Evaluation of Liquefaction Potential by CPTU and SDMT

Transcription

1 Evaluation of Liquefaction Potential by CPTU an SDMT Ergys Anamali Doctoral School, Faculty of Geology an Mines, Tirana, Albania. Luisa Dhimitri In Situ Site Investigation, Battle, UK. Neritan Shkorani Faculty of Civil Engineering, Polytechnic University, Tirana, Albania. Darren War In Situ Site Investigation, Battle, UK. Keywors: liquefaction, Piezocone test (CPTU), Seismic Dilatometer Marchetti (SDMT) ABSTRACT: This paper eals with the evaluation of liquefaction potential of the soils at the marsh of Porto Romano, locate in the western central part of Albania, where will be constructe the Energetic Park of Porto Romano. These analyses are performe base on the ata taken from ifferent in situ testing techniques. Piezocone tests (CPTU) were carrie out uring the two ifferent site investigations (before an after the groun improvement) for the characterization of soil layers an etermination of soil properties. In aition, uring the first phase of site investigation a full seismic stuy was performe to estimate the shear wave velocity, V s, an the peak groun acceleration. During the secon site investigation two seismic ilatometer Marchetti tests (SDMT) were also carrie out very close to the CPTU locations. The paper shows the results of the liquefiability assessment by CPTU an SDMT tests carrie out uring the two site investigations. 1 INTRODUCTION Piezocone Test (CPTU) an Seismic Dilatometer Test (SDMT) are wiely recognize as valuable techniques to etermine the soils properties in site an the change of stratigraphy. But these tests are also very useful to evaluate the liquefaction potential of soils. The most use methos assess the liquefaction potential by relating the results of CPTU or SDMT with numerous empirical proceures, in orer to compare the cyclic resistance ratio, CRR with the cyclic stress ratio, CSR. This paper aims to eal with the assessment of liquefaction potential of a construction site in Albania, by means of ifferent methos base on CPTU an SDMT ata. The soils in this site, accoring the seismic stuy, are classifie as Category III after Albanian Earthquake Design Regulation KTP-N.2-89 (1989). The peak groun acceleration is specifie to be 0.39g. The methos of Robertson (2009) an Boulanger & Iriss (2014) are use to calculate the factor of safety against the triggering of liquefaction, by using the input ata from CPTU tests. Cone tip resistance, sleeve friction, the moment magnitue of the earthquake, maximum surface acceleration uring earthquake, groun water level an the unit weights are involve in the calculations. The cone resistance, q t, ajuste by a correction factor, K C, that takes into consieration the fine content an soil plasticity, employ an iterative approach to etermine CRR base on the ata taken by CPTU (Robertson & Wie 1998)The cyclic stress ratio, CSR, is estimate base on See & Iriss (1971). Base on the seismic stuy, the maximal earthquake magnitue for this construction site is 6.8. The magnitue scaling factor, MSF, is calculate accoring to consistent methos with the calculation of CSR, base on the number of equivalent uniform stress cycles an earthquake magnitue. Meanwhile, SDMT tests provie two very important parameters for evaluating the liquefaction potential, the horizontal stress inex, K D, an the shear wave velocity, V s. Evaluation of liquefaction resistance at each test epth can be obtaine from K D an V s accoring to recommene CRR-K D an CRR-V S correlations. The SDMT ata engage in these calculations inclue the horizontal stress inex,

