SciELO - Scientific Electronic Library Online

 
 número20Presas de relaves de arena de gran altura: desafíos principales índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Servicios Personalizados

Revista

Articulo

Indicadores

Links relacionados

  • En proceso de indezaciónCitado por Google
  • No hay articulos similaresSimilares en SciELO
  • En proceso de indezaciónSimilares en Google

Compartir


Obras y proyectos

versión On-line ISSN 0718-2813

Obras y Proyectos  no.20 Concepción dic. 2016

http://dx.doi.org/10.4067/S0718-28132016000200001 

 

Comparison of mean shear wave velocity of the top 30 m using downhole, MASW and bender elements methods

Comparación del promedio de la velocidad de onda de corte en los primeros 30 m usando ensayos downhole, MASW y bender elements

 

Ricardo Moffat1, Nicolle Correia2 and Cesar Pastén2

1 Faculty of Engineering and Sciences, Universidad Adolfo Ibáñez, Diagonal Las Torres 2640, edificio D, Peñalolen, Santiago, Chile, ricardo.moffat@uai.cl
2 Department of Civil Engineering, University of Chile, Blanco Encalada 2002, Santiago, Chile, nicolle.correia@gmail.comcpasten@ing.uchile.cl


Multichannel Analysis of Surface Waves MASW tests were performed in different seismic stations where boring information and downhole tests were available. Active MASW tests were performed using 12 geophones of 4.5 Hz of frequency repeating 5 tests in each location. From the readings, dispersion curves were obtained using a f-k analysis with the software Geopsy. The shear wave velocity VS profiles were obtained by inverting the deduced dispersion curves. Downhole tests were analyzed using the direct approach in four stations (Maipú, Peñalolen, Casablanca and Melipilla), and using existing Vs results for Llolleo site. VS profiles obtained from MASW and downhole tests are compared and the average shear wave velocity of the top 30 m (VS30) calculated for each station. Vprofiles obtained from downhole and MAS Wactive tests are similar up to 30 m. Therefore, a good testing methodology and analysis of the MASW data allows reliable results and the same seismic classification of the soil. The major differences were found where there is a large impedance of two layers of soils such as it was found in Melipilla site.

Keywords: shear wave velocity, MASW, downhole, seismic classification


Ensayos geofísicos MASW (Multichannel Analysis of Surface Waves) son ejecutados en diferentes estaciones sísmicas donde existe información de estratigrafía de sondajes y ensayos del tipo downhole. Ensayos activos MASW se ejecutaron usando 12 geófonos de 4.5 Hz de frecuencia repitiendo 5 ensayos en cada ubicación. Se obtienen las curvas de dispersión usando un análisis f-k con el software Geopsy. El perfil de velocidad de onda de corte VS se obtiene de la inversión de dichas curvas de dispersión. Los resultados de ensayos downhole son analizados mediante el método directo en cuatro estaciones (Maipú, Peñalolén, Casablanca y Melipilla), además se usan datos mostrados en estudios anteriores para el sitio de Llolleo. Los perfiles de VS obtenidos de ensayos MASW y downhole son similares hasta los 30 m de profundidad. Por lo tanto mediante un ensayo y metodología de análisis adecuados de MASW se puede obtener resultados adecuados de perfiles VS y la misma clasificación sísmica del suelo que con el ensayo downhole. La mayor diferencia se observan en sitios donde existe una gran impedancia de dos capas de suelo tal como se observó en Melipilla.

Palabras clave: velocidad de onda de corte, MASW, downhole, clasificación sísmica


 

Introduction

Non-invasive geophysical methods have been used to estimate the shear wave velocities of soil layers in order to interpret physical properties such as stiffness and liquefaction potential of the soil. Applications using the surface waves started in the fifties with the geophysical method Steady State Rayleigh SSRM. In particular, surface wave methods (using active or passive source) have been applied since the 1980s (Nazarian and Stokoe, 1986) and still being broadly used nowadays (Everett, 2013). In the eighties, the Spectral Analysis of Surface Waves SASW was introduced to obtain the VS profile of the subsoil. Multistation approach become widespread in the late 1990s and today MASW (Multiple Analysis of Surface Waves) is the most used approach.

