ASCE GSP 175-2007 FMGM 2007.pdf

1、 GEOTECHNICAL SPECIAL PUBLICATION NO. 175 FMGM 2007 PROCEEDINGS OF THE SEVENTH INTERNATIONAL SYMPOSIUM ON FIELD MEASUREMENTS IN GEOMECHANICS September 2427, 2007 Boston, Massachusetts SPONSORED BY The Geo-Institute of the American Society of Civil Engineers EDITED BY Jerry DiMaggio, P.E. Peter Osbor

2、n Published by the American Society of Civil Engineers Copyright and Disclaimer ISBN-13: 978-0-7844-0940-4 Any statements expressed in these materials are those of the individual authors and do not necessarily represent the views of ASCE, which takes no responsibility for any statement made herein.

3、No reference made in this publication to any specific method, product, process or service constitutes or implies an endorsement, recommendation, or warranty thereof by ASCE. The materials are for general information only and do not represent a standard of ASCE, nor are they intended as a reference i

4、n purchase specifications, contracts, regulations, statutes, or any other legal document. ASCE makes no representation or warranty of any kind, whether express or implied, concerning the accuracy, completeness, suitability, or utility of any information, apparatus, product, or process discussed in t

5、his publication, and assumes no liability therefore. This information should not be used without first securing competent advice with respect to its suitability for any general or specific application. Anyone utilizing this information assumes all liability arising from such use, including but not l

6、imited to infringement of any patent or patents. Copyright 2007 by the American Society of Civil Engineers. All Rights Reserved. ASCE and American Society of Civil EngineersRegistered in U.S. Patent and Trademark Office. Photocopies and reprints. You can obtain instant permission to photocopy ASCE p

7、ublications by using ASCEs online permission service (www.pubs.asce.org/authors/RightslinkWelcomePage.html). Requests for 100 copies or more should be submitted to the Reprints Department, Publications Division, ASCE, (address above); email: permissionsasce.org. A reprint order form can be found at

8、www.pubs.asce.org/authors/reprints.html. American Society of Civil Engineers ASCE International Headquarters 1801 Alexander Bell Drive Reston, VA 20191-4400 USA Call Toll-Free in the U.S.: 1-800-548-2723 (ASCE) Call from anywhere in the world: 1-703-295-6300 www.pubs.asce.org Preface Measurements of

9、 field performance during and after construction of constructed facilities has taken on more importance as structures become more complex and sensitive to deformations, the costs of surprise performance grow rapidly, owners seek to more actively manage risks to their projects, and the public demands

10、 diminished impacts from construction activities. Major developments in measurement technologies are providing more ways to monitor performance with near real-time frequency for decreasing cost. It is therefore beneficial to give those interested in instrumentation and performance monitoring an oppo

11、rtunity to meet regularly to exchange ideas and experiences to stimulate further advancement within field instrumentation. The international symposia for Field Measurements in GeoMechanics, FMGM as an acronym, are organized to serve this purpose. This publication covers the 7thInternational Symposiu

12、m on Field Measurements in Geomechanics, FMGM 2007, held in Boston, USA, from September 24-27, ,2007 under the sponsorship of the ASCE Geo-Institute. The great international interest and large attendance at this symposium have again clearly demonstrated the need for specialty conferences of this kin

13、d to build on the benefits from the previous six FMGM symposia held in Switzerland (1983), Japan (1987), Norway (1991), Italy (1995), Singapore (1999) and Norway (2003). The great interest in the FMGM 2007 symposium is reflected by the large number of abstracts received (155). After independent revi

14、ews by two reviewers, 105 papers were accepted for presentation and publication in the proceedings. These papers were written by 174 authors and co-authors from 19 countries. The topics of the papers cover a wide spectrum within the three main conference themes: Case studies demonstrating the role o

15、f field measurements in problem-solving, research, safety assessment, risk assessment and improving the design of civil engineering structures and works. State-of-the-art and futures trends in measurement technologies, equipment, communications, data management and interpretation, and visions for fu

16、ture development. Business side of instrumentation demonstrating and quantifying the benefits of field measurements to project management teams, owners, engineers, contractors, regulators and insurers. In addition three special workshops covering geotechnical instrumentation to measure performance,

17、inclinometers and innovations in instrumentation, installation and data acquisition were held. The efforts of many people were essential to the success of this symposium. As Chair of the Organizing Committee, my appreciation and thanks go to all those involved, especially to the authors of papers, t

18、he reviewers, the members of the Organizing Committee and the Editors of these proceedings, Jerry DiMaggio and Peter Osborn of the United States Federal Highway Administration. Without the extraordinary efforts of these people, there would have been no symposium and no proceedings. September 2007 W.