2 K D, an hence the stress history, into CPT liquefaction correlations in orer to reuce the uncertainty of CRR estimate by CPTU. (Marchetti, 2014) At the en of this paper, the CRR will be estimate from correlations, using at the same time Q cn an K D. The evelope SDMT base methos have the potential to be a very goo complementary test to the CPTU base methos for liquefaction analysis. As long as CRR can be evaluate by a correlation base at the same time on both Q cn an K D, many uncertainties relate to liquefaction potential evaluation can be over passe. 2 SITE INVESTIGATIONS AND GROUND IMPROVEMENT 2.1 Site investigations The project consiere in this paper is the Oil Prouct Storage in Porto Romano, locate on a flat area north of Durresi city, in Albania. The area is a marsh, at an average initial level of m below the sea level. A set of oil prouct storage tankers, LPG tankers, pipelines, roas an railway system is foreseen to be constructe. The plan of the project facilities an in situ tests carrie out uring the geotechnical investigations of this project is presente in Fig. 1. Fig. 2. SPT measurements. Fig. 3. CPTU 04 measurements. Fig. 4 gives the erive parameter of fine content particles (<0.06 mm), FC base on CPTU measurements. FC lies between 5% an 15% for the layers between 2 m an 10 m epth an then increases with epth. Fig. 1. Plan of the construction site. The first investigation starte in the beginning of year 2009, incluing 12 boreholes, SPT tests an laboratory tests. Due to very poor quality ata taken by a few laboratory tests carrie out on very isturbe samples, in September 2009 was ecie to carry out 4 CPTU tests. 3 of them were carrie out own to m, until the ense layer of san was met an 1 of the them was carrie out own to m. Fig. 2 an 3 present the results of SPTs an one of the CPTU tests carrie out in that time. Fig. 4. Fine content, FC (%) erive from CPTU.

3 From CPTU tests measurements were etecte that the layers susceptible to liquefaction are loose saturate sans, silty sans or sany silts, locate between 2 m an 15 m epth. The analysis of liquefaction potential is than performe on the basis of Iriss & Boulanger, 2004 proceure taking into account firstly the fill layer at the top an some of the results are presente in Fig. 5. (Logar, 2009) groun by stone columns uner all major structures on the oil terminal. In the geotechnical report, it was recommene to install the stone columns own to m, mae by crushe stone particles with a iameter 16 mm - 32 mm (Logar, 2009). Table 2. Require an achieve cone resistance with 15% replacement ratio Depth (m) Require q c (MPa) New B q c (MPa) B with B A s = 15% B Fig. 4. Interpretation on liquefaction potential, CPTU 03. Due to the high liquefaction potential, uring the secon analysis was consiere also the effect of the installation of stone columns, by means of the replacement ratio, A s, efine as the ratio of cross sectional area of one pile with the total cross sectional area of the gri corresponing to this pile. In orer to avoi liquefaction, the installation of stone columns, shoul increase the value of cone en resistance, q c as a function of the replacement ratio, A s, as given in Table 1 (Japanese Geotechnical Society, 2006). Table 1. Increase in q c values as a function of replacement ratio A s Replacement ratio (A s ) 10% 15% 20% Increase in q c (Δq c ) 4 MPa 6 MPa 8 MPa The emonstration of effect of 15% replacement ratio on the safety against liquefaction is represente in Table 2, where are shown the values of require q c to prevent liquefaction. Also the values of the expecte new q c after executing groun improvement for achieving 15% replacement ratio for each CPTU; B3, B4, B7, B9, which procures sufficient safety against liquefaction until the epth of aroun 10 m, but is still too low at the epths below 10 m. The propose solution for safety against liquefaction was ecie to replace 15% of the 2.2 Groun improvement In orer to prevent liquefaction an ifferential settlements of oil storage tankers an other facilities of the project, the first phase of groun improvement consiste in the construction of an embankment, with the height of 1 m, in the area of railway system an access roas. In the area of tankers an other oil proucts storage places, in orer to prevent the liquefaction the stone columns were execute, with iameter of 80 cm, installe until 14 m of epth, measure from the surface of the embankment. The axial istance between them is 1.8 m below the oil tankers an 2.0 m below the area of the LGP tanker. To accelerate the consoliation process in the area below the tankers, another system of vertical wick rains was execute, that was installe until 24 m below the embankment surface. The prefabricate vertical wick rains have a with of 10 cm an thickness of 0.5 cm an they are place in site in the same axial istance as the stone columns. The water coming out from the vertical rainage system is collecte by a horizontal rainage system an sent out of the construction site. At the en of this phase of groun improvement, 4 CPTU tests an 2 SDMT tests are carrie out, in orer to control the results of the first phase of groun improvement an also to ecie for the secon an the thir phase of groun improvement, by aing other 3 m an 4 m of filling material, respectively, over the ol embankment of 1 m. 2 CPTU tests, CPTU 04 an CPTU 06 are use in this paper to assess the liquefaction potential at the en of the first phase of groun improvement. The

