Renovation and Reoperation of a Geotechnical Centrifuge at the University of New Hampshire

Transcription

1 Renovation and Reoperation of a Geotechnical Centrifuge at the University of New Hampshire M. Ghayoomi & S. Wadsworth Department of Civil Engineering, University of New Hampshire, Durham, NH, USA ABSTRACT: The geotechnical centrifuge at the University of New Hampshire (UNH) is a historic piece of equipment, being one of the first centrifuges in the United States. The centrifuge was constructed in the early 70s at the California Institute of Technology and was recently relocated to UNH for renovation and reoperation. It is a 5 g-ton centrifuge with a 1 m radius arm. The centrifuge has a pay-load capacity of 100 kg and a centrifugal acceleration capacity of 175 g. The facility has been involved in numerous research projects over the years and includes several geotechnical modeling containers and an in-flight servo-hydraulic shaker. Since its relocation, the centrifuge has been substantially renovated and modernized. The old MTS controller was replaced with a new National Instruments motion controller and state-of-the-art sensing and measuring modules were appended to the system. The re-assembly process, centrifuge specifications, and new data acquisition system are explained in this paper. 1 INTRODUCTION The application of centrifuge modelling in geotechnical engineering education and research has been boosted in recent years (Scott 1983; Dewoolkar et al. 2003, Garnier et al. 2007). These applications range from material characterization problems (Ghayoomi and McCartney 2011) to fully modeled soil-structure systems (Dashti et al. 2010), from completely dry soils (Pitilakis et al. 2004) to partially and fully saturated soils (Ghayoomi et al. 2013, Dashti et al. 2010), and from static to dynamic problems (Madabhushi 2004). The purpose of centrifuge modelling is to produce a physical scaled model to replicate the real prototype system by increasing the gravitational acceleration level and proportionally reducing the dimensions. Similitude scaling factors are employed to correlate the model parameters to those of the prototype system (Garnier et al. 2007). Due to high demand for centrifuge modelling in geotechnical engineering research, many universities around the world have been acquiring geotechnical centrifuge facilities. Geotechnical centrifuges are categorized into two structural systems, i.e. drum and arm centrifuges. They can vary in size from small educational centrifuges to high capacity centrifuges with hundreds of g-ton pay load capacity. Geotechnical engineering group in the Department of Civil Engineering at the University of New Hampshire (UNH) recently attained and installed a geotechnical centrifuge. This centrifuge is one of the oldest centrifuges in the US originated from the California Institute of Technology (Scott 1983), which was later moved to the University of Southern California. In 2012, the centrifuge was relocated to UNH and has been installed and re-operated. Since then, some older pieces were replaced and a new controller and Data Acquisition system has been installed. This paper summarizes some of the main characteristics of this centrifuge. The installation and system renovation process are explained. In addition, the associated testing system, some of the past and present capabilities, and the future planned improvements are described. 2 CENTRIFUGE SPECIFICATIONS The centrifuge is a Model A1030 Genisco g- accelerator with a 2 m diameter aluminum-alloy arm rotating in a horizontal plane. It is rated at 5 g-ton payload capacity. The 0.46 by 0.59 m magnesium mounting frames at both end of the arm can carry 100 kg or 30 kg payload up to 50g or 175g respectively. The absolute acceleration capacity of the basket is 175g. A schematic of the original centrifuge arrangement and a picture of partly-assembled and upgraded centrifuge (without the protective casing) are shown is shown in Figure 1 and 2, respectively.