19、 Allen Marr, Chairman of Organizing Committee, FMGM 2007 President and CEO of Geocomp Corporation, Boxborough, MA, USA Geotechnical Special Publications 1 Terzaghi Lectures 2 Geotechnical Aspects of Stiff and Hard Clays 3 Landslide Dams: Processes, Risk, and Mitigation 7 Timber Bulkheads 9 Foundati

20、ons the second phase,dealing with advanced statistical analyses started in 2005 and ended in 2007.A surveillance system is a complex system resulting from applying statisticalsurveillance and diagnostics to a monitoring system and encompassing enough logicsand data understanding to give warning sign

21、als on an automatic basis.Generally speaking a monitoring system is involved in real-time data collectionfrom the corresponding monitored real system, which is considered as an input-output system. It is considered here as a collection of integrated single measuringinstruments and measuring systems

22、or chains. For example, in geomechanics and/orstructural monitoring, the system under monitoring may be a landslide or a build-ing or bridge and the monitoring system is composed of various interconnectedmeasuring systems which monitor positions, displacements, inclinations etc as wellas temperature

23、 precipitation, car tra c etc. Moreover, the monitoring system has adata collection unit which is typically made up of wireless connections and a centralmanaging and storage unit.Statistical surveillance and diagnostics is based on hierarchical multivariate de-tectors which are composed of various

24、advanced control charts applied to a mul-tidimensional dynamical grey-box model which describes the important dynamics,physical relationships and correlations among the data given by the monitoring sys-tem. The surveillance modules are designed to give a warning signal when thevariation of the corre

25、sponding components exceeds a statistical threshold due to anongoing anomaly in the monitored system. The statistical grey-box model exploitsthe relationships between system environmental input and geomechanics or struc-tural systemoutput by the use of correlations. It gives assessment of uncertaint

26、y dueto partial knowledge and measurement uncertainty of the real system under moni-toring. Due to this approach, the measurements can be adjusted for environmentaleects, correlations between measurements, uctuations over time and inertia.The rest of the paper is organized as follows. Section 2 pres

27、ents the source ofthe problem from the engineer point of view. Then section 2.2 extends statisticalanalysis and modelling of measuring systems introduced by Fass et al. (2004 and2005) to a monitoring system as a whole. Using the introduced model, section 3de nes a surveillance and diagnostics approa

28、ch based on a hierarchical multivariatedetector. A conclusion section concludes the paper.2 MODEL BASED UNCERTAINTY ANALYSISFig. 1showsatypical plotof monitoringdata, whichrefertoajointmetermeasuringthe relative movement of a bridge slab on its bearing devices for an observationperiod of approximate

29、ly two years, see section 2.1 for details. It is apparent at aglance that temperature aects the measurements. This is a typical situation, whichariseswhenmeasuringinstrumentsareoperatinginanopen eld, wheretemperatureplays a major role. The example refers to temperature eects since this is the mostco

30、mmon case (see e.g. Wah, 1971), but the same type of approach can be used forexample, when trying to separate the eect of the atmospheric pressure variationto water level measurements in a wall by means of absolute pressure transducers orthe eect of reservoir levels from dam body movement.Observing

31、Fig. 1, engineers are asked to answer to the following three naturalquestions:2FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics 2007 ASCEFigure 1: Joint meter row data (black line) and temperature (gray line).a) are measurements inuenced by instrument thermal drift ?b

32、) are measurements related to structural thermal behaviour ?c) is structural behaviour correlated to the temperature only or are some otherphenomena aecting it ?The answer to question a) is normally based on the analysis of the instrumentdata sheets and metrological specs supplied by the manufacture

