1、Field Experiments for Monitoring the Dynamic Soil StructureFoundation Response of a Bridge-Pier Model Structure at a Test SiteG. C. Manos, M.ASCE1; K. D. Pitilakis2; A. G. Sextos, M.ASCE3;V. Kourtides4; V. Soulis5; and J. Thauampteh6Abstract: Summary results from a series of field experiments at a t

2、est site in Greece are presented, involving an in situ instrumented bridge- pier model built on realistic foundation conditions, to study the dynamic behavior of structure-foundation-soil system. It was attempted to link the variation of its dynamic characteristics to certain changes in its structur

3、al system, including the development of structural damage. This measured response was next utilized to validate numerical tools capable of predicting influences arising from such structural changes as well as from soilfoundation interaction. This bridge-pier model was supported on soft soil deposits

4、 allowing the study of structurefoundation soil interaction effects during low-to-medium intensity artificial excitations. The in situ experiments provided measurements that were used to verify fundamental analytical solutions for soilstructure interaction. They were also used to validate numerical

5、simulations that were de- veloped to predict the response of the studied structure and thus, back-evaluate modeling assumptions. The obtained accuracy of the numeri- cal predictions must be partly attributed to sound knowledge of the mechanical properties of the pier model and of the soil, not neces

6、sarily the case in all practical applications. It is evident that more complex finite-element models can improve the quality of the prediction only in cases where their parameters can be defined equally well. A special study further focused on the radiation of the waves generated by the vibration of

7、 the bridge-pier model through the soil medium. It is deemed that this comprehensive experimental investigation of soilstructure interaction provides measurements of the system response and enhances our understanding of the physical phenomenon as a whole. DOI: 10.1061/ (ASCE)ST.1943-541X.0001154. 20

8、14 American Society of Civil Engineers.Author keywords: System identification; In-situ measurements; Bridge Pier; Soil-structure-foundation interaction; Test site; Radiation damping; Structural identification.IntroductionTo investigate the effects of earthquakes on structures including influences fr

9、om soilstructure interaction (SSI) a variety of exper- imental approaches have been utilized during the last decades to study in situ the coupled response of structures and the underlying soil. Such an effort must be seen as complementary to experiments conducted at earthquake simulators or centrifu

10、ge facilities (Elgamal et al. 2007). Many prototype structures have been instrumented in areas of high seismicity to monitor in situ their response during an earthquake, taking advantage of realistic foundation conditions that cannot always be reproduced in the laboratory (Trifunac et al. 2001; Akta

11、n and Brownjohn 2013; elebi 2003; Kijewski-Correa et al.2013; Mucciarelli et al. 2004; Rathje and Adams 2008; Stewart et al. 1999; Todorovska 2009). Instrumented prototype structures can be excited by earthquakes of medium to strong intensity generated at relatively short distances (i.e., R15 km); u

12、nder such excitations the nonlinear response of prototype structures is recordable (Kreger and Sozen 1989). All this, however, incurs high deployment and maintenance costs while the probability of experiencing a strong earthquake event within the life cycle of the structure or of the instrumentation

13、 system is quite low. Moreover, the in situ deployment of instrumentation dedicated to studying SSI effects on prototype structures often faces considerable obstacles.A less costly and attractive alternative is the instrumentation of model structures at test sites like the one presented here. Such m

14、odel structures can be excited either with lower levels of seismic ground-motion intensity, which are statistically more frequent, or by artificial excitations, as outlined in this paper. In addition, when these model structures are appropriately designed, they can also develop ad-hoc nonlinear mech

15、anisms of interest for less intense ground-motion or artificial excitations. The installation and main- tenance of proper instrumentation, as well as the performance of in situ experiments can be achieved more easily and safely at such a controlled test site, rather than within prototype structures

16、that con- tinue functioning. Another advantage of performing such in situ experiments with structural models of considerable size is the fact that such in situ models have realistic foundation conditions that permit the simultaneous study of soilfoundationstructure interac- tion effect; an objective