2 3 INSTALLATION PROCESS Figure 1. Schematic of the original centrifuge at Caltech (Hushmand 1984a). The centrifuge has been installed in room S166 Kingsbury Hall. The room has bay door access for sample preparation and equipment maneuverability. Electrical service has been installed to support the operation of the centrifuge drive motor (208V 40Amp single phase), and the hydraulic pump (208V 20A 3 phase), with a lock out disconnect switch. The centrifuge frame was mounted to the existing reinforced eight inch thick concrete slab floor. Hilti HIT-RE 500 Epoxy Adhesive anchoring system was used to secure ASTM A193 Grade B7 Alloy Steel threaded rod with a plain finish. Pictures of the anchoring system are shown in Figure 4. The centrifuge has been re-assembled with existing structural hardware. However, the dated bolt and connections were renewed. The bolt and connections were re-evaluated for possible imbalance loads and safety checks. The centrifuge includes an enclosing chamber to protect the system. Figure 2. Geotechnical centrifuge at UNH (during the assembly process). The machine is driven by a RG 2600, singlephase, full wave generative, static, DC drive with a 40.5 kw, 1725 rpm, 230 V, 3-phase, constant torque, double-ended electric drive motor from Sabina Electric and Engineering. The centrifuge arm is protected in an extruded aluminum casing. The electrical power and signals can be transferred to and from the arm through 44 slip rings with various capacities ranging from 10 to 30 amp. A picture of the slip ring system is shown in Figure 3. The centrifuge control system was completely renewed with a National Instrument PXI/SCXI system with an embedded controller, which will be running in LabView realtime module. The centrifuge control system is comprised of a Sabina wave regenerator, an electric motor, NI-based control modules (NI USB-9263 and NI USB-6009), relay motor activation switch, power supply, interlock hardware lockout system, and an emergency stop system. Figure 3. Slip rings system picture. Figure 4. Centrifuge ground anchoring system.

3 4 FACILITIES This centrifuge has been equipped with several mechanisms and containers with wide modeling capabilities for several geotechnical engineering and science applications. These systems include a shake table, a laminar container, fault mechanism container, several rigid and transparent boxes, saturation chamber, and pluviaton hopper. 4.1 In-flight shake table A 1-D in-flight hydraulic shake table is mounted on one side of the arm for dynamic modelling of geotechnical systems. A special smart mounting frame shown in Figure 5 was used. On each side of it, there is a row of four vertical cylindrical rods to support the test container. A hydraulic double chamber piston with a servo-valve is located at the center. Four horizontal bars are used to guide the test container. Figure 6. Hydraulic accumulators. Figure 5. Shake table mounting frame. The electro-hydraulic system controlling the shake table includes the following parts. A HPS-6A Team Vibration Testing system hydraulic pump, the servo-valve, accumulators, the controller and signal analyzer, and a LVDT. Shake table servo valve control is a G P-I servo-amplifier with compact DIN rail housing. An LVDT was used to provide a feedback mechanism. This enables the controller to assess the performance of the valve and correct the movement if necessary. NI PXI-6259 was used for shaker control and signal generation. A general schematic of the hydraulic system is shown in Figure 7. The pump includes a 75.7 liter reservoir with 3.7 kw motor and can supply up to 3000 psi (20.7 MPa) pressure. To control the flow capacity of the hydraulic pump, two small accumulators (ZEMARC standard bladder accumulators) (3.8 liter each) were fixed underneath the centrifuge arm, as shown in Figure 6. They can deliver extra flow of hydraulic oil very quickly during dynamic tests. A 71.5 liter accumulator outside the centrifuge delivers oil to the inside ones. A second line returns the oil to the hydraulic pump. Figure 7. Hydraulic system diagram (Hushmand 1984a) 4.2 Model containers Several model containers and testing mechanisms have been developed for this centrifuge including rigid containers, transparent boxes, and a laminar container. Laminar container has been used to control the boundary effects on the soil response (Arulanandan and Scott 1993, Hushmand et al. 1988). It was found to be an optimal solution because it simulates shear beam conditions in which the shear waves propagate uniformly through the soil specimen while minimizing the end effects (Whitman and Labe 1986). The centrifuge facility at UNH includes a laminar container with an inside dimensions of mm long, mm wide, and 254 mm height. It is composed of 19 rectangular aluminum rings, each 12.7 mm thick, and separated by a 0.6 mm gap between them. Linear bearings between the rings reduce friction during differential movement of the rings. A picture of the laminar container is shown in Figure 8.

