Downhole seismic testing and seismic cone penetration testing (SCPT) have shown little change since the 1990’s, with essentially the same sensors, sources, test procedures and analytical methods being used. In these tests, the time differences of first-arrivals or other reference points early in the time-domain signals have been used to calculate shear and compression wave velocities in soil and rock layers. This time-domain method requires an operator to pick the first arrival or other reference point of each seismic wave in the time record. Picking these reference points correctly is critical in calculating wave velocities. However, picking these points in time records is time consuming and is not always easy because of low signal-to-noise ratios, especially in the case of shear waves which arrive later in the time record. To avoid picking reference points, a cross-correlation method is sometimes applied to determine travel times of the seismic waves, especially in traditional downhole testing. One benefit of the cross-correlation method is that it can be automated. The cross-correlation method is not, however, appropriate for evaluation other body wave characteristics such as wave dispersion and material damping. An alternate approach is to use frequency-domain analysis methods which are well suited for evaluating time changes between all types of waveforms measured at spatially different points. In addition, frequency-domain methods can be automated and attenuation measurements can also be performed. Examples of such testing procedures with Rayleigh-type surface waves in geotechnical earthquake engineering are the Spectral-Analysis-of-Surface-Waves (SASW) and Multi-Channel-Analysis-of-Surface-Waves (MASW) methods. In this research, an automated procedure for calculating body wave velocities that is based on frequency-domain analysis is presented. The basis for and an automated procedure to calculated body wave dispersion is also presented. Example results showing shear wave velocity and material damping measurements in the SCPT are presented. The objective of this study is to improve downhole seismic tests with boreholes, cone penetrometers or flat-plate dilatometers by developing a frequency-domain analysis method which overcomes many of the disadvantages of time-domain analyses. The frequency-domain method is called the Spectral-Analysis-of-Body-Waves (SABW) method. The SABW method does not require an operator to pick the first-arrival or other reference times. As a result, the shear wave velocities and wave dispersion can be calculated in real time using the interpretation method with an automatic calculation procedure, thus reducing human subjectivity. Also, the SABW method can be used to determine additional information from the dispersion curves such as the material damping ratio and an estimate of soil type based on the dispersion relationship. In this research, field SCPT measurements are presented as an example to illustrate the potential of the SABW method. Measurements with shear waves are highlighted because these measurements are most often required in geotechnical earthquake engineering studies. / text

Enhanced Integration of Shear Wave Velocity Profiling in Direct-Push Site Characterization Systems Alexander V. McGillivray 370 Pages Directed by Dr. Paul W. Mayne Shear wave velocity (VS) is a fundamental property of soils directly related to the shear stiffness at small-strains. Therefore, VS should be a routine measurement made during everyday site characterization. There are several lab and field methods for measuring VS, but the seismic piezocone penetration test (SCPTu) and the seismic dilatometer test (SDMT) are the most efficient means for profiling the small-strain stiffness in addition to evaluating large-strain strength, as well as providing evaluations of the geostratigraphy, stress state, and permeability, all within a single sounding. Although the CPT and DMT have been in use for over three decades in the USA, they are only recently becoming commonplace on small-, medium-, and large-size projects as more organizations begin to realize their benefits. Regrettably, the SCPTu and the SDMT are lagging slightly behind their non-seismic counterparts in popularity, in part because the geophysics component of the tests has not been updated during the 25 years since the tests were envisioned. The VS measurement component is inefficient and not cost effective for routine use. The purpose of this research is to remove the barriers to seismic testing during direct-push site characterization with SCPTu and SDMT. A continuous-push seismic system has been developed to improve the integration of VS measurements with SCPTu and SDMT, allowing VS to be measured during penetration without stopping the progress of the probe. A new type of portable automated seismic source, given the name RotoSeis, was created to generate repeated hammer strikes at regularly spaced time intervals. A true-interval biaxial seismic probe and an automated data acquisition system were also developed to capture the shear waves. By not limiting VS measurement to pauses in penetration during rod breaks, it is possible to make overlapping VS interval measurements. This new method, termed frequent-interval, increases the depth resolution of the VS profile to be more compatible with the depth intervals of the near-continuous non-seismic measurements of the SCPTu and the SDMT.

