UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Evaluation of soil-structure interaction effects in the dynamic response of instrumented bridges based… Fraino, Miguel 2013

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2013_fall_fraino_miguel.pdf [ 11.65MB ]
Metadata
JSON: 24-1.0074210.json
JSON-LD: 24-1.0074210-ld.json
RDF/XML (Pretty): 24-1.0074210-rdf.xml
RDF/JSON: 24-1.0074210-rdf.json
Turtle: 24-1.0074210-turtle.txt
N-Triples: 24-1.0074210-rdf-ntriples.txt
Original Record: 24-1.0074210-source.json
Full Text
24-1.0074210-fulltext.txt
Citation
24-1.0074210.ris

Full Text

EVALUATION OF SOIL-STRUCTURE INTERACTION EFFECTS IN THE DYNAMIC RESPONSE OF INSTRUMENTED BRIDGES BASED ON SEISMIC RECORDS  by Miguel Fraino  B.Sc., Universidad de Carabobo, Venezuela, 2004  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Civil Engineering)   THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  August 2013  ? Miguel Fraino, 2013 ii  Abstract  The main objective of this study is to gain a better understanding of soil-structure interaction (SSI) and how it affected the response of the bridges under earthquake excitation. Detailed studies of signals recorded at key locations of a bridge ?i.e. the bridge deck, pier base, and free field? were conducted. A total of 6 instrumented bridges subjected to 12 earthquakes were selected for the analysis, focusing on the behavior in the transverse direction for these particular cases. The first step of the evaluation process was the investigation of the modal properties by means of a system identification process.  Then response spectra were calculated for all the records, and the effects of SSI were determined by comparing the acceleration spectrum of the free field motions with the spectrum of the bridge motions recorded at the foundation slabs or on the pile caps. The frequency contents of the signals were compared based on the Fourier amplitude amplifications of the records. The analysis of the comparisons of column base vs. free-field and column base vs. deck allowed to highlight behaviors related to SSI effects. The possible variability of the fundamental transverse period as a function of the amplitude of shaking with time was investigated using wavelet analysis techniques. The results from both response spectra and Fourier spectra analyses showed clearly that the free field motions are not always de-amplified at the foundation due to the soil-structure interaction effect, as it has been generally accepted. It was also demonstrated that for the same site and bridge, amplification or de-amplification varies from one earthquake to another. For almost all cases in this study the time-frequency analysis results showed that the peak response at the deck corresponds to the natural transverse period. This observation does not hold true only for one case, where the dynamic behavior was highly affected by the approaching embankments. iii  Preface  This work was part of a research project entitled ?Soil-Structure Interaction in Performance Based Design of Bridges? carried on at the University of British Columbia, focused on providing an efficient approximate procedure to include SSI effects in structural analysis and design of bridges. The funding for this research project was provided by the Natural Sciences and Engineering Research Council of Canada (NSERC).  Professor Carlos E. Ventura was the Project Leader and also the Supervisor of this research work. He provided important feedback during the different stages of the research and was the main reviewer of this document for its preliminary and final version. As the Director of the Earthquake Engineering Research Facility (EERF), Prof. Ventura granted the author access to hardware and software available in the Earthquake Engineering Research Facility (EERF). Professors W.D. Liam Finn and Mahdi Taiebat were Co-Supervisors of this work and provided crucial feedback and advices during its different stages. Prof. Taiebat carried out a complete revision of this document that allowed preparing an improved final document. The comments and advices from Dr. Anoosh Shamsabadi were very valuable to achieve the research objectives.   Parts of the material included in this document were published in three conference articles, listed below: ? Fraino M. and Ventura C.E. (2011) Analysis of Seismic Records to Evaluate Soil-Structure Interaction Effects on Bridges. Topics on the Dynamics of Civil Structures, Volume 1. Conference Proceedings of the Society for Experimental Mechanics Series 2012, pp 145-153. iv  ? Fraino M., Ventura C.E., Finn, W.D.L., and Taiebat, M. (2012) Seismic Soil-Structure Interaction Effects in Instrumented Bridges. Proceedings of the Fifteenth World Conference on Earthquake Engineering, Lisbon, Portugal. Paper ID: 5307, 10 pages.  ? Fraino M. and Ventura C.E. (2013) Soil-Structure Interaction Effects in the Dynamic Behavior of Bridges. Proceedings of the International Operational Modal Analysis Conference 2013, Minho, Portugal. Paper ID: 204, 8 pages.  The author contributed to this research work by carrying out all the numerical analyses and preparing the drafts and final document. The software described in Chapter 4 was prepared in collaboration with Indira Pandey as the programmer. The author provided the calculation algorithms, along with the design of the software interface and the logo.   v  Table of Contents  Abstract .......................................................................................................................................... ii Preface ........................................................................................................................................... iii Table of Contents ...........................................................................................................................v List of Tables ..................................................................................................................................x List of Figures .............................................................................................................................. xii Acknowledgements .................................................................................................................... xix Dedication ................................................................................................................................... xxi Chapter  1: Introduction ...............................................................................................................1 1.1 Background ..................................................................................................................... 1 1.2 Motivation and purpose .................................................................................................. 5 1.3 Objectives ....................................................................................................................... 6 1.4 Scope of this study .......................................................................................................... 7 Chapter  2: Theoretical background ............................................................................................8 2.1 Soil-Structure Interaction (SSI) ...................................................................................... 8 2.1.1 General description ..................................................................................................... 8 2.1.2 Inertial interaction ....................................................................................................... 9 2.1.3 Kinematic interaction ................................................................................................ 12 2.2 System identification .................................................................................................... 13 2.2.1 General overview ...................................................................................................... 13 2.2.2 Frequency Domain Decomposition (FDD) ............................................................... 13  vi  Chapter  3: Method of analysis ...................................................................................................15 3.1 General overview .......................................................................................................... 15 3.2 Bridges description ....................................................................................................... 16 3.3 Ground motions description .......................................................................................... 17 3.4 Comparative analysis .................................................................................................... 18 3.4.1 Overview ................................................................................................................... 18 3.4.2 System identification ................................................................................................ 18 3.4.3 Response spectra evaluation ..................................................................................... 19 3.4.4 Fourier spectra evaluation ......................................................................................... 20 3.4.5 Free field record vs. spectral analysis results comparison ........................................ 22 3.5 Time-frequency domain analysis .................................................................................. 24 3.5.1 Overview ................................................................................................................... 24 3.5.2 Evaluation of time-frequency spectra ....................................................................... 24 Chapter  4: BRIDGE-SSI System ...............................................................................................27 4.1 Overview ....................................................................................................................... 27 4.2 BRIDGE-SSI Seismic Records Database Website ....................................................... 27 4.3 BRIDGE-SSI Software ................................................................................................. 30 4.3.1 Bridge description and records ................................................................................. 30 4.3.2 Time-history .............................................................................................................. 32 4.3.3 Response spectra ....................................................................................................... 34 4.3.4 Frequency domain ..................................................................................................... 34 4.3.5 SSI effect ................................................................................................................... 34 4.3.6 Time domain ............................................................................................................. 34 vii  Chapter  5: Bridges description ..................................................................................................36 5.1 Bridges selection criteria ............................................................................................... 36 5.2 Bridges description ....................................................................................................... 36 5.2.1 Meloland Overpass ................................................................................................... 37 5.2.2 Via California ............................................................................................................ 38 5.2.3 Highway 395 Bridge ................................................................................................. 39 5.2.4 Alviso Overpass (Bridges K and L) .......................................................................... 40 5.2.5 Murray Road Bridge ................................................................................................. 42 Chapter  6: Results.......................................................................................................................44 6.1 Comparative evaluation of selected bridges ................................................................. 