17、 that cannot be easily met at the laboratory. Moreover, instrumentation dedicated to studying SSI effects can be1Professor, Dept. of Civil Engineering, Aristotle Univ., Thessaloniki 54124, Greece (corresponding author). E-mail: gcmanoscivil.auth.gr2Professor, Dept. of Civil Engineering, Aristotle Un

18、iv., Thessaloniki54124, Greece.3Assistant Professor, Dept. of Civil Engineering, Aristotle Thessaloniki 54124, Greece.Univ.,4Civil Engineer, Dept. Thessaloniki 54124, Greece. 5Civil Engineer, Dept.Thessaloniki 54124, Greece. 6Civil Engineer, Dept. Thessaloniki 54124, Greece.ofCivilEngineering,Aristo

19、tleUniv.,ofCivilEngineering,AristotleUniv.,ofCivilEngineering,AristotleUniv.,Note. This manuscript was submitted on April 26, 2013; approved onJuly 15, 2014; published online on August 14, 2014. Discussion period open until January 14, 2015; separate discussions must be submitted for individual pape

20、rs. This paper is part of the Journal of Structural Engineer- ing, ASCE, ISSN 0733-9445/D4014012(11)/$25.00. ASCED4014012-1J. Struct. Eng.J. Struct. Eng.Downloaded from by Hebei University of Engineering on 09/25/14. Copyright ASCE. For personal use only; all rights reserved.deployed

21、 relatively easily prior to the construction stage of such structures at a test site, as will be presented here.This paper presents a comprehensive overview of the insight gained in soilfoundationstructure interaction at a test site in northern Greece (EuroSeis Test Site, Volvi) since 1995, through

22、a series of in situ tests that involved low-to-moderate intensity artificial excitations of model structures as seen in Fig. 1 (Manos 1998). This test site is located at a distance of 5 km from the epi- center of the damaging Thessaloniki 1978 earthquake (Papazachos et al. 2010) (40.8N, 23.2E, Ms 6.

23、5, Imax VIII+ MSK). It is a re-gion of high seismicity, surrounded by major faults (at a distance of 1020 km) as also reflected in the seismic zonation prescribed in the Greek National Annex of Euro code 8-Part 1 (CEN 2004), where the design peak ground acceleration (10% probability of being exceede

24、d in 50 years) is set to 0.24 g for this particular area. Prior to the construction of two such model structures this test site was equipped with means to generate artificial dynamic excitations as well as with permanent instrumentation capabilities to measure and record the excitation and the struc

25、tural response either during an earthquake or during artificial excitations, as will be explained here (Manos et al. 2004). The tested structures represent structural mock-ups designed (1) to activate the desired interaction mecha- nisms between the foundation and the surrounding soil, and(2) to dev

26、elop certain nonlinear mechanisms of interest under either feasible medium intensity artificial excitations or small-to-medium intensity earthquake ground motions.The aim of this paper is to present the most significant obser- vations of the investigation of the SSI effects as they were studied at t

27、his test site. This was initially attempted by the first model struc- ture built at this test site in 1994; it represents a small-scale (13)mock-up of a six-story R/C building with or without masonryinfill (right side of Fig. 1). The overall dimensions of this model are 3.5 3.5 m in plan and 6.5 m i

28、n height. Its weight together with the foundation slab and the added mass, placed at each slab, is ap- proximately 350 KN. Due to space limitations, only basic informa- tion is given here as further details can be obtained from the literature (Manos et al. 1996; Manos 1998; Okada and Tamura 1985).Th

29、e experience gained from the in situ tests of this building model were utilized in testing the second model structure that rep- resents a small-scale mock-up of a single R/C bridge pier with itsdeck and its foundation block. This bridge-pier model, the focus of this paper, was built in 2004 and is a