4 results of these tests are presente in Fig. 5a an Fig. 5b. Fig. 6b. SDMT 10 measurements. Fig. 5a. CPTU 04 measurements. 3 METHODOLOGIES FOR ASSESSING LIQUEFACTION POTENTIAL 3.1 CPTU Base Methos Fig. 5b. CPTU 06 measurements. 2 SDMT tests, SDMT 03 an SDMT 10 are use in this paper to assess the liquefaction potential at the en of the first phase of groun improvement. The results of these tests are presente in Fig. 6a an Fig. 6b. Fig. 6a. SDMT 03 measurements Overall Approach The Robertson (2009) an Boulanger & Iriss (2014) methos use a eterministic relation expresse as a factor of safety, FoS, as the ratio of CSR to CRR multiplie to the magnitue scaling factor, MSF, as given in Eq. (1) below: CRR FoS = MSF (1) CSR where, FoS = factor of safety; CRR = cyclic resistance ration, CSR = seismic inuce cyclic stress ratio. If FoS > 1, no liquefaction is expecte CSR Calculations Both methos estimate the CSR base on the simple approach propose by See & Iriss (1971) given in Eq. (2) below: τ av vo a CSR = = σ max r (2) ' ' σ v 0 σ v 0 where, τ av = average cyclic shear stress; σ vo = total vertical overburen stress, σ' vo = effective vertical overburen stress, a max = maximal acceleration of the groun surface in horizontal irection; r = stress reuction factor, 0.65 = factor use to convert peak groun cyclic shear stress ratio to a cyclic stress ratio, representative of the most significant cycles over full uration of loaing. The calculation of r iffers between Robertson (2009) an Boulanger & Iriss (2014) methos as follows: - Robertson (2009) metho for liquefaction evaluation uses the r value accoring Liao an

5 Whitman (1986), which is estimate using epth epenent relationships, given in Eqs. (3) below: r = z, if z < 9.15 m (3a) r = z, if 9.15 < z <23 m (3b) r = z, if 23 < z < 30 m (3c) r =0.5, if z > 30 m (3) - Boulanger & Iriss (2014) metho for liquefaction evaluation uses the value of r accoring to Iris (1999), which is estimate as a function of epth an earthquake magnitue given in Eqs. (4) below: ln( r ) = α ( z) + β ( z) M, if z 34 m (4a) z α ( z) = sin (4b) z β ( z) = sin (4c) where, z = the epth below the groun surface CRR Calculations Both methos estimate CRR base on cone resistance. Boulanger & Iriss (2014) has been evelope for clean sans with FC < 5%. If the FC > 5%, a correlation factor is engage an the correction becomes constant if FC > 35%. But Robertson (2009) uses the correction factor for fine contents an plasticity inex of soils. Both methos imply iterative functions to etermine CRR. The CRR for earthquake magnitue 7.5 accoring to Robertson (2009) consiers the initial stress exponent n = 1.0 an calculates the normalize cone resistance, Q tn, normalize friction ration, F R an soil behaviour type inex, I c. The stress exponent is calculate by using Eq. (5) below, which is involve in the calculations of overburen correction factor of penetration, C N, as given in Eq. (6) (Robertson 2009) ' σ = n I v 0.15 c P, n 1.0 (5) a C N n P a = σ ' v0 An iteration until the change in n prouces Δn 1 is require. Depening on the values of I C, which varies from 1.64 until 2.70, the correction factor for soil (6) plasticity, fines content, mineralogy, soil sensitivity, age an stress history, K C is calculate, which is than invlove in the calculations of Q tn,cs an CRR 7.5 by using Eq. (7) an (8). (Robertson 2009) Q = tn, cs K C Q tn 3 Q, = 93 tn cs CRR , if 50 Q tn,cs 160 (8) If I C 2.70, than CRR is calculate by Eq. (9) below: CRR = Q K (9) tn α where, K α = correction factor to account to static shear stress. The values of CRR for earthquake magnitue 7.5 accoring to Boulanger & Iriss (2014) is calculate also base on the value of cone resistance, q c. In this metho is also involve in the calculations the C N, calculate as follows by Eq. (10): (7) β P C = a 1.70 ' (10) N σ v where, β = ( q ) c1n (10a) where, q = C q c1 N N c At this point, an iteration between C N an q c1n is require. (Boulanger & Iriss 2014) The erive value of FC is use to calculate the equivalent clean san ajusments, Δq C1N by using Eq. (11). q C1N q = ( c N ) exp( FC + 2 FC + 2 (11) The equivalent clean san of CPTU is calculate by using Eq. (12) below: = q q (12) qc 1N, cs c1n c1n Therefore, CRR 7.5 is calculate by Eq. (13) (Boulanger & Iriss 2014). 2 q q c1n, cs c1n, cs CRR = exp (13) q q c1n, cs c1n, cs

