 2009, 2011). In the numerical simulation o the propagation o tsunami, we used a leap-rog scheme with staggered grid or the inite dierence method (Goto et al., 1997) Stress variables or the mixture The total stress tensor is assumed to be composed o two partial stresses or each phase. (2.1) s s a where is the total stress tensor, and,, and are the partial stress tensors or solid, liquid, and gas, respectively. The partial stress tensors or unsaturated soil can be given as ollows: ns p (2.2) r s (1 np ) (2.3) where is the eective stress tensor, p is the pore water pressure, respectively, n is the porosity. Using Eqs. (2.1), (2.2) and (2.3), we have the total stress tensor as p (2.4) 2.2. Constitutive equations or solid and luid The constitutive equation or soil skeleton is given based on the elasto-plasticity theory (Oka et al., 1999) as d D d (2.5) EP s kl kl EP where D kl is an elasto-platic modulus, d is the eective stress increment and increment o soil skeleton. For the pore water, the material is elastic luid as s d kl is the strain p K ii (2.6) where liquid. K is the volumetric elastic coeicient o the pore water, ii is the volumetric strain o 2.3. Equation o motion and continuity equation The equations o motion or the whole mixture are deined as s u i bi (2.7) x j where is the mass density o the mixture, s u i is the acceleration vector o the solid phase, b i is

3 the body orce vector and is the total stress tensor. Incorporating the constitutive equation or pore water, p K ii ( K : volumetric elastic coeicient o liquid, ii : volumetric strain o liquid) into the above equation leads to the ollowing continuity equation or the liquid phases: k s p s n ui bi ii p r 0 (2.8) xi w xi K where w is the unit weight o water and k is the permeability coeicient Compressibility o the pore water Following Okusa(1976), we use a relationship between saturation K o the water as Sr and volumetric elastic modulus m 1 S 1 m S (2.9) r w w0 r Pmg K where K =5880kPa or the saturation o 98%, m w is compressibility o water considering a air bubble, m v0 is a value o mw when saturation is 100%, P is an absolute pore pressure. mg 3. RESPONSE OF SAND GROUND DUE TO TSUNAMI INUNDATION DEPTH CHANGE 3.1. Analysis Conditions We assume a ground model at Gobo city o Wakayama preecture in Japan where the huge tsunami is assumed by the expected Tokai-Tounankai-Nankai earthquake. The ground model is shown in Figure 3.1. with a loose sandy layer ollowed by the impermeable clay layer. Several ground are used with dierent N1 value (converted N value; Japan Road Association, 2009). N1 =5 is the most loose sandy soil. The material parameters are listed in Table 3.1. The liqueaction strength curve is shown in Figure 3.2. The soil parameters have been determined by itting the liqueaction strength curve shown in Figure 3.2. Figure 3.1. Ground model

4 The tsunami inundation depth change as the input data o this analysis is shown as a broken line in Figure 3.4..This wave proile has been computed at the distant o 1.0km on the land area in Figure 3.3. by the tsunami inundation simulation. The sine wave has been imputed rom a let hand side boundary shown in Figure We used a sine wave or simplicity and the amplitude is 5.0m and the period is 20 minutes. Cyclic shear stress ratio N1=5 N1=10 N1= Number o cycles Figure 3.2. Simulated liqueaction strength curve Table 3.1. Material parameters or sandy soil Parameters N1=5 N1=10 N1=20 Initial void ratio e Compression index λ Swelling index κ Int.shearcoeicient ratio G 0 /σ m Permeability k (m/s) Density ρ(t/m 3 ) Stress ratio at PT M m Stress ratio at ailure M Hardening parameter B Hardening parameter B Hardening parameter C Elastic modulus o water K (kpa) Quasi-OCR OCR Dilatancy parameters D 0,n 1.0, , , 3.0 Plastic re. strain γ re P Elastic re. strain γ re E Figure 3.3. Topographic model