44 6.1.1 Meloland Overpass ................................................................................................... 44 6.1.1.1 Description of the ground motions ................................................................... 44 6.1.1.2 Modal identification results .............................................................................. 46 6.1.1.3 Response spectra analysis ................................................................................. 47 6.1.1.4 Fourier spectra analysis ..................................................................................... 49 6.1.1.5 Free field record vs. spectral analysis results comparison ................................ 51 6.1.2 Via California Bridge ................................................................................................ 53 6.1.2.1 Description of the ground motions ................................................................... 53 6.1.2.2 Modal identification results .............................................................................. 54 6.1.2.3 Response spectra analysis ................................................................................. 55 6.1.2.4 Fourier spectra analysis ..................................................................................... 57 6.1.2.5 Free field record vs. spectral analysis results comparison ................................ 59 6.1.3 Highway 395 Bridge ................................................................................................. 61 viii  6.1.3.1 Description of the ground motions ................................................................... 61 6.1.3.2 Modal identification results .............................................................................. 62 6.1.3.3 Response spectra analysis ................................................................................. 62 6.1.3.4 Fourier spectra analysis ..................................................................................... 65 6.1.3.5 Free field record vs. spectral analysis results comparison ................................ 67 6.1.4 Alviso Overpass (Bridges K and L) .......................................................................... 68 6.1.4.1 Description of the ground motions ................................................................... 68 6.1.4.2 Modal identification results .............................................................................. 70 6.1.4.3 Response spectra analysis ................................................................................. 70 6.1.4.4 Fourier spectra analysis ..................................................................................... 72 6.1.4.5 Free field record vs. spectral analysis results comparison ................................ 74 6.1.5 Murray Road Bridge ................................................................................................. 75 6.1.5.1 Description of the ground motions ................................................................... 75 6.1.5.2 Modal identification results .............................................................................. 77 6.1.5.3 Response spectra analysis ................................................................................. 77 6.1.5.4 Fourier spectra analysis ..................................................................................... 78 6.1.5.5 Free field record vs. spectral analysis results comparison ................................ 79 6.1.6 Summary of results for comparative evaluation ....................................................... 80 6.2 Time-frequency analysis ............................................................................................... 82 6.2.1 Meloland Overpass ................................................................................................... 83 6.2.2 Via California bridge ................................................................................................ 86 6.2.3 Highway 395 bridge .................................................................................................. 89 6.2.4 Alviso Overpass (Bridges K and L) .......................................................................... 93 ix  6.2.5 Murray Road Bridge ................................................................................................. 99 6.2.6 Summary of results for time-frequency analysis .................................................... 102 Chapter  7: Conclusions and recommendations .....................................................................104 7.1 Summary ..................................................................................................................... 104 7.2 Conclusions ................................................................................................................. 105 7.3 Recommendations ....................................................................................................... 106 Bibliography ...............................................................................................................................108 Appendices ..................................................................................................................................111 Appendix A BRIDGE-SSI System manual ............................................................................ 111 A.1 BRIDGE-SSI Database management ..................................................................... 111 A.2 BRIDGE-SSI Software user manual ....................................................................... 122 Appendix B Response spectra plots ........................................................................................ 136  x  List of Tables  Table 6.1.    Ground motions included in the analysis of the Meloland Overpass ....................... 44 Table 6.2.    Identified fundamental mode parameters in the transverse direction. Meloland Overpass ........................................................................................................................................ 46 Table 6.3.    Spectral ratios for the identified natural transverse period. Meloland Overpass ...... 48 Table 6.4.    Free field record properties vs. spectral analysis results, Meloland Overpass ......... 51 Table 6.5.   Ground motions included in the analysis of the Via California Bridge ..................... 54 Table 6.6.    Identified fundamental mode parameters in the transverse direction for the Via California Bridge .......................................................................................................................... 54 Table 6.7.    Spectral ratios for the identified natural transverse period, Via California Bridge .. 56 Table 6.8.    Free field record properties vs. spectral analysis results, Via California Bridge. ..... 60 Table 6.9.    Ground motions included in the analysis of the Highway 395 Bridge ..................... 61 Table 6.10.    Identified fundamental mode parameters in the transverse direction for the Highway 395 Bridge ..................................................................................................................... 62 Table 6.11.    Spectral ratios for the identified natural transverse period, Highway 395 Bridge . 64 Table 6.12.    Free field record properties vs. spectral analysis results, Highway 395 Bridge. .... 67 Table 6.13.    Ground motion included in the analysis of the Alviso Overpass (Bridges K and L)....................................................................................................................................................... 69 Table 6.14. Identified fundamental modal parameters in the transverse direction for the Alviso Overpass (Bridges K and L) ......................................................................................................... 70 Table 6.15.   Spectral ratios for the identified natural transverse period. Highway 395 Bridge .. 70 xi  Table 6.16.    Free field record properties vs. spectral analysis results. Alviso Overpass, Bridges K and L. ........................................................................................................................................ 74 Table 6.17.    Ground motion included in the analysis of the Murray Road Bridge ..................... 75 Table 6.18.    Identified modal parameters in the transverse direction for the Murray Road Bridge....................................................................................................................................................... 77 Table 6.19.    Spectral ratios for the identified natural transverse period. Murray Road Bridge .. 77 Table 6.20. Free field record properties vs. spectral analysis results, Murray Road Bridge. ....... 79 Table 6.21.   Spectral response ratios for acceleration (Sa), velocity (Sv) and displacement (Sd) corresponding to the identified fundamental transverse period (Tn) for all the studied cases ..... 80 Table 6.22. Summary of observations and findings from the evaluation of response spectra ..... 82 Table 6.23. Comparison between observed frequencies of peak PSD-TISA at the center of the deck and identified fundamental frequencies, Meloland Overpass .............................................. 83 Table 6.24.    Comparison between observed frequencies of Peak PSD-TISA at the center of the deck and identified fundamental frequencies, Highway 395 Bridge ............................................ 90 Table 6.25.    Comparison between observed frequencies of Peak PSD-TISA at the center of the deck and identified fundamental frequencies. Alviso Overpass (Bridge K) ................................ 95 Table 6.26.    Comparison between observed frequencies of Peak PSD-TISA at the center of the deck and identified natural frequencies. Alviso Overpass (Bridge L) .......................................... 96 Table 6.27.    Comparison between observed frequencies of Peak PSD-TISA at the center of the deck and identified natural frequencies. Murray Road Bridge. .................................................... 99 Table 6.28.    Summary of results for time-frequency analysis. ................................................. 102 Table 6.29.   Summary of findings from the time-frequency analysis. ...................................... 103  xii  List of Figures  Figure 2.1. Displacements induced on a SDOF system under seismic excitation without inertial SSI effects ....................................................................................................................................... 9 Figure 2.2. Displacements induced on a SDOF system under seismic excitation with inertial SSI effects ............................................................................................................................................ 10 Figure 2.3. SDOF model considering SSI effects  ........................................................................ 11 Figure 3.1. Example graphics used for Fourier spectra evaluation ............................................... 22 Figure 3.2. Example of comparison between free field record properties and spectral analysis results ............................................................................................................................................ 23 Figure 3.3. Example of 2D wavelet time-frequency spectrum ..................................................... 25 Figure 4.1. Search screen in BRIDGE-SSI Seismic Records Database Website  ........................ 28 Figure 4.2. Search results list in BRIDGE-SSI Seismic Records Database Website ................... 29 Figure 4.3. Bridge information section in BRIDGE-SSI Software. ............................................. 31 Figure 4.4. Records section in BRIDGE-SSI Software. ............................................................... 31 Figure 4.5. Time-history section in BRIDGE-SSI Software. ....................................................... 32 Figure 4.6. Filtering algorithm implemented in BRIDGE-SSI Software. .................................... 