It is important to notice that seismic waves travel through rock and soil following paths that depend on the density, stiffness, and stratification of soils. Waves that pass through these materials are compressive waves P and shear waves S. P-waves can travel through any material that supports compressional forces (solids and liquids) whereas S-waves can only travel through materials that have shear strength therefore they are not transfer in liquids or gases. Soils P- and S-wave velocities, VP and VS, respectively, depend mainly of the type of soil, their stiffness and density.

Additionally, on the soil surface there are other types of waves; Love and Rayleigh waves. These waves travel along the surface with an amplitude that decreases with the depth (Biot, 1962). Rayleigh waves are commonly used on practice. Similarly, analyzing Rayleigh waves velocity VR, it can be obtained a relationship between VR, VS, and the Poisson’s ratio u or between VR, VS and Vp as follows:

Details on this derivation can be obtained in Foti et al. (2014). When considering the seismic design of structures it has been studied the seismic bearing capacity (Tiznado and Paillao, 2014) and the seismic response of the soil that could induce amplification factors on the loads that affects the structure. Seismic response of soil deposits depends between other factors on the soil stratification that exist in a specific location. Shear wave velocity profiles have been used to predict the behaviour of a soil during an earthquake. Different site classification systems use a representative average of the shear wave velocity up to a depth of 30 m (Dobry et al., 2000). The depth is basically determined by the usual depth of borings and the cost of these. Obtaining this value allow us to categorize the soil in different classes that define different seismic response.

Seismic codes generally use VS30 to characterize "site effects" and to select an appropriate response spectrum according to soil characteristics. The value of VS30 is calculated using (2) as follows:

where n is the number of layers up to 30 m depth and h and VSi are the thickness and shear wave velocities of these layers, respectively. There are many other factors that have been found to be important in the seismic response that are not always considered, among them, the impedance ratio between surface and underlying deposits and the depth to bedrock.

Different in situ seismic techniques have been developed to determine shear wave velocity profiles. In Chile and in many parts of the world, it has become very popular the use of surface seismic methods such as MASW (Multichannel Analysis of Surface Waves). However, as the analysis of the disperse nature of surface waves is complicated, there are doubts whether the method is able to characterize a site up to 30 m depth measuring only from the surface. The MASW method consists in three main steps; (1) Data acquisition on the field, (2) Dispersion analysis of the data, and (3) Inversion to obtain VS profile. This test can be performed using an active or passive source. The active source usually consists in a hammer blow or a heavy weight drop on the surface. The maximum depth of VS that can be achieved using this method varies between 10 to 30 m depending on the site and the type of active source employed (Park et al., 1999). The maximum depth that reaches the method also depends of the spread length used by the geophones D, defined as the distance between the source and the further geophone, because this determines the maximum wavelength of the Rayleigh wave that can be measured (see Figure 1). In general, it is considered that zmax D. On the other hand, the minimum depth zmin, is determined by the separation between geophones, dx, as it is shown in Figure 1. It is considered that zmin 0.5λmin = 0.5cmin/ƒmax, where cmin and λmin are phase velocity and wavelength, respectively, corresponding to a particular frequency ƒmax. It is known that the receiver spacing affects the maximum wavenumber, which corresponds to the minimum wavelength (shallow layers). The minimum wavelength can be as low as the receiver spacing (λmin = dx). Additional details of this method and the main factors that influence the measurements can be found in Park et al. (1999) and Foti et al. (2014).

 

Figure 1: Scheme of MASW setup on the field

 

Geophones with natural frequency of 4.5 Hz are used to measure the vertical component of the surface waves. The time of sampling is in the order of 1 to 2 seconds, the latter is used in the case of low shear wave velocities. The testing parameters typically advised for this type of measurements are shown in Table 1. These parameters would allow to measure depths zmax in the order of 20 to 30 m.

 

Table 1: Typical values for parameters setup in active
MASW (Park et al., 2002)

 

As it was mentioned previously, during the signal analysis there are two main processes that have to be performed: (1) Finding the dispersion curve, and (2) Inversion process.

The dispersion curve is found directly from the measurements of the geophones used on the field. In Figure 2 it is shown, as an example, the data for the Casablanca site after 5 blows in the ground. The dispersion curve is the most critical step to generate an adequate shear-wave velocity profile. In this work it has been applied a total of 5 blows in each site and the data has been stacked together to characterize a clear dispersion curve. From this figure, the representative curve that will be used in the next steps of the analysis is obtained. The analysis of this dispersion curve is obtained by using the software Geopsy.

