The Comparison of in-situ seismic and penetration test for the determination of geotechnical properties

Tez (Doktora) -- İstanbul Teknik Üniversitesi, Fen Bilimleri Enstitüsü, 1993Thesis (Ph.D.) -- İstanbul Technical University, Institute of Science and Technology, 1993Geoteknik Mühendisliğinde zemin kesitinde yer alan tabakaların mühendislik özellikleri laboratuvarda ve arazide yapılan deneyler yardımı ile belirlenebilmektedir. Laboratuvar yöntemleri elastik ve elastik olmayan davranışların incelenmesinde ve gerilme, boşluk oranı, malzeme sönümü ve deformasyon ilişkilerinin incelenmesinde parametrik çalışmalar yapılmasına imkan tanımaktadır. Ancak bu deneylerde elde edilen sonuçlar kullanılan zemin numunelerinin özelliklerine bağlı kalmaktadır. Zemin numuneleri ise örselenmenin etkisindedir ve alındıkları tabakanın küçük bir bölgesini temsil etmektedir. İlgilenilen derinlik boyunca tabaka özelliklerinin belirlenmesinde çok sayıda deney numunesine gerek duyulmaktadır. Zemin dinamik özelliklerinin yerinde yapılan deneyler ile bulunmasına olanak sağlayan Karşıt Kuyu ve Aşağı Kuyu gibi sismik yöntemler doğal koşullarda arazi parçasının daha geniş bölümünde uygulanabilmekte ve sonuçları tüm tabaka için geçerli olmaktadır. Yapay olarak üretilen sismik dalgaların belli mesafelerde gözlenmesini içeren bu yöntemler ile ölçülen hızlara bağlı olarak zemin özellikleri kolayca bulunabilmektedir. Diğer arazi deneyleri ve laboratuvar yöntemleri ile karşılaştırıldığında bu deney teknikleri önemli üstünlüklere sahiptir. Arazide dinamik zemin özelliklerinin sismik yöntemler yardımı ile belirlenmesi amacıyla Karşıt Kuyu ve Aşağı Kuyu sismik deney sisteminin kurulması, kurulan bu sistem ile seçilen çeşitli sahalarda arazi uygulamalarının gerçekleştirilmesi ve ölçülen sismik dalga hızlarının, arazi penetrasyon deneyleri ile ilişkisi incelenip aralarında korelasyon bağıntıların geliştirilmesi bu çalışmanın amacını teşkil etmektedir. Bu amaç doğrultusunda, oluşturulan bilgisayar destekli sismik deney düzeni ile üç ayrı sahada arazi uygulamaları yapılmış ve zemin kesitinde yer alan tabakaların hız profilleri elde edilmiştir. Bu sonuçlar, sondajlarda alınan numuneler üzerinde yapılan sonuçları ile karşılaştırılmıştır. Ayrıca zeminler için önemli bir özellik olan sönüm ölçümleri de gerçekleştirilmiştir. Sismik dalga hızları ile arazi penetrasyon deneyleri arasındaki ilişki incelenirken, SPT-N darbe sayısı, CPT-qc uç mukavemeti ve Dinamik sonda darbe sayısı değişken olarak alınmıştır. SPT-N sayılarının derinlik düzeltilmesi yapılmış değerleri de bir değişken olarak alınmış ve pratik amaçlar için belli bir güvenlik içinde kullanılabilir amprik bağıntılar çıkarılmıştır. Bu incelemelerde zemin tipi, derinlik ve efektif düşey gerilmenin etkisi de incelenmiş, kayma ve basınç dalgası hızını bu değişkenler cinsinden tahmin için çeşitli bağıntılar sunulmuştur. Arazide uygulanan SPT, CPT ve DS deney sonuçları arasında da korelasyon bağıntıları geliştirilmiştir. In-situ testing methods involving seismic velocity measurements of wave propagation which is generated from an artificial source are widely used for determining the dynamic soil properties. Like most engineering materials, soils also behave elastically at very small strain levels. The threshold strain level where soils behave as an elastic material is about 10- 3 percent. The seismic waves generated below this threshold strain level propagate with characteristics depending on physical properties of medium such as density, elastic moduli and material damping. As a result, observation and interpretation of seismic waves may contain very relevant information about the soil properties. Shear and constrained moduli, for example, are directly related to shear and compression wave velocities, respectively. Seismic methods simply consist of measuring the time required for the waves generated from a source to travel given distances. Once the travel time and distances monitoring waves have been measured, wave velocities can be calculated by dividing distance of travel by travel time. If wave velocities are known, other elastic constants can be determined from following equations; G=(67g)Vs2 (1) M=(07g)Vp2 (2) ju=[Ö.5(Vp/Vs)2-l]/[(Vp/V8)2-l] (3) E=2G(1+m) (4) Shear Modulus Constrained Modulus Poisson's Ratio Young's Modulus Where Vs is shear wave velocity, Vp is compression wave velocity, 0 is total unit weight and g is gravitational acceleration. Shear modulus and its variation with shearing strain amplitude is a very important parameter to evaluate the response of soil and soil structures during dynamic loading. Other important dynamic soil properties include material damping and Poisson's Ratio. When a disturbance is initiated in any kind of medium, a stress wave field generated and this field contains both body waves and surface waves. Body waves propagate through the body of the soil as either compression (P) or shear (S) waves while surface waves propagate along the surface of the soil as either Rayleigh (R) or Love wave. The direction of particle motion relative to the direction of wave propagation defines the real seismic wave type. P-wave propagates in the direction of particle motion while S-wave propagates perpendicular to the direction of particle motion. If shear wave particle motion confined to a vertical or horizontal plane, it is referred to vix vertically polarized shear wave (SV) or horizontally polarized shear wave (SH), respectively. Techniques of