4 Figure 8. Laminar container. 5 INSTRUMENTATION AND DATA ACQUSITION SYSTEM 5.1 Instruments The centrifuge lab provides wide variety of sensing system for different geotechnical modeling applications. These sensors include accelerometers, miniature pore pressure sensors, strain gauges, dielectric water content sensors, tensiometers, LVDTs, and miniature load cells. All sensors are capable of high sampling frequency measurement during dynamic centrifuge testing. 5.2 Data acquisition The Data Acquisition System (DAQ) was completely renovated with the state-of-the-art National Instrument chassis and modules. NI PXI-1031 and NI SCXI-1000 chassis with 3 PXI slots and 4 SCXI slots will be implemented along with a NI PXI-8115 controller (2.5 GHz Intel Core i5-2510e). NI PXI- 6259, M-Series DAQ is available for use as signal and SCXI control through SCXI-1349 interface. This multifunction DAQ can be employed for pressure, tachometer, and strain measurements. In addition, they have been used as the servo-valve signal analyzer. SCXI channel ICP accelerometer module with BNC connectivity and SCXI channel LVDT input module have been installed in this system. The DAQ computer chassis were mounted on the center of arm and it operates in-flight. This provides a shorter distance between the DAQ and the testing system. A simple PC was remotely connected to the DAQ providing the user-interface to control the test and adjust the measurements through a local wireless connection. 5.3 Shake table program The shake table is controlled through a NI LabView interface. The details of shake table hardware were explained in section 4.1. The Shaker Control software provides the user with abilities to configure sensor channels and data logging file parameters. This program is capable of generating earthquake and cyclic motions. Earthquake profiles are read through a text file and a file path control allows the user to select the earthquake file. The sampling frequency can be adjusted in the program. The data can be recorded before triggering the shaking and after the end of the shaking using a buffer data recording system, pre-trigger time, and post-trigger time mechanisms. The Shaker control software is implemented using multiple loop master-slave architecture in order to accommodate the various independent software functions such as configuration, data collection, data logging, and user interface control and display. Separation of the software functions provides isolation of the functionality within the software architecture, it makes the software development less complicated, and it allows future functions to be implemented without the need for major re-design of the architecture. 6 PAST, PRESENT AND FUTURE This historic centrifuge was one of the first centrifuges in the United States established in the early 1970s. Since then, it has been used in numerous pioneering research studies in geotechnical and earthquake engineering, geophysics, and geology. The following are some examples of these research projects: simulation of rupture processes on reverse fault (Roth et al. 1981, Burridge et al. 1989); earthquake response of earth dams (Abiom et al. 1986); dynamic testing of shallow foundations subjected to earthquakes (Hushmand 1984b); measurement of stresses in piles and soil during pile driving (Allard 1984); liquefaction studies (Hushmand et al and 1988); modeling of slope failure under static and dynamic loading (Burridge et al. 1989); base isolation studies using smooth geosynthetic liners/soil interface (Kavazanjian et al. 1991, Hushmand et al. 1994); and study of convection in earth s mantle (Nataf et al. 1984). After a few years of operation curtailment, this centrifuge was renovated and re-operated for geotechnical engineering research and education at UNH. In addition to upgrading the DAQ, several new testing systems are under development. For example, the laminar container is modified for steady state infiltration processes. The infiltration system during centrifugation is a strong tool in controlling