33 Figure 4.7. Time-history plots in BRIDGE-SSI Software. ........................................................... 33 Figure 5.1. Elevation and plan view of Meloland Overpass including sensors layout (from http://www.strongmotioncenter.org/) ........................................................................................... 37 Figure 5.2. Elevation and plan view of Via California Bridge including sensors layout (from http://www.strongmotioncenter.org/) ........................................................................................... 38 xiii  Figure 5.3. Elevation and plan view of Via California Bridge including sensors layout (from http://www.strongmotioncenter.org/) ........................................................................................... 39 Figure 5.4. Elevation and plan view of Alviso Overpass ?Bridges K and L? including sensors layout (from CESMD: http://www.strongmotioncenter.org/) ...................................................... 41 Figure 5.5. Elevation and plan view of Murray Road Bridge including sensors layout (from http://www.strongmotioncenter.org/) ........................................................................................... 42 Figure 6.1. Displacement orbit plots of the ground motions included in the analysis. Meloland Overpass ........................................................................................................................................ 45 Figure 6.2. Response spectra for Calexico 22May2010 earthquake (PGA=0.03g), Meloland Overpass ........................................................................................................................................ 47 Figure 6.3. Response spectra for Calexico 30Dec2009 earthquake (PGA=0.17g), Meloland Overpass ........................................................................................................................................ 47 Figure 6.4. Response spectra for Calexico 04Apr2010 earthquake (PGA=0.21g), Meloland Overpass ........................................................................................................................................ 47 Figure 6.5. Spectral ratios for the three analyzed earthquakes, Meloland Overpass .................... 49 Figure 6.6. Fourier spectra for Calexico 22May2010 (PGA=0.03g), Meloland Overpass ........... 50 Figure 6.7. Fourier spectra for Calexico 30Dec2009 (PGA=0.17g), Meloland Overpass ............ 50 Figure 6.8. Fourier spectra for Calexico 04Apr2010 (PGA=0.21g), Meloland Overpass ............ 50 Figure 6.9. Free field record properties vs. spectral analysis results, Meloland Overpass ........... 52 Figure 6.10. Displacement orbit plots of the ground motions included in the analysis, Via California Bridge .......................................................................................................................... 53 Figure 6.11. Response spectra corresponding to the Borrego Springs 07Jul2010 earthquake (PGA=0.006g), Via California Bridge .......................................................................................... 55 xiv  Figure 6.12. Response spectra corresponding to the Calexico 04Apr2010 earthquake (PGA=0.020g), Via California Bridge .......................................................................................... 55 Figure 6.13. Response spectra corresponding to the Chino Hills 29Jul2008 earthquake (PGA=0.023g), Via California Bridge .......................................................................................... 56 Figure 6.14. Spectral ratios for the three analyzed earthquakes, Via California Bridge .............. 56 Figure 6.15. Fourier spectra for Borrego Springs 07Jul2010 (PGA=0.006g), Via California Bridge ............................................................................................................................................ 58 Figure 6.16. Fourier spectra for Calexico 04Apr2010 (PGA=0.020g), Via California Bridge .... 58 Figure 6.17. Fourier spectra for Chino Hills 29Jul2008 (PGA=0.023g), Via California Bridge . 59 Figure 6.18. Free field record properties vs. spectral analysis results, Via California Bridge ..... 60 Figure 6.19. Displacement orbit plots of the ground motions included in the analysis, Highway 395 Bridge ..................................................................................................................................... 61 Figure 6.20. Response spectra for Qualeys Camp 18Sep2004 (PGA=0.014g), Highway 395 Bridge ............................................................................................................................................ 63 Figure 6.21. Response spectra for Toms Place 26Nov2006 (PGA=0.015g), Highway 395 Bridge....................................................................................................................................................... 63 Figure 6.22. Response spectra for Mammoth Lakes 12Jun2007 (PGA=0.051g), Highway 395 Bridge ............................................................................................................................................ 63 Figure 6.23. Spectral ratios for the three analyzed earthquakes, Highway 395 Bridge ................ 64 Figure 6.24. Fourier spectra for Qualeys Camp 18Sep2004 (PGA=0.014g), Highway 395 Bridge....................................................................................................................................................... 66 Figure 6.25. Fourier spectra for Toms Place 26Nov2006 (PGA=0.015g), Highway 395 Bridge 66 xv  Figure 6.26. Fourier spectra for Mammoth Lakes 12Jun2007 (PGA=0.051g),. Highway 395 Bridge ............................................................................................................................................ 66 Figure 6.27. Free field record properties vs. spectral analysis results, Highway 395 .................. 68 Figure 6.28. Displacement orbit plot of the ground motion included in the analysis. Alviso Overpass (Bridges K and L) ......................................................................................................... 69 Figure 6.29. Response spectra corresponding for Gilroy 13May2002 (PGA=0.019g), Alviso Overpass Bridge K ........................................................................................................................ 71 Figure 6.30. Response spectra for Gilroy 13May2002 (PGA=0.019g), Alviso Overpass Bridge L....................................................................................................................................................... 71 Figure 6.31. Spectral ratios for the Gilroy 13May2002 earthquake (PGA=0.019g), Alviso Overpass (Bridge K) ..................................................................................................................... 72 Figure 6.32. Spectral ratios for the Gilroy 13May2002 earthquake (PGA=0.019g), Alviso Overpass (Bridge L) ...................................................................................................................... 72 Figure 6.33. Fourier spectra for Gilroy 13May2002 (PGA=0.019g), Alviso Overpass (Bridge K)....................................................................................................................................................... 73 Figure 6.34. Fourier spectra for Gilroy 13May2002 (PGA=0.019g), Alviso Overpass (Bridge L)....................................................................................................................................................... 73 Figure 6.35. Free field record properties vs. spectral analysis results. Alviso Overpass (Bridges K and L) ............................................................................................................................................ 75 Figure 6.36. Displacement orbit plot of the ground motion included in the analysis, Murray Road Bridge ............................................................................................................................................ 76 Figure 6.37. Response spectra for Ferndale 09Jan2010 (PGA=0.077g), Murray Road Bridge ... 78 Figure 6.38. Spectral ratios for the analyzed earthquake, Murray Road Bridge .......................... 78 xvi  Figure 6.39. Fourier spectra for Ferndale 09Jan2010 (PGA=0.077g), Murray Road Bridge ....... 79 Figure 6.40. Time-frequency spectrum of the signal at the free field. Calexico 22May2010 earthquake (PGA=0.031g), Meloland Overpass ........................................................................... 83 Figure 6.41. Time-frequency spectrum of the signal at the column base. Calexico 22May2010 earthquake (PGA=0.031g), Meloland Overpass ........................................................................... 84 Figure 6.42. Time-frequency spectrum of the signal at the center of the deck. Calexico 22May2010 earthquake (PGA=0.031g), Meloland Overpass ...................................................... 84 Figure 6.43. Frequency Domain Decomposition chart and modal shapes associated to frequencies of interest for Calexico 22May2010 earthquake (PGA=0.031g), Meloland Overpass....................................................................................................................................................... 85 Figure 6.44. Time-frequency spectrum of the signal at free field. Chino Hills 29Jul2008 earthquake (PGA=0.023g). Via California Bridge. ...................................................................... 87 Figure 6.45. Time-frequency spectrum of the signal at column base. Chino Hills 29Jul2008 earthquake (PGA=0.023g). Via California Bridge ....................................................................... 87 Figure 6.46. Time-frequency spectrum of the signal at the center of the deck. Chino Hills 29Jul2008 earthquake (PGA=0.023g). Via California Bridge...................................................... 88 Figure 6.47. Time-frequency spectrum of the signal at the right end of the deck. Chino Hills 29Jul2008 earthquake (PGA=0.023g). Via California Bridge...................................................... 88 Figure 6.48. Time-frequency spectrum of the signal at free field. Mammoth Lakes 12Jun2007 earthquake (PGA=0.051g). Highway 395 Bridge ........................................................................ 90 Figure 6.49. Time-frequency spectrum of the signal at column base. Mammoth Lakes 12Jun2007 earthquake (PGA=0.051g). Highway 395 Bridge ........................................................................ 91 xvii  Figure 6.50. Time-frequency spectrum of the signal at the center of the deck. Mammoth Lakes 12Jun2007 earthquake (PGA=0.051g). Highway 395 Bridge. ..................................................... 91 Figure 6.51. Time-frequency spectrum of the signal at the right end of the deck. Mammoth Lakes 12Jun2007 earthquake (PGA=0.051g). Highway 395 Bridge. ..................................................... 92 Figure 6.52. Time-frequency spectrum of the signal at free field. Gilroy 13May2002 earthquake (PGA=0.019g). Alviso Overpass (Bridges K and L). ................................................................... 93 Figure 6.53. Time-frequency spectrum of the signal at column base. Gilroy 13May2002 earthquake (PGA=0.019g). Alviso Overpass (Bridge K). ............................................................ 94 Figure 6.54. Time-frequency spectrum of the signal at the center of the deck. Gilroy 13May2002 earthquake (PGA=0.019g). Alviso Overpass (Bridge K). ............................................................ 94 Figure 6.55. Time-frequency spectrum of the signal at the left end of the deck. Gilroy 13May2002 earthquake (PGA=0.019g). Alviso Overpass (Bridge K). ........................................ 95 Figure 6.56. Time-frequency spectrum of the signal at column base. Gilroy 13May2002 earthquake (PGA=0.019g). Alviso Overpass (Bridge L). ............................................................ 96 Figure 6.57. Time-frequency spectrum of the signal at the center of the deck. Gilroy 13May2002 earthquake (PGA=0.019g). Alviso Overpass (Bridge L). ............................................................ 