initiating seismic waves and extracting viable information about the medium is called seismic methods in geotechnical engineering. Seismic wave measurement methods such as surface reflection and refraction have been used by geophysicists for the exploration of deposits or the depth of bedrock by measuring P-wave velocity. More recently, the surface refraction technique has been used to measure shear and compression wave velocities to determine dynamic soil properties. Although wave travel times generally are measures of distances up to hundreds of meters, when using these methods, the soil layers between bedrock and surface are considered to be little importance of a few layers with constant properties. However, when it is thought that the most civil engineering problems occur within depth of 30 meters below the surface, seismic reflection and refraction methods are difficult to apply directly to civil engineering problems. Consequently, seismic methods have been developed with which detailed soil profile characteristics can be obtain at shallow depths for engineering purposes. The requirement of more detailed dynamic soil properties necessitates the use of short distance between seismic energy source and receivers, and also the generation of identifiable waves. Among the various seismic methods used for engineering application, the Cross-Hole and Down-Hole seismic methods are most widely used and most reliable for determining dynamic soil properties, like initial elastic moduli at small strain levels. Dynamic soil properties are determined from soil samples using laboratory testing methods such as cyclic triaxial and sample shear, and resonant column or in-situ seismic testing techniques such as Cross-Hole and Down- Hole methods performed on natural conditions. Besides of laboratory methods having some advantages for performing parametric studies on the elastic and inelastic properties of soils, these methods are influenced from sample disturbance, stress relief and simulating field stress conditions in the laboratory. The results of laboratory test represent only the limited region of soil layer from which soil sample obtained. In comparison with laboratory and other in-situ tests, seismic velocity measurement methods have distinct advantages, i) They can be performed to represent larger zones of soils, ii) They have a strong theoretical basis, based on elasticity theory, iii) Measured wave velocities are independent of equipment used, so no need any correction factor, iv) Initial field stress conditions and anisotropy are automatically incorporated, v) Large zones of soil which represent the macroscopic nature of the site can be sampled. In-situ seismic methods continue to play an important role to play in geotechnical engineering, especially when dealing with low deformational characteristics of soil. Seismic wave velocity measurements are performed initially by generating waves at one point in or on the soil and measuring the required time for the direct waves of concern to travel to one or more receiver points. Wave velocity can be calculated from the distance between wave source and receiver, and travel time. In application, seismic methods are divided into two groups according to the geometrical configuration of the source and receivers just like surface and borehole seismic methods. In surface wave methods, source and receivers are on the soil surface, and having vi n economic advantage in comparison with borehole methods, because no boreholes are necessary. The steady-state Rayleigh wave method and spectral analysis-of surface-waves (SASW) methods are known as surface methods. Borehole seismic methods are typically grouped again according to the geometrical arrangement of source and receivers, such as Cross-Hole, Down-Hole, Up-Hole, In-Hole and Bottom-Hole methods. In this study, the Cross-Hole and Down-Hole seismic methods are the only borehole methods considered. In the Cross-Hole method, the time for seismic wave to travel horizontally from a source in a borehole to one or more receivers at the same depth as the receivers in other boreholes is measured. With Down-Hole method, the time for waves to travel almost vertically from a source on the surface to one or more receivers at different depths in a single borehole is measured. This study is concerned with the determination of dynamic soil properties based on the in-situ testing techniques such as Cross-Hole and Down-Hole seismic wave velocity measuring. These kind of in-situ measurements have two basic aspects. The first aspect is involved with the technique of testing at the site and obtaining the necessary data. The second one is the analytical interpretation of field data to evaluate the soil properties of concern. Initially, for the Cross-Hole methods, boreholes with spacing in the order of 3-5 m are drilled and cased with plastic casing material to the desired depth. Logging is important for each borehole during drilling to determine soil type and layering. Borehole diameters should be as small as feasible to minimize factors affecting the results of measurement such as disturbance. Typical borehole diameters range from 7 to 15 cm. The Cross-Hole and Down-Hole seismic testing system used in implementation in this study at different sites in various soil condition consists of a source to generate seismic wave, receiver with proper coupling and frequency response and a 12-channel recording device. The source used in