 

Figure 2: Dispersion curve using MASW in
Casablanca site; stacking of five blows.
Magenta indicates the maximum spectral amplitude.

 

Inversion of the curve obtained in the previous step is a mathematically complicated process. It is required the estimation of Poisson’s ratio and density of the subsoil. It is an iterative process where a starting shear wave profile is established and the corresponding theoretical dispersion curve is deduced for that VS profile. This is performed many times comparing the obtained dispersion curve with the measured curve on the field and using a least-squares approach to estimates how adequate is the VS profiles adopted. In this form, it is deduced the VS profile that generates the dispersion curve that is most similar to the measured curve on the field (through the assignation of a fitting error for each curve). More details of this process can be found in (Penumadu and Park, 2005).

In this work, results of shear wave velocities obtained from laboratory tests using bender elements are also used. Bender elements have been used since the late 1970s, for example Shirley (1978) and Shirley and Hampton (1978). These elements consist in a piezoceramic transmitter, which generates shear waves through the soil sample, and another piezoceramic that acts as a receiver and converts the shear movement in an electrical signal. The time from the shear wave to travel from the transmitter to the receiver is measured to calculate the shear wave velocity as the soil sample length is known. A detailed explanation for the use and interpretation of bender elements reading to deduce shear wave velocity is given in Lee and Santamarina (2005) and Leong et al. (2005).

The objective of this work is to compare VS values obtained from MASW method and downhole tests in five sites. Shear waves velocity obtained from laboratory tests on undisturbed soil samples are also compared with the Vprofiles deduced from MASW.

Experimental setup and procedure

Sites were chosen as there were previous measurements of downhole testing performed in these locations. The locations form part of a Chilean network of seismometers that recorded the 2010 Chile earthquake (Saragoni et al., 2010). Stratigraphic information is also available at these site as boring were performed and will be shown for each location. MASW tests were performed using a Geode seismograph (from Geometrics), as shown in Figure 3a. This equipment has 24 channels of 24 bits each one with the main characteristics shown in Table 2.

 

Figure 3: MASW field setup, a) seismograph, laptop and peripherals and b) impact load

 

Table 2: Main characteristics of the equipment used on the field

 

The 12 geophones used corresponds to the model GS-11D/PC-803 with a natural frequency of 4.5 Hz, damping of 50%, cut off frequency of 80 Hz and a sensibility of 0.7 V/in/sec. The active source used in this research, corresponds to the Standard Penetration Test SPT weight that is 63.5 kg. This weight is released from an approximate height of 1.3 m as it is shown in Figure 3b. For measurements performed in Peñalolén, Maipú, Melipilla and Llolleo, it was used a source offset X0 of 8 m, and geophone distance dx = 5 m. In Casablanca, due to space constrictions, the source offset was 10 m and the geophone distance dx = 4 m. In all locations 5 blows were performed to obtain the dispersion curve of the site.

Downhole tests were performed as part of the geotechnical study of the Chilean network of seismometers (Boroschek et al., 2012). Tests were performed taking measurements every 2 m of depth, and using a three directional geophone. Downhole results are analyzed according to the interval method (Kim et al., 2004). In Llolleo, there is a downhole test published nearby reported by Verdugo (2009). Additional details on how the downhole tests are performed can be found in ASTM (2014).

Test results

In total, 25 MASW tests were performed during this study, these are complemented by the analysis of 5 downhole tests and the stratigraphy found in each boring performed previously. All MASW tests were carried out as close as possible to the location of the boring. Shear wave velocity profiles were deduced directly from the downhole tests and through the MASW equipment. Differences in the Vprofile for each site are analyzed in terms of the stratigraphic profile and the values of VS30 deduced. The results obtained by the surface methods are compared with downhole tests. Additionally, results are complemented by existing bender elements tests on soil samples obtained from the field (Boroschek et al., 2012). Tests are performed on reconstituted specimen or undisturbed samples depending on the type of soil/rock tested. More information about obtaining shear wave velocities from bender elements can be found in Lee and Santamarina (2005) among others.