Apresenta-se neste trabalho a aplicação de um método para a previsão da curva tensão-recalque de fundações diretas assentes em solos tropicais arenosos a partir de resultados de ensaios de cone sísmico (SCPT). Os locais estudados foram os campos experimentais de fundações da EESC/USP - São Carlos e da UNESP-Bauru, onde existem resultados de provas de carga realizados a diferentes profundidades, assim como resultados de ensaios SCPT. As previsões realizadas apresentaram bons resultados, após ajustes dos parâmetros f e g, pois as curvas tensão-recalque estimadas foram próximas a aquelas obtidas a partir de provas de carga em placa, para as profundidades maiores que 1,5 metros. Verifica-se assim a aplicabilidade do método, após seu ajuste, para reproduzir a curva tensão-recalque neste tipo de solo, empregando uma abordagem mais racional, com menor dependência de correlações empíricas. Destaca-se nesta pesquisa que existe uma variabilidade dos resultados de ensaios SCPT e de provas de carga que está relacionada com a mudança de sucção no solo. Para o campo experimental de São Carlos foi possível ainda fazer uma avaliação da variabilidade nas previsões realizadas, pois existe maior número de resultados de ensaios de campo e provas de cargas disponíveis. / It is presented in this dissertation the use of a method for predicting the stress-settlement curve of shallow foundations on tropical sandy soils based on seismic cone (SCPT) test results. The studied sites were the experimental research sites from USP - São Carlos, and UNESP - Bauru, Brazil, where there are results from plate load tests conducted at various depths, as well as SCPT test results. The stress-settlement curve predictions show good results, after adjusting the parameters f and g, because the estimated curves were close to those obtained from plate load tests, to depths greater than 1.5 meters. The applicability of the method, after its adjustment, to reproduce the stress-settlement curve for this type of soil, was verified employing a more rational approach with less reliance on empirical correlations. It is highlighted in this research that there is variability on SCPT and plate load test results, which is related to the change in soil suction. It was also possible to access the variability on the prediction for the USP São Carlos site, since there is a greater number of in situ and plate load tests in this site.

Apresenta-se neste trabalho a aplicação de um método para a previsão da curva tensão-recalque de fundações diretas assentes em solos tropicais arenosos a partir de resultados de ensaios de cone sísmico (SCPT). Os locais estudados foram os campos experimentais de fundações da EESC/USP - São Carlos e da UNESP-Bauru, onde existem resultados de provas de carga realizados a diferentes profundidades, assim como resultados de ensaios SCPT. As previsões realizadas apresentaram bons resultados, após ajustes dos parâmetros f e g, pois as curvas tensão-recalque estimadas foram próximas a aquelas obtidas a partir de provas de carga em placa, para as profundidades maiores que 1,5 metros. Verifica-se assim a aplicabilidade do método, após seu ajuste, para reproduzir a curva tensão-recalque neste tipo de solo, empregando uma abordagem mais racional, com menor dependência de correlações empíricas. Destaca-se nesta pesquisa que existe uma variabilidade dos resultados de ensaios SCPT e de provas de carga que está relacionada com a mudança de sucção no solo. Para o campo experimental de São Carlos foi possível ainda fazer uma avaliação da variabilidade nas previsões realizadas, pois existe maior número de resultados de ensaios de campo e provas de cargas disponíveis. / It is presented in this dissertation the use of a method for predicting the stress-settlement curve of shallow foundations on tropical sandy soils based on seismic cone (SCPT) test results. The studied sites were the experimental research sites from USP - São Carlos, and UNESP - Bauru, Brazil, where there are results from plate load tests conducted at various depths, as well as SCPT test results. The stress-settlement curve predictions show good results, after adjusting the parameters f and g, because the estimated curves were close to those obtained from plate load tests, to depths greater than 1.5 meters. The applicability of the method, after its adjustment, to reproduce the stress-settlement curve for this type of soil, was verified employing a more rational approach with less reliance on empirical correlations. It is highlighted in this research that there is variability on SCPT and plate load test results, which is related to the change in soil suction. It was also possible to access the variability on the prediction for the USP São Carlos site, since there is a greater number of in situ and plate load tests in this site.