97 Figure 6.58. Time-frequency spectrum of the signal at the left end of the deck. Gilroy 13May2002 earthquake (PGA=0.019g). Alviso Overpass (Bridge L). ........................................ 97 Figure 6.59. Time-frequency spectrum of the signal at free field. Ferndale 09Jan2010 earthquake (PGA=0.077g). Murray Road Bridge ........................................................................................... 99 Figure 6.60. Time-frequency spectrum of the signal at column base. Ferndale 09Jan2010 earthquake (PGA=0.077g). Murray Road Bridge ....................................................................... 100 xviii  Figure 6.61. Time-frequency spectrum of the signal at the center of the deck. Ferndale 09Jan2010 earthquake (PGA=0.077g). Murray Road Bridge .................................................... 100 Figure 6.62. Time-frequency spectrum of the signal at the left extreme of the deck. Ferndale 09Jan2010 earthquake (PGA=0.077g). Murray Road Bridge .................................................... 101  xix  Acknowledgements  I owe the main acknowledgement to Prof. Carlos E. Ventura, my Supervisor in this Master of Applied Science program. His academic and financial help, his advice and particularly his strong confidence in my work were great supports of this life-changing journey from the first stages. As part of his class I could appreciate the importance of Dynamics as a fundamental topic for any structural engineer, which principles are the source of a wide range of solutions. His guidance as the leader of the research project where I worked was crucial to learn how to generate good-quality research products on time, even more during the preparation and reviewing process of this thesis document. And especially his example as a person showed me how talent and constant effort should be braced with human quality to assure long-lasting success.  I also owe special thanks to Prof. W.D. Liam Finn. He was a great Professor of Earthquake Engineering inside and outside the classroom. I will always thank him the opportunity and honor of working as a Teaching Assistant of his course titled ?Seismicity and Seismic Design Parameters? for two years. His valuable comments and guidance for my research were crucial to achieve this final result. His exceptional career as a very talented researcher and engineer is a source of inspiration for me, as it has been for his students and co-workers for many years.  I would like to acknowledge Prof. Mahdi Taiebat for his guidance in my research work as one of the leading Professors of this project. His excellent classes in the Geotechnical Earthquake Engineering course were fundamental for me to understand this subject. And also his reviewing and comments on this document were very important to shape the final product. xx  I would also like to thank Dr. Anoosh Shamsabadi for his important contributions to my work and especially for his encouragement after every time we discussed the topics included here.  One of the important supports that I was blessed to have during this program is my great group of friends. My gratitude to Seku Catacoli, Manuel Archila, Jose Centeno, Bishnu Pandey, Jason Dowling, Majid Baradaran, Ana Bola?os, Manuel Monroy, Brook Robazza, Laura Quiroz, Amin Rahmani, Antone Dabeet, Miguel Guillen and Dana Malaguti. I thank all of them for their unconditional help in several ways and moments, and for the very refreshing conversations during these years.  There are no words big enough to describe how thankful I am to my family. My mother Nora, my father Jose Miguel and my sister Maria Claudia provided the core support that I needed to complete this program, the greatest goal I have achieved so far. Their continuous and strong encouragement will always remain as my best memory from these years. I will always remember the everlasting support from my cousin Carlos Daniel too, since the first steps of this project. The example from my uncle Elio Gonzalez Barboza has been there to guide me from my very first stages as an engineer. My grandfather Jose Miguel was also a great supporter that helped me to achieve this goal. I offer my greatest gratitude to all of them, and through them I acknowledge my stronger protector and source of energy: The Holy Trinity.     xxi    Dedication  To Nora, Jose Miguel and Maria Claudia  And especially to my grandfather Elio Gonzalez Urdaneta (1927-1986),  a former UBC student and the main source of inspiration for all my journeys   1  Chapter  1: Introduction  1.1 Background  The seismic behavior of any structural system is highly dependent on the level of fixity of the superstructure to its supporting foundation. Soil-structure interaction (SSI) in bridges is the term used to define the type of interaction between the structural system and the soil at the foundation level and the abutments. This reciprocal action has an important effect on the response of the structure under horizontal and vertical accelerations.  This problem is described by defining two types of interaction: inertial and kinematic. The inertial interaction is related to the inertial forces developed by the structure ?due to its own vibration? and transmitted to the base, which consequently act on the foundation soil and affect the shaking at that level. The kinematic type is mainly related to the process of the stiff foundation and the surrounding soil trying to develop mutually compatible displacements. Both types of interaction cause differences between the shaking at the column base and the motion at free field location, which is defined as a point located close to the structure but not affected by its movement.  The complexity and importance of this problem have encouraged research works for decades. A comprehensive study carried out at the Massachusetts Institute of Technology (Kausel 2009) reports the first studies in the mid-19th century when Sir William Thompson analyzed the problem of determining the displacements of an arbitrary point in an elicited elastic infinite solid 2  where concentrated static forces are acting. The solutions to this problem were based on static analysis until the beginning of the 20th century, when Sir Horace Lamb studied this phenomenon based on a dynamic approach. The influence of soil-structure interaction effects on structural response was initially assessed by Sezawa and Kanai in the 1930?s.   Several research works done in the mid-1970?s established simplified analytical methodologies to predict inertial interaction effects on dynamic response, such a period lengthening and soil/foundation damping factor. The work done by Bielak (1975) and Veletsos & Nair (1975) studied SSI effects on Single Degree of Freedom (SDOF) structures founded on visco-elastic halfspaces. Their results defined the basis for future design provisions developed by the Applied Techonology Council in 1978 and the US National Earthquake Hazards Reduction Program in 1997.  The results from some experimental research works carried out in the 1970?s and early 1980?s are the base of the approach that has been widely used in practice until now. In those studies specific types of piles were subjected to static lateral static or cyclic loading applied at the pile head. The results included non-linear load-deformation curves recommended for simulation of soil behavior in the near-field zone, called ?p-y curves?. Matlock (1970) proposed correlations for soft clays. Reese et al. (1975) proposed p-y curves for stiff clays, and O'Neill and Murchinson (1983) proposed p-y relationships in sands.   However, this approach has been found not accurate in several studies since the mid 1980?s. Murchison and O?Neill (1984) evaluated the reliability of p-y relations ?four types including the 3  API curves? in cohesionless soils based on full-scale tests, and found poor predictions in some cases. Another experimental study (Gazioglu and O?Neill 1984) showed that the API curves also had poor results predicting deflections and moments of piles in clayed soils.  Regarding the evaluation of soil-structure interaction effects based on seismic records, a comprehensive research work carried out at the University of California-Berkeley by Stewart et al. (1998) shows the influence of structural and geotechnical properties on the inertial interaction effect, expressed in terms of first-mode period lengthening and soil-foundation interaction damping. This work is based on empirical evaluation of seismic records from instrumented structures located in 58 sites, applying system identification techniques to get the dynamic properties of the assessed structures. The results show a strong influence of the structure-to-soil stiffness ratio on the inertial interaction. A secondary influence was related to the structure aspect ratio and to the foundation properties, defined as embedment ratio, type shape and non-rigidity.  Recent studies use simplified and continuum models that have been developed to reflect an overall response that approximates the real behavior of the soil-structure system. The most common method used to model the soil-pile system is Beam on a Non-linear Winkler Foundation (BNWF), where the pile is modeled as a beam and the lateral soil pressure acting on the pile is modeled using a nonlinear discrete spring?dashpot model (Rahmani, et al. 2012). The p-y curves are used as the non-linear load-deformation relations that define the springs.    4  The report released by the American Petroleum Institute (2007) related to fixed offshore platforms provides p-y curves is widely used in practice. The document includes non-linear p-y curves recommended for simulation of soil behavior in the near-field zone, defined for three types of soil: soft clay, stiff clay, and sands. These relations are based in the previously described works done by Matlock (1970), Reese et al. (1975), and O'Neill and Murchinson (1983).   The ASCE/SEI 41-06 provisions (ASCE 2007) include the evaluation of soil-structure interaction effects for the analysis in seismic rehabilitation of existing buildings. The document suggests that in most cases considering the SSI effects modifies the seismic demands on buildings by reducing the spectral acceleration, and also by increasing lateral displacements and secondary forces due to P-? effects. The geometry and properties of the foundation slab are expected to cause filtering and reduction of the shaking at the foundation if compared with the free-field motion. The increment in spectral acceleration related to the period lengthening effect is attributed to rare cases ?such as near-field motions and soft soil sites? where the SSI effect on the seismic response must be studied.  A recent study done by Rahmani, et al. (2012) evaluated the efficiency of API curves based on the results of four centrifuge tests. The peak responses of the system such as maximum bending moment and bending moment distributions at maximum pile head deflection were found to be seriously underestimated.   Another recent research work carried out by Pandey, et al. (2012) evaluates the response of 22 instrumented buildings to 99 earthquake records. The results indicate that the spectral response at 5  the foundation level is not always deamplified if compared with the free field record. About 30% of cases indicate spectral amplification, contradicting the expected behavior based on ASCE/SEI 41-06 provisions.   1.2 Motivation and purpose  The differences between the expected behavior and the actual response observed in the research works described in the previous section indicate that the models currently used in practice may be inadequate to represent the SSI effects on the seismic response. Based on this, the main motivation to carry out this research is the need to gain a better understanding of the effects of the soil-structure interaction in dynamic response of bridges to earthquakes. Since the data used in the analysis comes from real bridges, the findings and observations are intended to be used as a reference for current and future research projects aimed to provide better ways to model the SSI condition in bridges.  