International Parallel Tests on Bender Elements at the University of Porto, Portugal

REPORT International Parallel Tests on Bender Elements at the University of Porto, Portugal Cristiana Ferreira PhD student, Research Assistant [ [email protected] ] António Viana da Fonseca Associate Professor, Director of LabGEO of FEUP [ [email protected] ] FEUP, January 2005 LABORATÓRIO DE GEOTECNIA FACULDADE DE ENGENHARIA DA UNIVERSIDADE DO PORTO Rua Roberto Frias, s/n | 4200-465 Porto PORTUGAL Tel.: 225081988/225081728; Fax: 225081835 Email: [email protected]; Website: http://www.fe.up.pt/sgwww/labgeo 1 Introduction The Geotechnical Laboratory (LabGEO) of the Faculty of Engineering of the University of Porto (FEUP) is one of the participants of the INTERNATIONAL PARALLEL TEST ON THE MEASUREMENT OF Gmax USING BENDER ELEMENTS, organized and supported by the Technical Committee TC-29 of ISSMGE. This initiative promotes the participation of institutions and laboratories worldwide currently using this state-of-the-art technique, in a series of bender element tests using the well-known Toyoura sand (from Japan), according to pre-established specifications. The aim of the International parallel tests using bender elements is to disseminate the experience of different practices (in terms of data acquisition, analysis and interpretation of results) used worldwide and to develop the recommendation for procedures referred to bender elements test. The present report describes and details the results obtained for a short series of tests carried out at the Geotechnical Laboratory of FEUP, under the specified testing procedures. 1 2 Background of bender element testing at LabGEO of FEUP The LabGEO of FEUP has been using bender elements, since 2000, when a recently acquired triaxial system equipped with bender elements and compression transducers, from ISMES- Enel.Hydro, was set up and put into operation. Subsequently, another conventional triaxial cell was locally modified and new bender and extender elements were implemented. These piezoelectric transducers originated from the University of Bristol, where significant efforts on the development of these devices were being made, under the supervision of Dr. David Nash and colleagues (Viana da Fonseca & Ferreira, 2002). These equipments have been continuously in use, mainly for research purposes. The first and most important application of this testing technique has been the comparative assessment of sampling quality on young residual soils from granite. This technique proved to show substantial differences on shear wave velocity Vs, hence shear modulus Gmax, of samples retrieved by different samplers, with distinct induced disturbances (Ferreira et al., 2004). The other research topic involving bender elements has been the issue of interpretation of the measurements. It is widely accepted that there is some uncertainty in the interpretation of the bender element traces and that however simple the transmitter wave is, a far more complex wave will be received (Moncaster, 1997). For this purpose, an extensive set of tests mainly on natural residual soils has been performed, using different interpretation methods, in order to further understand the differences in the results, towards determining which method most closely estimates the true travel time. In this framework of research, it is therefore with enthusiasm and great expectations that the LabGEO of FEUP has embraced this initiative. 3 Outline of the test apparatus The triaxial test apparatus was used for carrying out these tests. Cylindrical samples of Toyoura sand were tested, with 70mm diameter and 140mm height. The triaxial testing device, equipped with bender elements and compression transducers, was originally developed at the Italian Research Centre ISMES, Spa and later commercialized by the company Enel.Hydro. Besides the triaxial chamber itself, this system consists of a very simple pressure control panel, a function generator, an input-output amplifier with data acquisition and an oscilloscope. The confining pressure is applied by means of an air-water 2 interface, inside the chamber, where the compressed air is regulated in the control panel. Figure 1a shows this apparatus in use at the LabGEO of FEUP. 2mm 2mm 136mm 70mm 2mm 2mm 136mm 70mm a) b) Figure 1 – Triaxial apparatus at LabGEO of FEUP: a) triaxial chamber; b) scheme of the tip-to-tip distance and protrusion of the bender elements 4 Details of the bender elements In terms of the piezoelectric transducers, only the bender elements will be hereby detailed. The bender elements have been provided already mounted in the platens of the triaxial cell. Each bender element consists of two plates glued together and polarised so that, when stressed by an electrical signal, one elongates and the other contracts. According to the manufacturer, the dimensions of the transducers are about 20mm x 10mm with a thickness of 3mm. It is possible to directly measure the protrusion of the bender elements, which is approximately 2mm on each platen. Hence, the tip-to-tip distance is calculated by subtracting 4mm to the sample height, as