6 3.1.4 Magnitue Scaling Factor, MSF The magnitue scaling factor, MSF is use to scale the CSR estimate for a magnitue of 7.5 for the esign earthquake magnitue. Accoring to Boulanger & Iriss, MSF is calculate by using Eq. (14) (Iriss 1999) M MSF = 6.9 exp( 0.058) (14) 4 Accoring to Robertson metho, MSF is calculate by using Eq. (15) below (You & Iriss 2001): 1.74 MSF = (15) M 2.56 where, M = the moment magnitue of esign earthquake. MSF an r must be etermine by using consistent methoology. 3.2 SDMT Base Methos K D - CRR correlations Robertson (2012) propose the latest correlation between K D an CRR, obtaine by replacing Q cn with 25K D. Q cn - CRR correlations is calculate using Eq. (16): 3 ( K ) CRR = 93 + (16) D Accoring to Iriss & Boulanger (2006), CRR is calculate using Eq. (17) (Marchetti 2014). 2 Q Q cn + cn CRR = exp (17) Q Q cn cn where, V S = measure shear wave velocity, p a = atmospheric pressure ( 100 kpa), σ' vo = initial effective vertical stress in the same units as p a, V * S1 = limiting upper value of V S1 for liquefaction occurrence, assume to vary linearly from 200 m/s for soils with fines content of 35 % to 215 m/s for soils with fines content of 5 % or less, K a1 = factor to correct for high V S1 values cause by aging, K a2 = factor to correct for influence of age on CRR. 3.3 CRR base on Q cn an K D The CRR is estimate from two one - to - one correlations, from Q cn or K D, or it can be estimate by these two parameters as given in Eq. (19), as an geometric average of these two values of CRR: [( ) ( )] 0. 5 AverageCRR = CRRfromQ cn CRRfromK D (19) 4 RESULTS OF CALCULATIONS 4.1 Results of calculations base on CPTU ata In Fig. 7a an Fig. 7b are presente the results of CSR an CRR, which are calculate by using the measurements of CPTU tests in site. Fig. 7a. CSR an CRR results calculate from CPTU 04. where, Q cn = 25K D V s - CRR correlations The relationship CRR-V S1, for earthquake magnitue M = 7.5, is approximate by the Eq. 18: CRR 7.5 = 2 Ka 1V S K * * a2 100 VS1 Ka 1V s1 V S1 (18) 0.25 where, V S1 = V S (p a /σ' vo ) (18a) Fig. 7b. CSR an CRR results calculate from CPTU 06.

7 In Fig. 8a an Fig. 8b are presente the results of the calculate factors of safety, FoS, by using the measurements of CPTU tests. Fig. 9b. CSR an CRR calculate from SDMT 10 ata. Fig. 8a. FoS results calculate from CPTU 04 ata. In Fig. 10a are presente the results of the calculate factors of safety, FoS, by using the measurements of SDMT 03 test. Fig. 8b. FoS results calculate from CPTU 06 ata. 4.2 Results of calculations base on SDMT ata In Fig. 9a an Fig. 9b are presente the results of CSR an CRR, which are calculate by using the measurements of SDMT tests in site. Fig.10a. FoS results calculate from SDMT 03 ata. Fig. 9a. CSR an CRR calculate from SDMT 03 ata. In Fig. 10b are presente the results of the calculate factors of safety, FoS, by using the measurements of SDMT 03 test.