Complete non-linear finite element models which include soil, foundation and structure, implicitly include the soil-structure interaction condition; therefore no special considerations regarding SSI effects are necessary in such cases. The use of these models in engineering practice is very limited due to the complexity of the modeling and analysis process. On the other hand, simplified models that can be applied with software tools used in current engineering practice require the inclusion of SSI effects through adequate approximate procedures (Finn 2010).  6  However, even the simplified models to account for SSI effects may not be included in common practice. The additional considerations, complexities and uncertainties incorporated in the analysis and design processes are the main reason to neglect them. The idea that the SSI effects will reduce the structural demand is also a cause for not considering them in design stages.  This situation also motivates the completion of this research work. Thus one of the purposes of this work is to show the importance of SSI effects on the seismic response of instrumented bridges currently in service, based on the records from several earthquakes. The evaluation is oriented to highlight the differences of considering the free-field record ?with no SSI effects involved? or the column base record as the input motion to the structure.  1.3 Objectives  The main objective of this research work was to have a better of the soil-structure interaction effects on the dynamic behavior based on the analysis of seismic records from instrumented bridges.  The following specific objectives were outlined: ? Determine the effects of soil-structure interaction in the seismic response of bridges by analyzing motions at free field and column base from a group of instrumented bridges ? Build a database of available records from instrumented bridges that can be used for evaluation of soil-structure interaction effects 7  ? Develop a software that permits fast and efficient processing of seismic records to evaluate the soil-structure interaction effects  The tasks associated to these objectives are listed below: ? Characterize a selected group ground motions and compare their predominant direction of shaking with the main axes of the bridge ? Identify the modal properties of the evaluated bridges based on the seismic records ? Compare the spectral response of the structure estimated based on free field records with the estimated based on column base records ? Evaluate the dynamic properties of the soil-foundation-structure system through the duration of the ground motion for all the records ? Report the main findings and the corresponding conclusions  1.4 Scope of this study  ? This study is delimited by the following criteria: The analysis procedures used in this study are focused on highlighting the inertial type of SSI effects. The evaluation of kinematic effects would need a different approach not included in this research work. ? The analysis is done based on the response of the analyzed bridges to the earthquake excitation in the transverse direction only. The procedure however can be used for records in any direction. ? Finite elements models are not developed as part of this study. 8  Chapter  2: Theoretical background  2.1 Soil-Structure Interaction (SSI)  2.1.1 General description  The seismic excitation is transferred to the structure by the movement of the soil mass that supports the system. A free field site refers to a point on the ground located close enough to the bridge to represent the surrounding soil, but without being affected by the seismic response of the structure.  A simplified analysis of the situation would assume that the movement of that free field location would represent the shaking that actually takes place at the top of the foundation system. This would be applicable if the structure is founded on rock ?or very stiff soil? because the movements at different surrounding locations and at the base of the structure can be considered similar. For less stiff foundation materials the seismic response of the structure at ground level and the surrounding soil become dissimilar, and this difference will increase for structures founded on softer soils.   The mechanism that leads to this modification of the motion at the column base is called soil-structure interaction. The overall phenomenon has been split in two mechanisms called inertial and kinematic interaction, detailed in the next sections.   9  2.1.2 Inertial interaction  The inertial part of the soil-structure interaction phenomenon occurs when the inertial forces induced in the superstructure by the seismic excitation produce base shear and moments at the ground level, which cause differences in the displacements of the foundation system compared to the surrounding soil.       Figure 2.1. Displacements induced on a SDOF system under seismic excitation without inertial SSI effects  Figure 2.1 shows a Single Degree of Freedom (SDOF) system subjected to earthquake excitation.  The surrounding foundation soil is represented inside the rigid box, and the further soil media corresponds to the box itself. The repose condition is shown at the left side, and the structure?s position labeled as ?0?. The figure at the right represents the situation in which the earthquake excitation is acting in the system, for an instant when the lateral movement is oriented to the right side. The final position of the mass is labeled as ?1?, and its total displacement deformation (?T1) is the sum of the ground movement (?G) and the displacement induced by structural deformations (?S). In this case there are no inertial soil-structure interaction effects acting on the system. 10  On the other hand, Figure 2.2 shows the case of a SDOF system under seismic excitation which is affected by soil-structure interaction effects. The initial position is labeled as ?0? and shown in the same way than the previous image; the deformed conditions due to the seismic excitation are shown at the right side labeled from ?1? to ?3?.        Figure 2.2. Displacements induced on a SDOF system under seismic excitation with inertial SSI effects  The total displacement for this case (?T2) of the mass is defined as the sum of following components: ? Ground movement (?G). ? Horizontal movement of the column base (?H), which leads to position 1. ? Contribution from rocking at the base (?R), that leads to position 2. This component is caused by the rotation of the column base and transmitted to the structure. ? Contribution from swaying of the structure (?S), that leads to position 3. This component is induced by the structural deformations in the system.  11  This previously described mechanism implies deformations in the foundation soil. The consequence of this from the structural point of view is to affect level of fixity at the foundation, and deviating from the condition of fixed-base system.  Based on this, the model that would describe the SDOF system of the second case is shown in Figure 2.3. ?Ks? represents the stiffness of the column; ?m? corresponds to the mass of the structure, and ?Cs? to its damping ratio.       Figure 2.3. SDOF model considering SSI effects    The fixed base is substituted with Kelvin-Voigt elements able to include the stiffness (Ki) and damping (Ci) added by the soil in the system. The lateral movement is considered with the horizontal Kelvin-Voigt element, and the rotation of the base is modeled with the vertical element including a rocking spring. The structural properties are kept as they were described in the original model. The physical phenomenon is expected to have an amount of mass from the surrounding soil that is coupled with the structure during the dynamic response, which is not included in the model.  12  Based on this model two major effects would be expected in the seismic behavior: ? Period lengthening, due to more flexibility included in the system ? Increment of damping ratio, due to the energy dissipation ability of the mass of soil that interacts with the structural system.  2.1.3 Kinematic interaction  This phenomenon is caused by the fact that the foundation system and the surrounding-supporting soil are two different elements not rigidly connected trying to move together. This naturally produces differences between free-field and structural base motions. The possible mechanisms behind this effect are:  ? Base-slab averaging: the waves from the free-field motion that enter to the system are ?averaged? within the footprint area of the base slab due to the kinematic constraint applied by the slab moving as a rigid body. ? Embedment effect: reduction of seismic ground motions with depth for embedded foundations ? Scattering of seismic waves off of corners and asperities of the foundation   The main effect expected on the overall response of the structure is an increment on the damping ratio, due to the energy dissipation induced by the described processes.  13  2.2 System identification  2.2.1 General overview  The process of obtaining the modal properties of a structural system based on measured data of vibration from that structure is called system identification.  There many techniques used to perform the system identification process, which can be classified in three groups: frequency domain, time domain and combined frequency-time domains. Two well-known examples are the Stochastic System Identification ?applied in time-domain? and the Frequency Domain Decomposition. The latter was chosen to perform the analysis in this research work, and is described in the next section.  2.2.2 Frequency Domain Decomposition (FDD)  This technique is explained in detail by Brinker, et al (2000) and Brincker, et al (2001). The process is based on obtaining the Singular Value Decomposition (SVD) of the spectral matrix. According to this technique the unknown input y(t) and the measured response x(t) are related by the following expression:     (  )   ?(  )   (  ) (  )          (2.1)   14  In which:    (  ) is the Power Spectral Density (PSD) matrix [r x r] of the unknown inputs, where r is the number of inputs;     (  ) is a m ? m PSD matrix of the responses, where m is the number of the measured responses; and   (  ) is the Frequency Reponse Function matrix of the system. The superposed dash refers to complex conjugate.  The technique decomposes the Power Spectral Density function matrix of the measured outputs by using the Singular Value Decomposition process. The result of this is to decompose the spectral response into a set of single degree of freedom systems, where each one corresponds to an individual mode.   The Artemis Extractor software (Structural Vibrations Solutions 2013) was used in this research work to perform the modal identification, using the FDD technique. The software presents the results of the singular value decomposition in a set of curves shown together in the same chart. Each curve represents one of the singular values plotted for all the frequencies considered in the analysis. The peaks in the curves are associated to particular modes, which shapes are also displayed. The interpretation of the results is based on the modal shapes, frequency values and damping ratios for each case.   15  Chapter  3: Method of analysis  3.1 General overview  The method of analysis applied in this research work is intended to accomplish two main goals: ? Identify the soil-structure interaction effects on the seismic response of the bridge ? Understand the mechanism that lead to the observed behavior  The procedure is based on the comparison of properties between the signals recoded at free field and at the base of the column. The first stage involves the description of the bridges. Then the evaluation is performed following two approaches: comparative evaluation and time-frequency analysis.   