shown in Figure 1b. As to the electrical connections, the manufacturer informs that the transducers can be used as transmitters or as receivers since they have been built in the same way. The electronic equipments used for running the bender element measurements consist of: ß A programmable function generator (TTi, Thurlby Thandar Instruments-TG1010) with several wave shapes (sinusoidal, square, ramp or other) and configurations (continuous, repeating pulses, sweep, etc.); ß An integrated unit with input and output amplifiers (specifically developed by ISMES-Enel.Hydro), which automatically amplifies 40x the input voltage and 3 enables the selection of four levels of output amplification (1x, 2x, 5x and 10x); and, ß An oscilloscope (Tektronix TDS 220), directly connected to the amplifier unit, which records the signals and enables the immediate identification of the travel time; this device is also connected to a PC and, by means of specific software, it is possible to transfer and save the obtained results under various formats, for further analysis and interpretation; ß Alternatively to the oscilloscope, a data acquisition wit spectral analysis device has been used (PICOScope ADC 216), also connected to a PC, where specifically designed software was used for acquisition and control. In Figure 2, a photograph of the transducers and a simplified scheme of the layout of the elements of the bottom platen are presented. 1. Compression transducer 2. Insulating support 3. Epoxy resin 4. Bender element 5. Silicon rubber 6. Electrical wires 7. Porous stone 8. Drainage line 9. Stainless steel bottom platen (not to scale) Figure 2 – Compression transducers and bender elements in the triaxial cell 5 Test description and procedure For this International Parallel Test series on bender element measurements, three extensive tests have been considered valid, named RR1, RR2 and RR3. The test procedure specifications provided were closely followed. Details of each test can be found in the respective test form attached. Toyoura sand samples were air pluviated, resulting on void ratios of about 0.70. All these samples were dry throughout testing. Four isotropic stress stages (50, 100, 200 and 400 kPa) were applied, according to the instructions provided. The nomenclature used to designate each test stage was “StageNumber-TestName”; for example, the third stage (200kPa) of the second test is “03-RR2”. 4 6 Interpretation of bender element measurements Time and frequency domain measurements were taken, for the application of the different interpretation methods, namely: ß First direct arrival, at various frequencies ß Discrete method of phase frequencies [π points] ß Continuous sweep input frequency [ABETS] In what follows, each of these techniques will be briefly described. 6.1 Time domain First direct arrival method, at various frequencies [time domain] The first direct arrival method is the most simple, common and usual procedure for interpreting bender element measurements. It consists on the identification of the first instant of arrival of the wave in the output signal, similarly to the techniques used in geophysical tests (namely Cross-Hole and Down-Hole). While it is sometimes easy to determine first arrival, it is often the cause of much uncertainty. For instance, Arroyo (2001) has estimated uncertainties of up to 100% in estimation of the small strain shear stiffness (Gmax). Many authors (Sanchez-Salinero et al., 1986; Viggiani & Atkinson, 1995; Jovičić et al., 1996; Arulnathan et al., 1998) have reported several sources and factors of error and inherent near field effects, which mask and compromise the identification of the arrival point. Figure 3 exemplifies a typical bender element trace, where the arrival instant can be determined within a margin of error of about 20%. t s 10 V 100 mV t s 10 V 100 mV Figure 3 – Determination of the travel time with the first direct arrival method 5 Several solutions have been proposed in order to minimize the subjectivity of this method, mainly by reducing the initial distortion of the signal. This can be achieved by changing the frequency and shape of the input wave, whilst ensuring the accomplishment of a number of technical requirements and boundary conditions. According to Jovicic (2004), these technical requirements comprise:  good electronics equipment, namely a function generator and an oscilloscope with screen;  well shielded and grounded cables;  properly connected and encased bender elements;  leak free connections;  noise free environment; Special attention should be put on design details, leaks and wear-and-tear issues. On the other hand, the boundary conditions refer to the spatial conditions (such as the alignment of the bender elements, the reflections of the wave on the edges of the sample, the near field effect or the relative distance between the transmitter and the receiver), the contact between the bender element and the soil (which might induce poor coupling) and the overshooting (at high frequencies, the bender element changes its mode shape and the response becomes very complex). In the present tests, the first arrival method was used and several input frequencies at relevant levels were applied, as well as simultaneous P- and S- wave recordings, in order to take into consideration the near field effect. Figure 4 – First direct arrival method at various input frequencies, considering near field effect 6 This procedure has evidenced the presence of near field effect, enabling the identification of the compressional component of the shear wave. It should be noted that, for all the tests carried out at FEUP, the polarity of the received wave is negative, i.e., inverted in relation to the input wave. For this reason, the first arrival point corresponds to the beginning of the first significant trough. 