Stratigraphy

The stratigraphy for each location was informed by the University of Chile (Boroschek et al., 2012). Stratigraphy is performed based on observations of samples obtained from drilling to depths larger than 30 m, but informed here only up to 30 m that is the interest zone for estimating VS. In Figures 4 to 6 the different stratigraphy of each site is shown.

 

Figure 4: Soil stratigraphy of Peñalolen and Melipilla

 

Figure 5: Soil stratigraphy of Maipú and Llolleo

 

Figure 6: Soil stratigraphy of Casablanca and symbols

 

Downhole testing

Downhole tests were performed every 2 m in each site. An example of waveforms obtained at Melipilla site is shown in Figure 7. Using these waveforms two independent interpretations of the arrival time for shear waves were performed in each site. This generates two VS profiles up to 30 m that have user depending variability.

 

Figure 7: Waveforms from a downhole test at Melipilla
(Boroschek et al., 2012)

 

Field shear wave measurements

VS profiles obtained from MASW tests at each site are shown in this paper. Dispersion curves are first shown and it is estimated the maximum depth that it is advisable to consider in each case based on the dispersion curve obtained. Then it is shown the VS profile obtained by using MASW data with the Geopsy software and the downhole deduction of VS by two different users. Also, when available, VS obtained from undisturbed or reconstituted (at the same density) specimens, are shown. These tests are performed in a triaxial cell under similar effective stress that has the soil at the corresponding depth in the field.

Maipú site

In Figure 8, it is shown the dispersion curve obtained after 5 blows on the field. In Figure 9, it is shown the dispersion curves adopted for each software. Finally, Figure 10 shows the estimate of VS profiles for MASW, downhole tests and one soil specimen reconstituted and tested with bender elements under a similar confining pressure as it is expected on the field. All the deduced profiles are very similar to each other. The larger difference is observed between 2 and 8 m deep where the downhole test gives a higher estimation of shear wave velocity. In this zone there is a dense sandy-gravel that is not detected in detail by the MASW analysis.

 

Figure 8: Dispersion curve obtained at
Maipu site; stacking of two blows

 

Figure 9: Dispersion curve used at Maipú site

 

Figure 10: VS profiles at Maipú site

 

On the other hand, the deduced shear wave velocity using bender elements in the reconstituted sample (reconstituted to the same density observed on the field) is lower, about 65% of the VS obtained by MASW and downhole tests. This is believed to be due to the loss of structure, cementation or aging effect due to sample reconstitution.

Considering expressions (2) and (3) (Park et al., 1999), it can be deduced the maximum depth and minimum definable thickness of the shallower layer, according to the frequencies and velocities derived from the dispersion curve (first mode of vibration). According to Stokoe et al. (1994):

where Hmin is the minimum definable thickness of the shallower layer, λmin and cmin are wavelength and phase velocity corresponding to a ƒmax. According to Rix and Leipski (1991):

where c1 is the phase velocity for the frequency ƒ1. For the Maipú site these values are: zmax = 29.4 m and Hmin = 5.3 m. Therefore, it can be confirmed from the measurements performed and the dispersion curves obtained from those measurements that it is possible to obtain the VS profile up to approximately 30 m deep and the best "resolution" of these measurements are in layers of 5.3 m.

Peñalolen

Similarly, at Peñalolen site it was derived the shear wave velocity from MASW (see Figure 11), downhole, and bender elements tests were performed in undisturbed soil obtained from that depth. The result of a bender elements test, considering the expected effective stress on the field, of an undisturbed sample, gives a very close value of V(related to downhole and MASW) at the same depth where the soil sample was extracted. VS results from downhole and MASW tests are very close up to 30 m.

 

Figure 11: Dispersion curve for Peñalolén site;
stacking of four blows

 

It is important to mention that according to the dispersion curve estimated it is possible to consider the maximum depth for using MASW tests equal to 17.9 m and a minimum thickness of 3.6 m. However, the method compares well with downhole tests from 20 to 30 m, it could be due to the small change in VS existing at those depths (see Figure 12).

 

Figure 12: VS profiles deduced at Peñalolén site

 

Casablanca

In this site the dispersion curve obtained from the 5 blows on the field is shown in Figure 13. The dispersion curve obtained is clear in a wide range of frequencies this allows us to estimate the VS profile up to depths of 37.9 m in this case and minimum thickness of 2.4 m.