5 the suction for partially saturated soil testing (Ghayoomi et al. 2011). The centrifuge facility is currently available for research to geotechnical engineering community. Also, it is used for the educational purposes, especially in a graduate geotechnical modeling class. 7 ACKNOWLEDGEMENT The authors would like to thank Dr. Ben Hushmand and Dr. J.P. Bardet for their assistance through the centrifuge relocation. 8 REFRENCES Aboim, C., Scott, R.F., Lee, J.R., and Roth, W.H Centrifuge Earth Dam Studies: Earthquake Tests and Analyses. Dames and Moore, Final Report to the National Science Foundation. NSF Grant No. CEE Allard, M. A Tests on piles installed in flight on the centrifuge. Proceedings of the Symposium on Recent Advances in Geotechnical Centrifuge Modeling. University of California, Davis, pp Arulanandan, K. & Scott, R.F Verification of Numerical procedures for the analysis of soil liquefaction problems. Proceedings of the International Conference on the Verification of Numerical Procedures for the Analysis of Soil Liquefaction Problems. Vols. 1 and 2, A. A. Balkema, Rotterdam, the Netherlands. Burridge, P., Scott, R.F., and Hall, J.F Centrifuge studies of faulting effects on tunnel. Journal of Geotechnical Engineering. ASCE, Vol. 115, No. 7, pp Dashti, S., Bray, J.D., Pestana, J.M., Riemer, M., and Wilson, D Centrifuge Testing to Evaluate and Mitigate Liquefaction-Induced Building Settlement Mechanisms. ASCE Journal of Geotechnical and Geoenvironmental Engineering. 136(7), Dewoolkar, M.M Centrifuge Modeling for Undergraduate Geotechnical Engineering Instruction. ASTM Geotechnical Testing Journal. 26(2), 1-9. Garnier J., Gaudin, C., Springman, S.M., Culligan, P.J., Goodings, D., Kong, D., Kutter, B., Phillips, R., Randolph, M.F., and Thorel, L Catalogue of Scaling Laws and Similitude Questions in Geotechnical Centrifuge Modeling. International Journal of Physical Modeling in Geotechnics. 3, Ghayoomi, M., Ko, H.-Y., and McCartney, J. S Measurement of seismically induced settlement in unsaturated sands. ASTM Geotech. Test. J. 34(4), Ghayoomi, M., McCartney, J. S., and Ko, H.-Y Empirical Methodology to Estimate Seismically Induced Settlement of Partially Saturated Sand. ASCE J. of Geotechnical and Geoenvironmental Eng. 139(3), Ghayoomi, M., and McCartney, J. S Measurement of small-strain shear moduli of partially saturated sand during infiltration in a geo- technical centrifuge, ASTM Geotech. Test. J., 34(5), Hushmand, B. 1984a. Experimental Studies of Dynamic Response of Foundations. PhD Thesis, California Institute of Technology, CA, USA. Hushmand, B. 1984b. Dynamic Behavior of Foundations: An Experimental Study in a Centrifuge. Proceedings of the Symposium on Recent Advances in Geotechnical Centrifuge Modeling. University of California, Davis, July. Hushmand, B., Scott, R.F., Crouse, C.B., and Martin, G.R Dynamic Centrifuge Testing of a Bridge-Soil Model. Proceedings of the Second International Conference on Geotechnical Centrifuge Modeling, Centrifuge '88, Paris, France, pp Hushmand, B., Crouse, C.B., and Martin, G.R Dynamic Soil-Structure Interaction of a Single Span Bridge. Journal of Earthquake Engineering and Structural Dynamics, Vol. 15, No. 6. Hushmand, B., Scott, R.F., and Crouse, C.B Centrifuge Liquefaction Tests in a Laminar Box. Geotechnique. Vol. 38, No. 2, pp Hushmand, B., Scott, R.F., and Rashidi, H. 1994, AVELACS Model No. 3: Liquefaction in Dense and Loose Sand Columns. Proceedings of VELACS Conference. Vol. 2; and "A Model No. 3 Primary Test Description and Test Results." Proceedings of VELACS Conference, Vol. 1. Kavazanjian, E., Jr., Hushmand, B., Martin G.R., and Beck, J.L Frictional Base Isolation with Smooth Geomembranes. Proceedings of the Third United States Conference on Lifeline Earthquake Engineering. Los Angeles, California, August, Nataf, H. C., Hager, B.H., and Scott, R.F Convection experiments in a centrifuge and the generation of plumes in a very viscous fluid. Annales Geophysicae. pp Madabhushi, G Modeling of Earthquake damage using geotechnical centrifuges. Current Science, 8(10), Pitilakis, K., Kirtas, E., Sextos, A., Bolton, M., Madabhushi, G., and Brennan, A Validation by Centrifuge Testing of Numerical Simulations for Soil-Foundation-Structure Systems. 13 th World Conf. on Earthquake Engineering, Vancouver, B.C., Canada, Roth, W. II, Scott, R.F., and Austin, I Centrifuge Modeling of Fault Propagation through Alluvial Soils. Geophysical Research Letters. 8(6), Scott, R. F Centrifuge Model Testing at Caltech. Journal of Soil Dynamics and Earthquake Engineering. 2(4), pp Whitman, R.V. & Lambe, P.C Effect of Boundary Conditions Upon Centrifuge Experiments Using Ground Motion Simulation. Geotechnical Testing Journal, 9(2),