The comparative evaluation involves using the information from the response spectra and Fourier spectra to assess the soil-structure interaction effects. The bridge and ground motions are described as an initial step. Then a modal identification process based on the records from the instrumented structure is performed, focusing on the properties of the fundamental mode. The response and Fourier spectra from the signals are evaluated to highlight amplifications or de-amplification of the response at the column base with respect to the free field. The final part consists of comparing the behavior observed in the previous sections with the properties of the free field record, also providing an overall summary of findings for the bridge.   16  The time-frequency evaluation section includes the analysis of spectra plots based on the wavelet transform technique that shows the Power Spectral Density (PSD) associated to each frequency throughout the duration of the record. The assessment is focused on comparing the peaks of PSD between the signals recorded at the deck, column base and free field. The objective is to notice similarities in the signals and amplifications/deamplifications. The frequencies of those peaks are also compared with the identified natural period.   The following sections describe in detail all the stages of the method.  3.2 Bridges description  The first step is intended to allow the understanding of the structural system and how it would behave under seismic excitation before the evaluation of the recorded signals. This is achieved by finding information from different sources.  The overall structural behavior can be outlined based on the structural geometry and properties, obtained from the structural drawings and/or sensor layouts. The following aspects play an important role in this step: ? Span support: continuous, simple, cantilever, cable stayed, etc. ? Type of superstructure: beam, truss, mixed, etc. ? Connectivity to abutments: monolithic, simply supported, rocker bearings, etc. ? Main geometry: height, length, number of spans, etc. ? Characteristics of bents: structural system, lateral stiffness, angle of skeweness, etc. 17  ? Foundation system: described based on the type ?i.e. shallow, deep or mixed? and the material ?e.g. concrete, steel, wood, etc. ? of the foundation system.  3.3 Ground motions description  The first characteristic reported from the ground motions are the peak ground acceleration, velocity, and displacement values. Additional parameters of the free-field signal can also be obtained to understand the input motion, such as Arias Intensity and significant duration.  The location of the bridge is reported along with the relative location of the epicenters of the ground motions. Any georeferenced map service ?online or offline? can be used to visualize all this by entering the coordinates of the bridge and epicenters.  The ?angle of attack? of the ground motion is an important factor to consider. This is defined from the main direction of shaking, which would be the predominant trend observed in orbit plots that show the two horizontal components of the free field records. This direction is compared with the main structural axes of the bridge. The intensity of the ground motion is also an important factor.     18  3.4 Comparative analysis  3.4.1 Overview  The comparative analysis is focused on highlighting the inertial effects of the response at the superstructure level on the response at the column base during the seismic excitation. This is done by comparing the spectral response calculated from the free field record ?not affected by the structural behavior? with the response from the column base record. The steps of this procedure are detailed in the following sections.  3.4.2 System identification  The first step of the comparative analysis is to identify the dynamic properties of the bridge based on the seismic records, for each ground motion included in the analysis. The results from this analysis allow a more accurate characterization of the bridge seismic response resulting in crucial information that supports the conclusions and recommendations for each case.  The modal identification procedure used in this research work is based in the Frequency Domain Decomposition, as it was developed in Chapter 2, and the Artemis Extractor 4.1 software was used to perform the analysis.   19  The data obtained from the databases included the recorded signals from instruments located on the deck, pier bases, and free field locations. The signals from the deck were loaded in the software and used to apply the system identification process.   The fundamental mode and estimated damping ratio in the transverse direction were then obtained based on the Singular Value Decomposition charts, by identifying prominent peaks and evaluating their corresponding modal shapes. The software reports for each selected frequency the corresponding damping factor associated to that modal shape.  3.4.3 Response spectra evaluation  This step represented the core of the evaluation process. The main purpose is to compare in both qualitative and quantitative ways the spectral values obtained at the column base and free field records.  The absence of soil-structure interaction effects would imply that the movement of the free field (a point located close to the bridge) would match the motion of the column base (the input motion that affects the bridge) therefore the signals would be the same. Thus the differences that actually occur between the records are associated to the interaction effects that take place in the soil-foundation-structure system.  20  This evaluation was done by determining the differences between the response spectra from both signals. The quantitative evaluation uses the Spectral Ratios (SR) as an index, calculated as shown in Equation 3.1:                             (3.1)  The S factor represents the spectral response ?acceleration, velocity, or displacement? for each period of the response spectra. The main factor to consider is the one calculated for the identified natural transverse period of the bridge. A spectral ratio of 1.0 is set as a threshold: ratios over that limit indicate amplification of the response and below the limit indicate de-amplification. The qualitative evaluation involves the analysis of the shape of both spectra. The presence of peaks is reported, as well as the amplification or reduction of the response at the column base for any interval of periods.  3.4.4 Fourier spectra evaluation  This step in the analysis intends to evaluate the differences in terms of frequency content between column base and free field signals. The amplification or deamplification of Fourier amplitudes at the column base with respect to the free field was reported, including the frequencies related to that phenomenon.   21  The other purpose of this stage is to identify two important frequencies in signal recorded at the free field: the frequency of peak Fourier amplitude and the mean frequency. The first one corresponds to larger Fourier amplitude value in the spectrum. The mean frequency is calculated based on the research work done by Rathje, et al (2004) by using the equation described below:     ?         ?       for 0.25 Hz ? fi ? 20 Hz,  with ?f  ?  0.05 Hz  (3.2)  where: Ci = Fourier amplitude coefficients fi = discrete Fast Fourier Transform frequencies between 0.25 Hz and 20 Hz ?f  = frequency interval used in the Fast Fourier Transform computation  Those values are compared with the fundamental transverse frequency of the bridge identified for that ground motion. The signals from the deck can be included to compare the frequencies of the energy peaks with the identified natural frequency of the bridge. All this parameters are also expressed as periods for future comparisons.  An example plot used for Fourier spectra evaluation is shown in Figure 3.1. The spectrum and parameters in blue represent the free field, and the black color represents the column base. The three parameters used for comparison are described below: ? Peak Point ? Free Field: point of peak Fourier amplitude for the free field spectrum. The number in the box located close to the point is the corresponding frequency in Hz. 22  ? Mean Frequency ? Free Field: calculated based on the mean period previously described. ? Identified fundamental frequency: obtained as it was explained in section 3.4.2.        Figure 3.1. Example graphics used for Fourier spectra evaluation  3.4.5 Free field record vs. spectral analysis results comparison  This last subsection is intended to gather the results from the previous sections and identify possible trends of the observed response.  For bridges with more than one recorded motion available, the results for the previous analysis ?i.e., identified fundamental period, period of peak Fourier amplitude for free field record, mean period of free field record, and spectral ratios? are compared with the PGA of the corresponding ground motion. The peak Fourier amplitude for each record is also included in the comparison.   An important parameter used in the analysis is the period difference (Tdif). This was calculated as an absolute value as stated in Equation 3.3:  23       |              |         (3.3)  Where Tfree-field refers to the period used to represent the free field record, which can be the period of peak Fourier amplitude (Tp) or the mean period (Tm) described in section 3.4.4.  An example of the results of this kind analysis is shown in Figure 3.2. The three graphs are shown together to facilitate the identification of trends.                 Figure 3.2. Example of comparison between free field record properties and spectral analysis results 24  The main objective of this part is to identify the trends in the spectral ratio, and how it is related to the trends observed in the other parameters. The example graphs in the figure show higher spectral ratios for lower period differences, and no clear trend with respect to PGA.  3.5 Time-frequency domain analysis  3.5.1 Overview  The time frequency analysis was outlined as a more advanced stage in the method of analysis. The main purpose is to evaluate both seismic response parameters and soil-structure interaction effects through the duration of the shaking. The following section describes the procedure.  3.5.2 Evaluation of time-frequency spectra  The spectra were obtained based on the Continuous Wavelet Transform (CWT) technique described in the work done by Torrence and Compo (1998). The Morlet function was used as the mother wavelet for all cases. The color scale in all the plots represents the values of Time-Integral Squared Amplitude Power (PSD-TISA), defined as the integral under the curve defined by the square of the amplitudes from the CWT.  The example plot in Figure 3.2 shows the PSD-TISA ?in color scale? corresponding to each time-frequency pair. Thus a vertical line shows the PSD-TISA of the signal for each frequency in that instant of time, and a horizontal line shows how the PSD-TISA changes for a fixed 25  frequency through the duration of the record. The acceleration time history of the record is located at the bottom of the spectrum. To facilitate the interpretation process, the identified natural frequency is highlighted with a horizontal white line, and the location of the recorded signal in the structure is shown with a corresponding icon.            Figure 3.3. Example of 2D wavelet time-frequency spectrum  The spectra are subjected to both quantitative and qualitative analysis. The quantitative evaluation is intended to compare the frequencies that correspond to the energy peaks (named Fenergy-peak) to the natural identified frequencies of the structure (Fidentified). This is done through an index named Frequency Ratio (FR) based on the signals recorded at the center of the deck, calculated as follows:  26                             (3.4)  An index of FR=1 indicates that highest power values are related to the frequency of the fundamental mode, as it would be expected for most types of bridges. Different FR values would indicate that the response is not governed by the fundamental mode of the structure, and further analysis should be done base on the results from the system identification process described in section 3.4.2.  