6.2 Frequency domain The use of continuous signals which require the shear wave velocity to be decoded from measurements of relative phase of transmitted and received signals is gaining in popularity (e.g. Blewett et al., 1999). While the technique is used very widely across a wide range of fields, Viggiani and Atkinson (1995) were the first to apply phase-delay method to bender element testing. According to Greening et al. (2003), these methods have a number of advantages over traditional pulse-based measurements. Chief amongst these is that it is convenient to create an algorithm to determine travel time by establishing the gradient of a graph of phase difference against frequency. Phase delay methods can be performed reliably using "traditional" equipment i.e. a signal generator and oscilloscope (Kaarsberg, 1975). A continuous harmonic sinusoid is used as the input signal. The frequency of the signal is changed and the frequencies at which the transmitted signal and received signal are exactly in and out of phase with one another (so called π-points) are noted. Greening and Nash (2003) showed that same information could be established less onerously using broadband (sweep) input signal and a spectrum analyser. Whereas the continuous sweep input method enables the acquisition of a continuous phase angle versus frequency relationship, for the discrete method only specific points can be observed, the π-points; hence, this is a simplified version of the more complete continuous sweep input frequency method. In the present work, for the frequency domain, two different methods were considered: the discrete method of phase frequencies and the continuous sweep input frequency. Discrete method of phase frequencies [π points] The discrete method of phase frequencies method consists on the identification of the πpoints, i.e., the frequencies to which the input and output signals are either in or out of phase. The phase angle corresponds to a measure of this relation between the input and the output signals, taking a value multiple of π, when an in or out of phase relation occurs. 7 In the absence of spectral analysis equipment, the identification of the phase frequencies is performed manually, by defining discrete points to which the phase angle is known to be multiple of π. The sequence procedure for the application of this simplified approach to the frequency domain can be divided in the following stages:  Input of a continuous sine wave, at a very low frequency, from the function generator;  Selection of a XY display in the oscilloscope;  Slow and gradual increase of the input signal frequency, until a relation between the two input and output waves is clearly visible (similar to any in Figure 5);  Registration of the frequencies values of the π-points, defined by the observation in the oscilloscope of the Lissajous figures (Figure 5), when the relation of the two waves becomes linear; this process is repeated for as long as it can be clearly observed (generally no further than 20kHz): In- and out-of-phase frequencies: π φ kf = )(  Plot of these frequency values versus the respective number of wavelengths N, which increase sequentially of 0.5 in each phase frequency; since the slope is only dependent on the relative difference between the ordinates and, as long as the sequential increase of 0.5 is maintained, the absolute value of the first N can be arbitrary; Number of wavelengths: 22 )( kf VfLL N== ⋅ == π φ λ  Determination of the average slope of the obtained plot, defined by linear regression of all, or a selection of, the points of the plot; the slope value coincides with the travel time. Wave velocity: )( 2f L ffV φ πλ ⋅=⋅= Travel time: f N V L t== (≡ slope) x y y xx yy y xx in phase out of phase a) b) c) d) e) Figure 5 – Different oscilloscope XY displays of the transmitted and received signals (Rio, 2002) 8 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 frequency (Hz) travel time (ms) 0.5 kHz 1 kHz 2 kHz 4 kHz 6 kHz first arrival first arrival Figure 14 – Time charts for test stage 04-RR1 Besides enabling an overview of the travel time in a continuous manner, the time chart also represents a new, useful and practical way of selecting the most adequate travel time, with less iterations and error. Therefore, it is clearly an improvement of the sweep method. Having been conceived in an attempt to understand the first and most complex test results, it is at this point considered a fortunate fruit of such incident. 