8 The metho evelope by Marchetti (2014), which is applie only in clean sans, is a goo complementary metho to the other methos use, because it is base on the ata measure by both tests. From CPTU 04 an SDMT 03 measurements it preicts liquefaction from 8.5 m until 13.5 m. From CPTU 06 an SDMT 10 measurements it preicts liquefaction from 2 m until 3 m an from 9 m until 13 m. Whereas, CRR-V s correlations on't provie similar estimations with CRR-K D correlation. It appears potentially superior than CRR-V S, ue to the higher sensivity of K D to relative ensity D R an to other factors that increase liquefaction resistance above all aging an stress history (Maugeri & Monaco, 2006). Greater weight in this stuy is given to CRR-K D correlations. Base on these results, it is better to use both CPTU an SDMT parameters to overpass uncertainties an to get a satisfactory reliability in liquefaction potential evaluation. Fig.10b. FoS results calculate from SDMT 10 ata. 5 CONCLUSIONS The measurements taken in site by CPTU an SDMT tests are use in this paper to calculate the liquefaction potential. The calculations presente are base on three methos. The results of CRR base on CPTU measurements are given in Fig. 7a, b. The factors of safety calculate from CPTU measurements are presente in Fig. 8a, b. Accoring the calculations base on CPTU 04 an CPTU 06 measurements, which are shown in Fig. 8a, b, from 1 m until a 3 m an from 9 m until 13.5 m, the soils are susceptible to liquefaction. CRR values calculate base on Q t are in goo agreement with the CRR values accoring to K D, as shown in Fig. 9a, b. Accoring to the results base on SDMT measurements, which are shown in Fig. 10a, b, using the relationship between CRR - K D, both methos give the same epths for the layers susceptible to liquefaction, which varies from 2.5 m until 4 m an from 9.5 m until 26 m for SDMT 03 an from 11 m until 26 m, for SDMT 10. These epths are in goo agreement for all methos use. Above 11 m the results iffer from one metho to the other for the calculations base on SDMT 10 ata. 6 REFERENCES Boulanger, R. W. an Iriss, I. M. (2014). CPT an SPT base liquefaction triggering proceures. Report No. UCD/CGM-14/01,Department of Civil an Environmental Engineering, College of Engineering, University of California at Davis, April, California. Iriss, I.M. & Boulanger, R.W Semi-empirical proceures for evaluating liquefaction potential uring earthquakes. Proc. 11 th Int. Conf. on Soil Dyn. & Earthquake Engineering & 3 3 Int. Conf. on Earthquake Geotech. Engineering, Berkeley, Japanese Geotechnical Association. (2006). Remeial measures against soil liquefaction. Taylor&Francis. Logar, J. (2009). Geotechnical report: Oil prouct terminal. Report no. E-28-09, Ljubjana, Slovenia. Koi Teknik i Projektimit (KTP-'89). (1989). Albania. Marchetti, S. (2014). Incorporating the stress history parameter K D by DMT into the liquefaction correlations. Personal Communication, October, Rome, Italy. Maugeri, M. an Monaco P. (2006). Liquefaction potential evaluation by SDMT. Proc.2 n Int. Conf. on the Flat Dilatometer, April, Washington DC, Robertson, P. K. an Cabal, K. (2012). Guie to Cone Penetration Testing for Geotechnical Engineering. 5 th Eition, Gregg Drilling an Testing, Inc., California. Robertson, P.K. & Wrie, C.E. (1998). Evaluating cyclic liquefaction potential using the cone penetration test. Canaian Geotech. Jnl, 35(3), See, H.B. & Iriss, I.M. (1971). Simplifie proceure for evaluating soil liquefaction potential. Jnl GED, ASCE, 97(9), You, T.L. & Iriss, I.M. (2001). Liquefaction Resistance of Soils: Summary Report from the 1996 NCEER an 1998 NCEER/NSF Workshops on Evaluation of Liquefaction Resistance of Soils. Jnl GGE ASCE, 127(4),