30、 small-scale (18) representa- tion of a single bridge pier on a shallow foundation block, with a part of its deck carrying additional weight (left side of Fig. 1) (Manos et al. 2006; Kawashima 2000). More details are givenin the subsequent sections. The thorough knowledge of the subsoilgeometry and

31、soil properties of the sedimentary valley of the Volvi graben (Raptakis et al. 2000) obtained during the years, further contributes to the reliable assessment of the dynamic response of the pierfoundationsoil system studied here.Specific Objectives of Bridge Pier Testing CampaignIn brief, the main o

32、bjective is to obtain measurements of the dy- namic response of the pierfoundationsoil system at this test site. These include elastic response and flexural nonlinear response at the bottom of the bridge pier together with foundationsoil structure interaction. Flexural damage is expected to develop

33、for such bridge piers designed and constructed according to relatively old design-code provisions. Subsequent objectives in utilizing these measurements are (1) study the coupling between the nonlin- ear mechanisms of both the structure and the supporting soil, under feasible artificial excitations

34、or for a small-to-medium intensity earthquake occurring at a relatively short distance from the test site (less than 10 km); (2) monitor the propagation of the wave field generated by the oscillating bridge-pier model during artificial vibration and emitted through the foundation to identify pattern

35、s of radiation damping within the surrounding soil. The waves gen- erated by the vibration of the bridge pier induce inertial loads back to the foundation, which are superimposed on those developed due to the main soil vibration. The effects of such coupling are not neg- ligible (Mylonakis and Gazet

36、as 2000); (3) investigate the validity of established analytical solutions (Mylonakis et al. 2006) predicting the dynamic impedance of a soilstructure interacting system rest- ing on surface footings; and (4) Assess the efficiency of different numerical modeling strategies in predicting the dynamic

37、response of the soilstructure system.Testing Capabilities of the EuroSeis Test SiteAn overview of the artificial excitations used to excite the model structures at this test site as well as the instrumentation schemes deployed to measure the level of excitation and the structural response are briefl

38、y presented here. These testing capabilities were changed and expanded based on the experiences gained by these in situ experiments of the building structure that preceded those of the pier-bridge structure.Artificial Excitations by Means of a Controlled Free-Vibration StateThe reinforced concrete w

39、alls of a storage house plus concrete blocks of sufficient dimensions embedded within the soil at proper locations at the test site served as reaction masses for cables used for pulling and releasing the pier-bridge structure to excite it in a controlled free-vibration state. This technique is herea

40、fter named low-intensity free-vibration.Controlled Explosions Performed at a Relatively Close DistanceArtificial excitations of an intensity higher than the one described before were achieved by performing single-source explosions at a distance approximately 150 m from the test site. They were used

41、to excite the building model. However, due to certain limitations, discussed in brief here, they were not fully utilized for the bridge model. The simple controlled explosions produced sufficiently strong waves, which were recorded by the acceleration sensors at the foundation block of the model str

42、uctures (Fig. 2). ThisFig. 1. Layout of the test site with the two model structures and a crane for lifting weights (in-plane: along two model axis) (image by G. C. Manos) ASCED4014012-2J. Struct. Eng.J. Struct. Eng.Downloaded from by Hebei University of Engineering on 09/25/14. Copy

43、right ASCE. For personal use only; all rights reserved.Fig. 3. Overall dimensions of bridge pier model together with added deck massFig. 2. Permanent instrumentation at middle of deck and at top of foun- dation block; center of each pressure transducer at 1,046.50 mm diag- onally from center of foun

44、dation block (x x is the in-plane direction,y y the out-of-plane direction)(Table 1). The measured response of the pier bridge for this exci- tation model is discussed in some detail. A crane for lifting weight (Fig. 1) was built together with a strong storage house for the power-generating equipmen

45、t. This storage house was also utilized as a reaction mass for free-vibration experiments for the pier-bridge structure.Table 1. Summary of Measured Eigenfrequencies for the Bridge PierBridge pier with cablesand strutsa (HZ)Bridge pier without cables and strutsa (HZ)Permanent Instrumentation Capabil