The qualitative evaluation is done first by identifying the peaks of PSD-TISA and its distribution through the duration of the record, which constitute a kind of ?fingerprint? for that signal. These characteristics are detailed for signals at the deck, column base, and free field, in order to identify similarities between the records. The presence of similar signals suggests the effect of one element of the system on the seismic response of the other.  The soil-structure interaction effects that can be highlighted in this analysis are: ? Amplification or deamplification of the response at the column base with respect to the free field record throughout the duration of the record, based on the values of PSD-TISA. ? Presence of peaks of PSD-TISA at the same frequency in two signals. This is used to evaluate similarities between the signals recorded at the column base and free field, and between center of the deck and the column base. This comparison helps to understand the physical phenomenon that leads to the seismic behavior for each case.  27  Chapter  4: BRIDGE-SSI System  4.1 Overview  This chapter describes the Bridge Records Information Database for General Evaluation of Soil-Structure Interaction (BRIDGE-SSI) created as part of this research work. The system is divided in two separate tools: ? BRIDGE-SSI Seismic Records Database Website: developed as an online databank used to store and manage the available records from instrumented bridges; ? BRIDGE-SSI Software: stand-alone program used to perform the comparative analysis of the stored seismic records.  Both website and software were developed using java as the programing language. The following sections describe each tool, including the main criteria used to develop their functions.  4.2 BRIDGE-SSI Seismic Records Database Website  This online tool can be accessed through the research project website under the ?Research Results? section, located at http://ssi-civil.sites.olt.ubc.ca.  The information is separated in three groups based on the source of the seismic records: British Columbia, California (from CESMD) and other sources. The user has to choose a source before entering to the database.   28  The following screen ?see Figure 4.1? allows to enter the search criteria that will be used to get records from the database. The user has to submit the information to get the list of records.             Figure 4.1. Search screen in BRIDGE-SSI Seismic Records Database Website   In the next step the website shows the list of records that matched the submitted criteria, as it is shown in Figure 4.2. The characteristics of each record are included in the list with a button for downloading the information.     29             Figure 4.2. Search results list in BRIDGE-SSI Seismic Records Database Website  Each file corresponds to one earthquake recorded by one station. A complete guide for preparing the records is included in this document in Appendix A. This guide is also in the software manual with the detailed procedure for uploading the information to the database.  The downloaded file corresponds to a ?.zip? compressed folder that has to be extracted to be readable by the BRIDGE-SSI Software. The following files are included in the folder: ? Information file: including the basic information of the bridge and the earthquake ? Image files: prepared to show the bridge location, pictures, sensor layouts, drawings, and soil profiles through the software interface. 30  ? Channel records: recorded signals from the instruments located at the bridge, labeled based on the numbering code from the sensor layout  It is recommended to save the folder in a directory located close to the root, to avoid a long address that may affect the visualization of the data.  4.3 BRIDGE-SSI Software  This tool was designed as a signal processing software that allows comparing the properties of three records at the same time, while showing the basic characteristics of the bridge and the ground motion. The following sections describe the main functions of the software and their corresponding calculation algorithms.   4.3.1 Bridge description and records  The bridge description section is intended to provide information about the structural system that is being analyzed. The first part contains basic information of the bridge, including location, pictures, structural elements description, and classification. The structural drawings and geotechnical (soil profile) drawings are also included in this part, as it is shown in Figure 4.3.  The records information section shows the sensor layout and available channels for the station. The ground motion basic information is also included, as it is shown in Figure 4.4.  31             Figure 4.3. Bridge information section in BRIDGE-SSI Software.           Figure 4.4. Records section in BRIDGE-SSI Software. 32  4.3.2 Time-history  This section has three purposes: to select the channels to be analyzed, to allow the pre-processing of the signals, and to show the time history plots of the processed signals. After a channel is selected from the list the corresponding time-history plot appears on the screen. Then the signal processing options have to be chosen before selecting the next channel. The related windows and options are shown in Figure 4.5.           Figure 4.5. Time-history section in BRIDGE-SSI Software.  The baseline correction is done in time domain based on polynomial formulas as they are described in Boore (2001). The filtering algorithm is shown in Figure 4.6. The resulting screen with three time history plots after baseline correction in the next figure. All acceleration, velocity and displacement plots are available. 33              Figure 4.6. Filtering algorithm implemented in BRIDGE-SSI Software.          Figure 4.7. Time-history plots in BRIDGE-SSI Software. 34  4.3.3 Response spectra  The response spectra are calculated based on Newmark?s method for numerical evaluation of dynamic response. Chopra (2006) in Chapter 5 describes this calculation method.  4.3.4 Frequency domain  The calculations in this section are done based on the Fast Fourier Transform technique. The Fourier and Power spectra are shown in this section.  4.3.5 SSI effect  This section is intended to help to highlight the soil-structure interaction effect by calculating the spectral ratios between the signals that were loaded before. If free-field and column-base signals were included in the analysis, the spectral ratios will help to visualize amplification or de-amplification of the response.  4.3.6 Time domain  This section allows first the evaluation of the energy of the signal through the duration of the record based on the Arias Intensity parameter. The parameter is calculated based on the acceleration values (a) for any value of time (t) by using Formula 4.1:  35          ? ? ( )?         (4.1)  Based on this, the percentage of the total Arias Intensity is calculated as:      ( )             ? ? ( )?         ? ? ( )?            (4.2)  Significant duration is then calculated by using Formula 4.3:            (4.3)  where     is the time value for which        ; and      is the time value for which         The bracketed duration is also calculated as the total time elapsed between the first and the last excursions of a specified level of acceleration (for example, 5% of PGA), using absolute values of acceleration comparison.  The last part in this section contains the orbit plots. This is a way to show the directionality of the ground motion when the horizontal components of the free field records are loaded in the software. 36  Chapter  5: Bridges description  5.1 Bridges selection criteria  The BRIDGE-SSI database contains several cases of instrumented bridges. The selection criteria for cases of study were based on the scope and objectives of this research work.  The chosen bridges to be analyzed had instrumentation on free field and column base locations, as well as several instruments distributed along the deck, including the pier where the column base instrument is located.  5.2 Bridges description  The following sections describe the selected cases for analysis. The instrumentation in all the stations is managed by the California Strong Motion Instrumentation Program (CSMIP). Their basic information and records were originally posted online by the California Earthquake Strong Motion Database (CESMD). Further information can be found in both CESMD (http://www.strongmotioncenter.org/) and BRIDGE-SSI (http://ssi-civil.sites.olt.ubc.ca/) websites.    37  5.2.1 Meloland Overpass  The Meloland Overpass is an integral-type two spans bridge built in 1972 near El Centro, California, United States of America, as part of the Highway 8. The following figure summarizes the structural characteristics and instrumentation for this case.            Figure 5.1. Elevation and plan view of Meloland Overpass including sensors layout (from http://www.strongmotioncenter.org/)  The superstructure consists of a concrete box girder connected monolithically to the abutments, supporting a 63.4m long and 10.4m wide deck. It has two same-sized spans separated by a pier with a single circular reinforced concrete column. The foundation system consists of a timber pile group supporting a square concrete pile cap under the central pier, located on a deep 38  alluvium with an average shear wave velocity Vs30=192 m/s. Abutments are also supported by timber piles. The instrumentation setup was installed in 1978 and upgraded in 1991. The current set up includes 29 accelerometers on the bridge and 3 accelerometers at a free-field site.  5.2.2 Via California  The Via California bridge is a 6 spans reinforced concrete straight bridge.  It was constructed in 1960 as part of the Interstate 5, near Capistrano Beach, California, United States of America. The bridge sketches and instrumentation are shown in Figure 5.2.             Figure 5.2. Elevation and plan view of Via California Bridge including sensors layout (from http://www.strongmotioncenter.org/) 39  The deck has a total length of 134.4m and 15.9m width. The superstructure is a 6-cell concrete box girder with cantilever abutments, supported by reinforced concrete bents with 2 columns on spread footing foundations, located on alluvial stratum. The bridge was retrofitted in 1996 and instrumented in 1999.  5.2.3 Highway 395 Bridge  The Highway 395 Bridge is located near Mammoth Lakes, California, United States of America. It was constructed in 1969 for the Highway 395 as the overpass on the South Landing Road.              Figure 5.3. Elevation and plan view of Via California Bridge including sensors layout (from http://www.strongmotioncenter.org/) 40  Figure 5.3 summarizes the structural characteristics and instrumentation. The deck bridge has a 61.9m straight longitudinal axis with two spans and a skewed supporting bent.  The superstructure is a continuous concrete box girder, with diaphragm abutments. The middle bent has 2 circular reinforced concrete columns with spread footings, founded on alluvial stratum. The bridge was instrumented in 1995.  5.2.4 Alviso Overpass (Bridges K and L)  The Alviso Overpass Bridges are located in Santa Clara, California, United States of America. They were constructed in 1994 as part of the Highway 237 to allow crossing over Lafayette Street and the Southern Pacific Railroad Track.   The Bridge K has 6 spans supporting a straight axis deck of 136.2m length and 8.38m width. The superstructure consists of a continuous concrete box girder without intermediate hinges, seated on abutments with elastomeric bearings. The bents have one rectangular concrete column with flares in the upper portion. The foundation system corresponds to pre-stressed concrete piles, located over a deep alluvium with a shear wave velocity of Vs30 = 155 m/s.  The Bridge L is located parallel to the other structure, also having 6 spans and supporting a straight axis deck of 132.9m length and 18.1m width. The superstructure consists of a continuous concrete box girder without intermediate hinges, seated on abutments with elastomeric bearings. The bents have two rectangular concrete columns with flares in the upper portion. The 41  foundation system is also formed by pre-stressed concrete piles, having a foundation material similar to the other bridge.  