5 3.2. Numerical Results From the analysis, we have obtained a time proile o the eective stress decreasing ratio (ESDR) in Figure 3.4. at depths o 1m and 2.5m. The ratio is deined as ESDR (3.1) m 1 m0 where is the initial value o the mean eective stress. ESDR indicates the degree o mo liqueaction. From Figure 3.4., we can see that the eective stress increases due to the increase o the water depth associated with the progress o tsunami at 1 stage illustrated as 1 shown in Figure 3.4. The at the second stage as 2 shown in the igure, The eective stress decreases associated by the seepage low induced by the gradient o the pore water pressure even at a depth o 2.5m.Then the eective stress gradually decreases but almost become constant. Table 3.2. shows the maximum value o ESDR or dierent N1 value. This table indicates that the relatively loose sandy soil, i.e., N1 is less than 10, the ground deteriorates below at a depth o 1m. 4. SENSITIVITY ANALYSIS OF TSUNAMI INUNDATION HEIGHT AND PERIOD In this chapter, we will examine the eect o tsunami waveorm on the soil deterioration Analysis Conditions We assume that the ground is loose sandy ground o N1=5 and the change o water depth associated with the tsunami wave is modeled by the simple sine wave. In particular eects o the period T and the inundation depth H are studied as: H(m)=5.0,10.0,15.0,20.0 and 30.0, and T(sec)= 5,10,20,30,40,50 and 60. Figure 3.4. ESDR at depths o 1m and 2m o the ground (N1=5) Table 3.2. Maximum value o ESDR or the ground with several N1 value Material parameters N1=5 N1=10 N1=20 The element o ground surace The element o GL.-1.0m The element o GL.-2.5m

6 4.2. Numerical Results The model ground is presented in Figure A relationship between the maximum value o ESDR and the inundation depth H is shown in Figure From this igure it is seen that ESDR depends on the inundation depth at all depths. Near the surace, the maximum value is large, i.e., at a depth o 1m the ground is ully liqueied or the inundation depth o 30m. But at the larger depth, the maximum value is not so large, i.e., soil is not deteriorated. In Figure 4.3., a relationship between the period o wave and the maximum value o ESDR. Maximum value o ESDR becomes large with increase o the period but becomes constant over the period o 20 minutes. Figure 4.1. Analysis model Figure 4.2. The relation between the maximum value o eective stress decreasing ratio and the tsunami inundation depth

7 Figure 4.3. The relation between the maximum value o eective stress decreasing ratio and the tsunami period (H=5m,GL-2.0m element) 5. CONCLUSION The ollowing conclusions have been obtained rom the present study. 1) The large scale o tsunami induced by huge earthquake such as the 2011 Tohoku earthquake in Japan may cause a deterioration o ground through the compressibility o the pore water. 2) Near the ground surace, the eective stress decreases at a level o 44% o the initial eective stress and this eect propagates downward. For the analysis conditions taken in this paper, it is probable that the relatively loose sandy soil, in which the converted N value (N1) is less than 10, the ground deteriorates below at a depth o 1m. 3) The soil deterioration increases in proportional to the amplitude o the tsunami and the eect becomes larger or the wave with the longer period although it becomes saturated over 20 minutes. REFERENCES Goto, C., Ogawa, Y., Shuto, N. and Imamura, F.(1997). Numerical Method o Tsunami Simulation with the Leap-Frog Scheme (IUGG/IOC Time Project) IOC Manual (UNESCO, PAaris), 35. Japan Road Association.(2009). Speciications or highway bridge, earthquake resistant design V, Liu, Z., D. S. Jeng, A. H. C. Chan and M. Luan(2009). Wave-induced progressive liqueaction in a poro-elastoplastic seabed: A two-layered model, Int. J. Numer. Anal. Meth. Geomech., 33, Oka,F., Ohmaki,M. Kamata,A.,Yashima,A. and Miura,K.(1994).Case study o the breakwater due to wave motion,north-east o Hokkaido, Honporoto ishery port-, Soil Engineering Symp., JSSMFE,39, (in Japanese). Oka, F. et al.(2009). LIQCA research and development group, Material or LIQCA2D09(2009 released Ver.( Oka, F., Yashima, A., Tateishi, A., Taguchi, Y. and Yamashita, S.(1999). A cyclic elasto-plastic constitutive model or sand considering a plastic-strain dependence o the shear modulus, Geotechnique, 49(5), Okumura,Y.,Kato,R. and Oka,F.(2010). Analysis o the deterioration o ground considering liqueaction due to various types o tsunami waves, Proc. 13th JSEE, (in Japanese). Okusa, S.(1976). Wave-induced stresses in unsaturated submarine sediments, Geotechnique, 32(3), Sassa, S., Sekiguchi, H. and Miyamoto, J.(2001). Analysis o progressive liqueaction as a moving boundary problem, Geotechnique, 51(10),