Cross-Hole method to generate identifiable seismic wave is a mechanical source, borehole shear wave hammer, which is repeatable, reversible and rich in creating of shear wave. This hammer which is developed to be used for Cross-Hole testing in this study can be coupled to the borehole wall by hydraulically expandable plates. These plates are pressed towards to borehole wall so that a falling weight can be hit the baseplate from top and bottom. Thus, shear wave can be produced in direction perpendicular to the borehole axis. Three component velocity transducers (borehole geophones) which are housed in one case and oriented to direction of one vertical and two horizontal have been used as receivers in both Cross-Hole and Down-Hole measurements in this study. The natural frequency of the receivers is 28 Hz. With borehole wall pressing system the geophones can be fixed at optional of borehole by expanding packer tube by means of air pressure. Two types of seismographs, analog and digital, had been used as recording system during case studies. At first in-situ applications for seismic wave velocity measurement had been performed by using surface refraction method at the location field of Istanbul Technical University Campus, Ayazağa, with the analog seismographs used for recording waveform. This IX device has 12 channels and waveform are recorded directly on special recording paper which is sensitive to the sun-light, and Xenon flash is used for timing. It is difficult to utilize seismic measurement using this analog seismograph because its triggering system does not automatically and simultaneously operate with the impulse given for generation wave. In this tests, a steel plate on the soil surface and a vertical impulse with the help of a sledge hammer had been utilized for the wave source. A transducer was mounted on this plate for the triggering system of seismograph. Immediately before the impulse activates, the seismograph is operated by driving the record paper at a selected sweep rate, and waveform is recorded. But the triggering system of multichannel digital seismograph starts the recording automatically, device as soon as impulse activates. Enhancement of signal to noise ratio is accomplished by its signal enhancement and digital stacking function. There are some advantages of using this type of instrument. Lower energy impacts can be used to generate the seismic wave, and survey can be performed for much greater distances using mechanical source. By utilizing signal enhancement unit, the impact for generating the wave can be repeated a number of times, so eliminating the errors present in timing a single impulse and effects of undesirable noise in waveform traces. Continuous confirmation feasible with waveform data can be stored in its memory and waveform can be displayed on the screen and printed on the special recording paper in variable modes. Obtained data is recordable in the computer by means of RS-232 data transmission. The data are transferred to the computer from the seismograph by a transfer computer code written for this study. This computer code can also be used to analyze the signal records for determining the first arrival time. Signal analysis of seismic waveform basically involves the determination of direct travel times for compression and shear waves by identifying characteristic point from the traces. The time between this characteristic point and zero time determined by the triggering system equals to direct travel time. If first arrivals of certain wave type at more than one point from the source are known then interval time or interval velocity can be calculated. The initial compression wave arrival at each trace is identified as the first excursion, while initial shear wave arrival is identified as the first high amplitude excursion. The Cross-Hole and Down-hole seismic methods set up for this study had been widely used for measuring shear and compression wave velocities at different sites. In the tests performed at the Campus of Technical University, three boreholes had been drilled, cased and grouted. Shear and compression wave velocity profiles had been determined and initial shear modulus (Gmax) and Poisson's Ra-tio calculated from these seismic wave values, including in-situ material damping ratio. Besides of in-situ measurements, laboratory tests of ultrasound and resonant frequency were performed on soil samples obtained from the boreholes. Determination of low amplitude material damping ratio is an important field measurement. Seismic wave amplitudes decay due to both geometrical damping and material damping while propagating within the soil. For geotechnical engineering applications material damping is expressed in terms of the damping ratio as the following relation; D=[ln(A1R1/A2R2)]/{(27rtI/T)2+[ln(A1R1/A2R2)]2}0.5 (4) x Where D is damping ratio, Aj and A2 are the amplitudes of same characteristic points on waveform at distances Rj and R2 from the wave source, tj is the interval time between Rj and R2, T is the period of the wave. In order to determine the damping ratio by using this equation, it is necessary to measure the wave amplitudes at least at two receiver points from the same source. For damping measurements performed at the campus site, frequency of receivers and the coupling of two receivers are the same so that receiver output is not affected. The time domain wave amplitudes in the field were measured by using Down-Hole seismic tests. Wave