46、ity to Record Structural Response under Either Artificial Excitations or Low-to-Medium Seismic Ground MotionsThe permanent instrumentation, depicted in Fig. 2 (bridge-pier model), included a combination of uniaxial and triaxial accelerom- eters of sufficient sensitivity and dynamic range (0.1 mg to

47、2 g, where g is the acceleration of gravity) together with a stable 32- channel data-acquisition system. The sensors were placed both at the deck for the bridge-pier model as well as at the foundation block; thus it was possible to capture all aspects of the dynamic structural response of interest.

48、This permanent instrumentation scheme was capable of fulfilling this objective for both artificial excitations and for prototype earthquake events. The in situ experi- ments with artificial excitations were performed either during late spring or early autumn with less variable climatic conditions an

49、d temperatures ranging from 15 to 25C. The ac celeration measure- ments together with the exact knowledge of all the masses and their distribution (Figs. 24) could then be utilized to calculate important response values (e.g., base shear/overturning moment, Fig. 10).State of damagePrior to cracking

50、in-plane (x x) Prior to cracking out-of-plane (y y) After cracking in plane (x x)2.801.932.542.601.122.25After cracking out-of-plane (y y)1.711.10After cracking torsional ( )2.783aWith extra mass.type of excitation was rich in high frequencies (1050 Hz) but lacked the dominant modal frequencies of e

51、ither the building model (fundamental frequencies in the range of 2.444.98 Hz) (Manos et al. 1996; Manos 1998) or the bridge-pier model (Table 1). The dynamic response of the bridge pier due to an excitation from an explosion, in terms of overturning moment and base shear, was not sufficient to prod

52、uce the desired structural damage at the bottom of the bridge pier. Moreover, these explosions, although of relatively limited intensity, required special permits from the authorities, and they inconvenienced the inhabitants of the nearby villages. For these reasons, limited use was made of this tec

53、hnique for the bridge-pier model. To overcome this force limitation result- ing from the use of either controlled free-vibration tests or simple controlled explosions, the capability of low-to-medium intensity forced vibration was developed and utilized for the bridge-pier model, as presented in wha

54、t follows.Additional Portable Instrumentation to Measure Important Response Mechanisms and Soil FoundationStructure InteractionThe upper soil layer (4 m deep) at the test site had a shear wave velocity of 135 m=s, density equal to 18 KN=m3, and a shear modulus value equal to 37.4 MPa (Pitilakis et a

55、l. 1999). The permanent and portable instrumentation captured both the excita- tion and response of the bridge pier, including soilfoundation structure interaction. Additional instrumentation to that shown in Fig. 2 was placed at the soil surface at specific locations surround- ing the foundation bl

56、ock. Triaxial accelerometers were employed to capture the soilsurface response at a relatively close distance. The experience gained from the in situ building model experiments, was used in the instrumentation and recording of the dynamic soilfoundationstructure interaction response of the pier-brid

57、ge model. The location of the soilsurface acceleration sensors isLow-to-Medium Intensity Forced VibrationA forced vibration technique was employed for the bridge-pier model to excite it in its in-plane (xx) direction at levels higher than before (Figs. 24). For this purpose, a uniaxial hydraulic ac-

58、 tuator with a moving lead mass located at the deck of this model was utilized (Fig. 4). This produced a dynamic excitation named low-to-mediumintensityforcedvibration, which couldbecontrolled to have frequency content within the range of in-plane (x x)fundamental frequencies for the bridge-pier model (2.252.80 Hz) ASCED4014012-3J. Struct. Eng.J. Struct. Eng.Downloaded from by Hebei University of Engineering on 09/25/14. Copyright ASCE. For personal use only; all rights reserved.Fig. 4. Layout of forced vibration tests dw, dx, dr, (in-plane dir