 

Figure 13: Dispersion curve for Casablanca site;
stacking of five blows

 

The shear wave velocity profiles are very similar up to approximately 22 m where some dispersion starts to be observed between MASW and downhole methods. Shear wave velocity VS, deduced on undisturbed samples gives a very close value to the measured on downhole tests at 22 m deep where firm clay was observed on the borings on the field. A lower value than deduced from the field values is obtained on bender elements tests where the soil was observed to pass from firm clay to medium dense sand (see Figure 14).

 

Figure 14: VS profiles obtained at Casablanca site

 

Melipilla

In this site, there is a considerable change of soil type and rigidity at 6 m. Up to 6 m it can be found layers of sandy clay and clay, then below those layers it is found gravels with boulders up to 6 inches in diameter. This creates a large impedance of the soil that makes difficult to interpret the tests results in both cases downhole and MASW. Bender elements tests were performed on reconstituted specimens at depths lower than 5 m and in an undisturbed rock sample at 60 m depth. Bender elements results are very close to MASW deduction of VS for depths lower than 5 m. Also there is a good agreement of VS between MASW and bender elements tests on a rock sample at depths of 60 m (VS = 1200 m/s) although according to the dispersion curve deduced from MASW, the maximum depth up to VS that can be obtained is of 25.4 m with a minimum thickness of 2.4 m.

 

Figure 15: Dispersion curve obtained at
Melipilla site; stacking of four blows

 

MASW VS profile using Geopsy shows a sudden increase at about 4.5 m. Downhole methods seem to give an average value of VS from 5 to 30 m. The VS values obtained with MASW for depths shallower than 5 m seems to correlate relatively well with bender elements (see Figure 16).

 

Figure 16: VS profiles obtained at Melipilla site

 

Llolleo

Near this site, there is a downhole test performed and informed in Verdugo (2009). Figure 17 shows the MASW profiles obtained during this research. It is possible to observe in Figure 18 a very good agreement on VS values up to 13 m, below that depth there is a considerable difference in the VS profiles obtained.

 

Figure 17: Dispersion curve obtained at
LLolleo site; stacking of three blows

 

Figure 18: VS profiles obtained at Llolleo site

 

According to the field measurements using active MASW, it is possible to deduce that the maximum depth where the VS can be obtained is of 41 m and that the minimum thickness that can be determined is equal to 3.2 m. From Figure 18 it is clear that the MASW method was not able to capture adequately the dense sandy-gravel layer existing between the 20 and 25 m deep. However, calculated values of VS30 are equal to 266 and 264 m/s for MASW and 227 m/s for downhole test (difference of about 17%). This change in VS30 values does not change the soil type in most classification systems (Dobry et al., 2000).

Conclusions

Different methods to determine the shear wave velocity in soils are compared in this paper. Even though measurements and analysis are very different, the results tend to be close between each other. If the objective of determination of the VS profile, is the seismic classification through the use of VS30, MASW and downhole tests give very similar results and no change on soil classification is expected unless the VS30 value is close to the limits for a specific seismic soil type. The measurement of shear wave velocities on undisturbed soil samples shows to be an adequate technique to confirm VS values in a particular soil depth. This confirms the robustness of the employed methods in this work and in the professional practice.

The main conclusions that are drawn from this study are the following:

• Deduction of shear wave velocity VS profiles performed using downhole and active MASW tests are very similar except in cases where there is an intermediate dense sandy-gravel layer that is not captured by MASW analysis, as in the case of Maipú and Lloleo sites.

• Shear velocities deduced from bender elements tests on undisturbed samples are very close to the shear velocities obtained from downhole and MASW tests.

• Shear velocities deduced from bender elements tests on remolded soil samples show a large difference with the values obtained from downhole and MASW tests. This is due to the loss of fabric or cementation of the soil sample.

• In general, the maximum depth that is possible to reach, using the active MASW with a weight of 63.5 kg that drops from 1.3 m, was found to be between 17.9 to 37.9 m. Data after this maximum depth has to be considered with precaution.

• The minimum shallower thickness that can be detected using active MASW was found to be in the range between 2.4 and 5.3 m. These are considered appropriate for most geotechnical projects.