8 Discussion of the interpretation methods The first important conclusion to be derived from this exercise is that the results from the frequency domain vary significantly for slight changes in the frequency range, meaning that the use of these methods may require an educated judgement and a critical assessment of the obtained values. As a final comparison, the travel time results for the time domain and the frequency domain have been gathered in a graph against the applied isotropic effective stress, shown in Figure 15. The convergence of the results in the higher stresses is evident. Despite following similar trends, the comparison between the different interpretation methods results in repetitive tests is not conclusive. In fact, the differences in the results for the first stress stages are quite variable among the three tests. While the RR1 test, the difference between the two domains is almost double, for RR2 it is less than 20%. Bearing in mind that these tests were carried out in identical samples, there is certainly room for further investigations. 15 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0 50 100 150 200 250 300 350 400 450 isotropic effective stress (kPa) travel time (ms) RR1-first arrival RR2-first arrival RR3-first arrival RR1-sweep RR2-sweep RR3-sweep Figure 15 – Travel time results for all tests, in the time and frequency domains Such discrepancies in the results clearly inform that there is still much to understand regarding how the input wave travels within the sample, whether and how much of it is reflected in its boundaries, and how it is picked up by the receiver on the other end. In our opinion, these series of tests have shown much uncertainty and many of the fragilities in the currently available interpretation methods. Nevertheless, these have also led to a continuing and renovated interest in a deeper research on this innovative and promising laboratory test. It should be pointed out that, alike all other testing equipments and technologies, there is a scope of materials and conditions to which the response of the system is more reliable and adequate. Outside that scope of applicability, there is a high probability of inconsistent results. We do not know whether this is the case for the bender element testing method. In any case, other geomaterials have been tested at labGEO of FEUP as well as in other laboratories, which show less dispersive and complex results. From this exercise derives the need to understand in greater depth how the input wave is actually traveling within the sample and the urgency to know the right track, whether it is the time or the frequency domain, or any other. 16 9 References Arroyo, M. (2001). Pulse tests in soil samples. PhD Thesis University of Bristol, UK. Arulnathan, R.; Boulanger, R.W.; Riemer, M.F. (1998). Analysis of bender element tests. Geotechnical Testing Journal, Vol. 21, No. 2, pp. 120-131. Blewett J, Blewett I. J. & Woodward P.K (1999). Measurement of shear wave velocity using phase sensitive detection techniques. Canadian Geotechnical Journal, No. 36, pp. 934-939. Ferreira, C.; Mendonça, A.A. & Viana da Fonseca, A. (2004). Assessment of sampling quality in experimental sites on residual soil from granite in Porto (in Portuguese). Proceedings of the 9th Portuguese Geotechnical Congress, Aveiro, Portugal, SPG, Lisboa, Vol. 1, pp. 27-38. Greening, P. D.; Nash, D. F. T.; Benahmed, N.; Ferreira, C. & Viana da Fonseca, A. (2003). Comparison of shear wave velocity measurements in different materials using time and frequency domain techniques. Proceedings of the 3rd International Symposium on Deformation Characteristics of Geomaterials, IS-Lyon '03, France. September 2003. Di Benedetto et al. (Eds). Swets & Zeitlinger, Lisse. Vol. 1, pp. 381-386. Greening, P.D & Nash, D.F.T. (2003). Frequency domain determination of G0 using bender elements. ASTM Journal of Geotechnical Testing. Jovičić, V. (2004). Rigorous bender element testing. Workshop on Bender Element Testing of Soils, UCL, London. Jovicic,V.; Coop, M.R.; Simic, M. (1996). Objective criteria for determining Gmax from bender element tests. Géotechnique, Vol. 46, No. 2, pp. 357-362. Kaarsberg, E.A. (1975). Elastic-Wave Velocity Measurements in Rocks and Other Materials by Phase- Delay Methods. Geophysics No.40, pp. 955-901. Moncaster, A.M. (1997). The shear modulus of sand at very small strains. MSc thesis. Department of Civil Engineering, University of Bristol, UK. Rio, J. (2002). Advances in laboratory geophysics using bender elements. Internal Report, University College London. Sánchez-Salinero I., Roesset J.M. & Stokoe II K.H. (1986). Analytical studies of wave propagation and attenuation Geotechnical report Viana da Fonseca, A. & Ferreira, C. (2002) Bender elements: excelent laboratory techniques for the evaluation of referencial geotechnical parameters (in Portuguese). Proceedings of the 8th Portuguese Geotechnical Congress, LNEC, Lisboa Viggiani, G. & Atkinson, J.H. (1995). Interpretation of bender element tests. Géotechnique, Vol. 45, No.1, pp.149-154. FEUP, January 2005 17