33、r. Usually instru-ments are compensated for standard thermal limits (i.e. -20 / + 80 C), this meansthat no major output signal change is expected. For example, instruments are testedin climatic chambers to determine the thermal behaviour and to correct it by actingon the mechanics or on the electron

34、ics of the instrument. As a result, a limitedand known thermal shift is expected and the output signal can be corrected. Muchmore complicated is the understanding of the measuring chain behaviour (Dunni-cli, 1988): cables with junctions, wiring panels, data loggers or read out units mayintroduce unc

35、ertainties which are very di cult to quantify and correct. The onlyway they can be understood is by testing the whole chain, which this is practicallyimpossible. In this case thanks to the analysis of a signi cant set of data, the statis-tical solution aims at splitting the dierent components and as

36、sessing the magnitudeof the parameter under analysis.The answer to question b) relies almost entirely on the practical situation. Refer-ring to the case of Fig. 1, the inuence of the temperature on the movement valuescan be easily understood: when temperature increases the movement increases andwhen

37、 temperature decreases, the movement decreases too. The general behaviouris clear and a correlation between movement and temperature can be found. Hence,3FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics 2007 ASCEat rst level it can be said that temperature is the caus

38、e of the movement. Nev-ertheless, sometimes the thermal eect overshadows other eects which can be ofgreat importance or even crucial for understanding the behaviour of the monitoredstructure. These eects may be of limited magnitude over a short period, but canincrease with variable speed over a long

39、 time span. If they are overshadowed byother phenomena such as thermal eects, are may only be recognized only whenthey have raisen to a level which is critical for the monitored system.The answer to question c) requires the identi cation of all the important externaleects, their separation and the c

40、orresponding data adjustment. Using the approachof section 2.2, we obtain a good approximation of the real behaviour of the bridgeslab as illustrated in Fig. 2; this is clearly stationary except for a possible linearvariation of about 3mm in two years, which will be assessed by methods of section3.F

41、igure 2: Joint meter row data (black solid line) and processed data (grayline) together with linear trend of processed data.2.1 The Reference ExampleThe bridge in Fig.3, which supports heavy vehicular tra c in an urban area, isthe reference example of this paper and, although it has been discussed i

42、n moredetail by Bruzzi et al. (2007), it is briey summarized here for the sake of self-containedness. It is a three spans (45 + 90 + 45 m) cable stayed bridge with fourantennaseach one sustaining eight cables. It has been instrumented in orderto monitor its behaviour over time. The monitoring system

43、 being equipped withsome forty instruments was designed and installed in order to collect informationregarding dierent parameters, including: tilt of the antennas,4FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics 2007 ASCE load into the cables, vertical deformations

44、of the bridge slabs, horizontal movement of the bridge slabs on the bearings, temperature of the structural elements.The tilt of the antenna is measured by means of servo type biaxial inclinometers;loadinto cables bymeans of estensimetric load cells; vertical deformationof the slabsby potentiometric

45、 linear transducers and horizontal movements of the slabs by linearpotentiometric transducers as well (joint meters). All the installed instruments areprovided with a temperature sensor and are connected to a Data Acquisition Unitwhich enables data collection and transmission to the control centre.F

46、igure 3: Picture illustrating the monitored bridge.An example of the real behaviour of the bridge slab is illustrated in Fig. 2for a single instrument, where the processed data are obtained from the statisticalanalysis of section 2.2, considering the correlation between measured movement andtemperat

47、ure and the linear trend shows the tendency of the movement during theobservation period.Question c) of section 2 above may be now related to bridge slab deterioration,due for example to material ageing. In this case, the structural change is expectedto be gradual, having a frequency which is genera

48、lly low and dierent from thatof the daily and seasonal variation of temperature. Moreover, structural changesmay intervene due to changes in tra c load and frequency which, once again, areconsidered as slow changes. Data of Fig. 1 and 2 are row high frequency data.For the rest of this paper, daily a

49、verages are used, being appropriate for such lowfrequency surveillance and are restricted to rst year analysis5FMGM 2007: Seventh International Symposium on Field Measurements in Geomechanics 2007 ASCE2.2 Model Setup and EstimationIn this section, some statistical methods for assessing and reducing the environmen-tal biases and uncertainty of a monitoring system as a whole are discussed. Todo this, the approach which Fass et al. (2004 and 2005) intr