 

References

ASTM (2014). Standard test methods for downhole seismic testing, ASTM D7400. West Conshohocken, PA., USA        [ Links ]

Biot, M.A. (1962). Mechanics of deformation and acoustic propagation in porous media. Journal of Applied Physics 33(4), 1482-1498        [ Links ]

Boroschek, R.L., Yañez, F., Bejarano, I., Molnar, S. and Torres, A. (2012). Geotechnical characterization University of Chile strong motion accelerograph stations. University of Chile        [ Links ]

Dobry, R., Borcherdt, R.D., Crouse, C.B., Idriss, I.M., Joyner, W.B., Martin, G.R., Power, M.S., Rinne, E.E. and Seed, R.B. (2000). New site coefficients and site classification system used in recent building seismic code provisions. Earthquake Spectra 16(1), 41-67        [ Links ]

Everett, M.E. (2013). Near-surface applied geophysics. Cambridge University Press        [ Links ]

Foti, S., Lai, C.G., Rix, G.J. and Strobbia, C. (2014). Surface wave methods for near-surface site characterization. CRC Press        [ Links ]

Kim, D.S., Bang, E.S. and Kim, W.C. (2004). Evaluation of various downhole data reduction methods for obtaining reliable Vs profiles. Geotechnical Testing Journal 27(6), 1-13        [ Links ]

Lee, J.S. and Santamarina, J.C. (2005). Bender elements: Performance and signal interpretation. Journal of Geotechnical and Geoenvironmental Engineering 131(9), 1063-1070        [ Links ]

Leong, E.C., Rahardjo, H., and Yeo, S.H. (2005). Measuring shear wave velocity using bender elements. Geotechnical Testing Journal 28(5), 1-11        [ Links ]

Nazarian, S. and Stokoe, K.H. (1986). Use of surface waves in pavement evaluation. Transportation Research Record 1070, 132-144        [ Links ]

Park, C.B., Miller, R.D. and Miura, H. (2002). Optimum field parameters of an MASW survey. Proceedings of the Society of Exploration Geophysicists, Japan, Tokyo, 22-23        [ Links ]

Park, C.B., Miller, R.D. and Xia, J. (1999). Multichannel analysis of surface waves. Geophysics 64(3), 800-808        [ Links ]

Penumadu, D. and Park, C.B. (2005). Multichannel analysis of surface wave (MASW) method for geotechnical site characterization. Proceedings of the Geo-Frontiers Conference, Austin, Texas        [ Links ]

Rix, G. J., and Leipski, E.A. (1991). Accuracy and resolution of surface wave inversion: Recent advances in instrumentation, data acquisition and testing in soil dynamics. Geotechnical Special Publication no. 29, American Society of Civil Engineers, 17-23        [ Links ]

Saragoni, G.R., Lew, M., Naeim, F., Carpenter, L.D., Youssef, N.F., Rojas, F and Adaros, M.S. (2010). Accelerographic measurements of the 27 February 2010 offshore Maule, Chile earthquake. The Structural Design of Tall and Special Buildings 19(8), 866-875        [ Links ]

Shirley, D.J. (1978). An improved shear wave transducer. The Journal of the Acoustical Society of America 63(5), 1643-1645        [ Links ]

Shirley, D.J. and Hampton, L.D. (1978). Shear-wave measurements in laboratory sediments. The Journal of the Acoustical Society of America 63(2), 607-613        [ Links ]

Stokoe, K.H., Wright, S.G., Bay, J.A. and Roesset, J.M. (1994). Characterization of geotechnical sites by SASW method. Geophysical Characterization of Sites, IBH Oxford Press, New Delhi, India, 15-25        [ Links ]

Tiznado, J.C. and Paillao, D. (2014). Analysis of the seismic bearing capacity of shallow foundations. Revista de la Construcción 13(2), 40-48        [ Links ]

Verdugo, R. (2009). Amplification phenomena observed in downhole array records generated on a subductive environment. Physics of the Earth and Planetary Interiors 175(1), 63-77        [ Links ]

 


Fecha de entrega: 5 de mayo 2016
Fecha de aceptación: 16 de septiembre 2016

 

Creative Commons License Todo el contenido de esta revista, excepto dónde está identificado, está bajo una Licencia Creative Commons