amplitudes generated from the same source impulse were simultaneously observed at two receivers in the same borehole. The equipments used in damping measurements consist of a source, receivers and a recording device, as in wave velocity measurements. The damping ratio were also measured on soil samples in the laboratory by resonant frequency test. In comparison of field and laboratory measurements, it is noticed that the results of each measurement are different due to the soil sample disturbance, nonrepresentative samples and inabilities in reproducing in-situ stress conditions. After March 13, 1992 Erzincan earthquake, a detailed site investigation were carried out utilizing Standard Penetration Test (SPT), Cone Penetration Test (CPT), Dynamic Penetration Test (DPT). For determining the dynamic soil properties Down-Hole and Cross-Hole seismic wave velocity measurements techniques were carried out in cased and grouted ten boreholes. Erzincan is located on a deep alluvial deposit layer formed with debris materials transported from the Fırat River and mountains around. The North Anatolian Fault zone is north of the Erzincan basin. The basin is filled mostly with silts, sands and gravels. Borehole data indicate sand and gravel series with small amount of silt and clay, and the layers are not continuous between different boreholes. The soil profile consists of alternating layers of silty sands, sandy gravels, gravelly sands and in some locations silty sandy clays. A dense, partly cemented gravel layer at depths of 4-5 m at the north and at depths 15-20 m at the south part of Erzincan had been observed. According to the results of various in-situ tests performed in Erzincan, the seismic wave velocities variation with SPT-N value, CPT qc tip resistance and dynamic penetration test results were studied and some correlations were obtained. The SPT-N values that' are greater than 120 and less than 1 were not taken into account but, on the other hand, SPT-N values for all kinds of soils were included in the regression analysis both without any correction and with depth-corrected value. A correlation with high regression coefficient between SPT-N and shear wave velocity were obtained. In comparison with the previous studies, the correlations obtained in this study rather overestimate shear wave velocity values for SPT-N greater than 20. This difference is most likely due to the variation in SPT techniques incorporated and gravelly nature of soil layers. The correlation between shear wave and depth-corrected SPT-N value, (Nj), had been found to be lower than the between shear wave and SPT-N. On the other hand, when the effective overburden stress was taken into account, the correlation between shear wave and depth-corrected SPT-N became higher. XI Other parameters, in addition to those described herein, were thought to influence shear wave velocity correlations. These parameters include soil type, mean grain size, (D50), effective overburden stress and depth. The variation of shear wave velocity with SPT-N, overburden stress and soil type had also been studied. The soil type was divided into three groups named as clay, sand and gravel. The results of the analysis gave the highest correlation coefficient for clay while the lowest coefficient was obtained in gravel. Besides of differences in correlation coefficients, for the same SPT-N value and vertical effective stress generally the same shear wave velocity had been evaluated from the correlations for all soil types, clays and sands. In the correlation analysis, mean grain size (D50) were also be taken as an input variable and it was found that as D50 values became higher, greater shear wave velocity were observed. So it can be said that as the grain size increases, the calculated shear wave velocity become larger. This fact is justified by higher shear wave velocity values obtained from gravels. When vertical effective stress was taken into account, the coefficient of correlation between shear wave and D50 found higher than the other. The variation of shear wave velocity with CPT-qc tip resistance and Dynamic Penetration Test (DPT) blow account had also been studied and some correlation equations were established. Since qc values are measured with depth at interval of few centimeters, the average of qc values were used in the obtaining correlation between shear wave velocity and CPT qc tip resistance. Both linear and nonlinear regression analyses were performed on the data set obtained for all kind of soil types. The coefficient of correlation between shear wave velocity and - qc tip resistance was found to be improved when effective overburden stress were utilized, as in correlation between shear wave velocity and SPT-N value. The correlations equations and coefficients obtained for shear and compression wave velocities in terms of selected variables are summarized in Table 1. Also the correlation compression wave velocity and mentioned variables was studied and it was found that compression wave has lower correlations than the shear wave. The results of the correlation analysis made between equations and coefficients compression wave velocity and other variables are given in Table 2. Another correlations considered in this study the correlations among in- situ penetration tests such as SPT, CPT and Dynamic Penetration Test performed in Erzincan during field investigation. Possible correlations among penetration resistances of these tests w