The variable angle of skewness is a noticeable characteristic in both structures. The instrumentation for both bridges was set up in 1995 as a single station, sharing the free field channels. The main structural characteristics and instrumentation are shown below.                Figure 5.4. Elevation and plan view of Alviso Overpass ?Bridges K and L? including sensors layout (from CESMD: http://www.strongmotioncenter.org/) 42  5.2.5 Murray Road Bridge  The Murray Road Bridge is located in Arcata, California, United States of America. It was constructed in 1964 as part of the Highway 101. The structural characteristics and instrumentation are summarized in the figure below.                Figure 5.5. Elevation and plan view of Murray Road Bridge including sensors layout (from http://www.strongmotioncenter.org/)  43  The bridge has four spans with a total length of 55.47m. The superstructure includes six concrete T-beams supported on open end cantilever abutments and concrete bents. The T-beams are also connected to the abutments with rocker bearings. Each bent has two rectangular concrete columns. The foundation system consists of spread footings, located on alluvium soil. The instrumentation was set in 1995, including 9 accelerometers on the bridge and 3 accelerometers at a free-field site.                  44  Chapter  6: Results  This chapter contain the results of the evaluation procedure for all of the analyzed bridges, with the corresponding comments and remarks. These results have been obtained using the methodology detailed in Chapter 3. The results are presented in two main sections: an initial comparative evaluation and the time-frequency analysis. Each section has its overall summary.  6.1 Comparative evaluation of selected bridges  The main objective of the comparative evaluation is to identify and extracting the main trends that are evident in the comparison between free field and column base records for the analyzed ground motions. These results are presented separately for each of the analyzed bridges.  6.1.1 Meloland Overpass 6.1.1.1 Description of the ground motions  Three example records from this station were chosen for the detailed analysis: the two cases with higher PGA and one of the cases with low PGA. Their key characteristics are listed in Table 6.1.   No. Epicenter Location / Date Epicentral Distance (Km) PGA (g) PGV (cm/s) PGD (cm) 1 Calexico 22May2010 35.2 0.031 0.670 0.068 2 Calexico 30Dec2009 41.2 0.174 16.610 3.899 3 Calexico 04Apr2010 58.9 0.213 19.050 13.928 Table 6.1.    Ground motions included in the analysis of the Meloland Overpass 45  The orbit plot for the Calexico 22May2010 ground motion included in Figure 6.1 shows oscillation without any outstanding direction. The Calexico 30December2009 motion has a clear trend of shaking almost exactly in the transverse direction. The Calexico 04April2010 ground motion shows important movements in both transverse and longitudinal directions. The amplitude of the shaking was related to the intensity of the ground motion as is expected.               Figure 6.1. Displacement orbit plots of the ground motions included in the analysis. Meloland Overpass    46  6.1.1.2 Modal identification results  Table 6.2 presents parameters corresponding to the fundamental mode in the transverse direction.  No. Ground Motion PGA (g) Frequency (Hz) Period (sec) Period Variation (%) Damping Ratio (%) Damping R. Variation (%) 1 Calexico 22May2010 0.031 3.861 0.259 - 0.49 - 2 Calexico 30Dec2009 0.174 3.472 0.288 11.2 0.84 71.4 3 Calexico 04Apr2010 0.213 3.663 0.273 5.4 0.48 2.0 Table 6.2.    Identified fundamental mode parameters in the transverse direction. Meloland Overpass  From Table 6.2 the following observations can be made. ? The identified period of the structure tends to be larger for higher intensity motions.  ? The damping ratio is similar for two cases, and shows a significantly higher value for the Calexico 30Dec2009 earthquake. A higher fundamental period was also identified in this case.  The observed behavior is very much associated with the Integral Abutment Bridge type of this case. In these structures there is a monolithic fixing between the bridge deck and the abutments, forming a coupled system in which both structure and soil in the approaching embankments have an important influence in the dynamic behavior of the bridge. The noted increase in the fundamental period of the system is associated with the amount of soil mass mobilized during the shaking (Carvajal 2011). The added soil mass also increases the ability of the system to dissipate energy, resulting in higher damping ratios.  47  6.1.1.3 Response spectra analysis  The resulting spectra for the three ground motions analyzed in this case are included in Figures 6.2 to 6.3 shown below:       Figure 6.2. Response spectra for Calexico 22May2010 earthquake (PGA=0.03g), Meloland Overpass      Figure 6.3. Response spectra for Calexico 30Dec2009 earthquake (PGA=0.17g), Meloland Overpass      Figure 6.4. Response spectra for Calexico 04Apr2010 earthquake (PGA=0.21g), Meloland Overpass 48  The shapes of the plotted acceleration, velocity and displacement response spectra at the column base and free field are similar. The spectral response at the column base tends to be lower than the response at free field for lower periods, up to a value slightly larger than the natural period. From that point the situation is the opposite, showing larger response at the column base.  The period of the highest peak in the response spectra at the column base is larger than the fundamentental period of the bridge for all records. This is result of the dynamic behavior of the system in this particular case, where the seismic response is governed by the bridge-embankment system instead of the structure itself.  The resulting spectral ratios of column base vs. free field are summarized as follows.  No. Record PGA Period SaBASE / SaFF SvBASE / SvFF SdBASE / SdFF (g) (sec) 1 Calexico 22May2010 0.031 0.259 0.87 0.85 0.86 2 Calexico 30Dec2009 0.174 0.288 0.66 0.61 0.66 3 Calexico 04Apr2010 0.213 0.273 0.71 0.66 0.71 Table 6.3.    Spectral ratios for the identified natural transverse period. Meloland Overpass  The ratios of spectral response of column base vs. free field plotted in Figure 6.5 show the following trends: ? Generally, the motions at the column base are de-amplified when compared with the free field records.  ? The peak in the spectral ratio plots corresponds to a period larger than the identified natural transverse period.   49           Figure 6.5. Spectral ratios for the three analyzed earthquakes, Meloland Overpass  6.1.1.4 Fourier spectra analysis  Figures 6.6 to 6.8 show the Fourier spectra plots for the analyzed records in this case. The following main trends were observed: ? The peak locations in the frequency content of both signals are similar, except for the natural frequency and its surroundings where the column base show less amplitude. This effect is more pronounced for higher intensity shaking. ? From the surroundings of the fundamental frequency to higher values the column base spectra show lower amplitudes. The reverse is true for the rest of the spectrum in all cases. ? The frequency of the highest peak is always lower than the mean frequency.   50        Figure 6.6. Fourier spectra for Calexico 22May2010 (PGA=0.03g), Meloland Overpass        Figure 6.7. Fourier spectra for Calexico 30Dec2009 (PGA=0.17g), Meloland Overpass        Figure 6.8. Fourier spectra for Calexico 04Apr2010 (PGA=0.21g), Meloland Overpass 51  6.1.1.5 Free field record vs. spectral analysis results comparison  Table 6.4 and Figure 6.9 summarize the results for all the studied ground motions in this bridge, including those not described in the detailed analysis of the previous sections.   ? Ground Motion Identified Fundamental Period [ Tn (sec) ] Free Field Record Spectral Acc.  Ratio Mean Period [ Tm (sec) ] Peak Amplitude Point Fourier Amplitude for Tn Location PGA (g) Period [ Tp (sec) ] Fourier Amplitude Borrego Springs 07Jul2010 0.012 0.269 0.622 1.260 6.576 4.644 0.749 Calexico 20Nov2008 0.017 0.297 0.709 0.987 6.014 2.102 0.841 Calexico 22May2010 0.031 0.259 0.230 0.259 7.398 7.398 0.865 Calexico 30Dec2009 0.174 0.288 0.709 0.987 56.552 28.192 0.657 Calexico 04Apr2010 0.213 0.273 0.580 0.964 118.900 67.914 0.714 Table 6.4.    Free field record properties vs. spectral analysis results, Meloland Overpass  The following trends are observed in the analyzed data: ? The first three ground motions correspond to low intensity shaking, with similar values of both PGA and peak Fourier amplitude. In these cases the spectral ratio is larger when the period of the peak point is closer to the fundamental period of the structure. The same behavior is observed when the mean period approaches the period of the structure. ? The last two ground motions in the list contain higher PGAs (0.174g and 0.213g). In these cases the distance between the period of the peak point and the period of the structure is similar. The mean period is closer for the case with higher PGA, which shows the higher response ratio. This case also has higher peak Fourier amplitude.  52  The spectral response ratio seems to be sensitive to both PGA and period difference, either using the peak point period or the mean period to represent the free field motion. Lower period differences tend to be associated with higher response ratios. The peak Fourier amplitude also seems to have an influence in the response. Higher peak Fourier amplitudes seem to be related with higher response ratios.                 Figure 6.9. Free field record properties vs. spectral analysis results, Meloland Overpass  53  6.1.2 Via California Bridge 6.1.2.1 Description of the ground motions  Figure 6.10 shows the displacement orbit plots for all the ground motions and their relative orientations to the bridge.               Figure 6.10. Displacement orbit plots of the ground motions included in the analysis, Via California Bridge  The key characteristics of the ground motions analyzed in this case are listed in Table 6.5. 54  No. Epicenter Location / Date Epicentral Distance (Km) PGA (g) PGV (cm/s) PGD (cm) 4 Borrego Springs 17Jul2010 110.1 0.006 0.600 0.140 5 Calexico 04Apr2010 337.1 0.020 4.000 2.350 6 Chino Hills 29Jul2008 52.1 0.023 1.140 0.280 Table 6.5.   Ground motions included in the analysis of the Via California Bridge  The Borrego Springs 07Jul2010 plot show oscillation with slight tendency to the direction of the free-field recording channels, but without a clear trend. A similar situation is seen for the Chino Hills 29Jul2009 ground motion, with a slight tendency of shaking in the longitudinal direction of the bridge. The Calexico 04April2010 ground motion looks to be mainly oriented in the longitudinal direction.  6.1.2.2 Modal identification results  In a similar fashion to Table 6.2, Table 6.6 presents parameters corresponding to the fundamental mode in the transverse direction. The variability of the results for the natural transverse periods is low, but larger differences are seen for the damping ratio. The only apparent trend in this case is damping ratio with period of variation.   No. Ground Motion PGA (g) Frequency (Hz) Period (sec) Period Variation (%) Damping Ratio (%) Damping R. Variation (%) 4 Borrego Springs 07Jul2010 0.006 2.688 0.372 - 0.83 - 5 Calexico  04Apr2010 0.020 2.518 0.397 6.8 2.52 203.6 6 Chino Hills 29Jul2008 0.023 2.623 0.381 2.5 1.75 110.8 Table 6.6.    Identified fundamental mode parameters in the transverse direction for the Via California Bridge 55  6.1.2.3 Response spectra analysis  Similarly to the presentation of the results for the Meloland Overpass, the spectra for the three ground motions analyzed in this case are summarized in Figures 6.11 to 6.13. The resulting spectral ratios for the natural transverse period are presented in Table 6.7 and Figure 6.14.       Figure 6.11. Response spectra corresponding to the Borrego Springs 07Jul2010 earthquake (PGA=0.006g), Via California Bridge       Figure 6.12. Response spectra corresponding to the Calexico 04Apr2010 earthquake (PGA=0.020g), Via California Bridge    56       Figure 6.13. Response spectra corresponding to the Chino Hills 29Jul2008 earthquake (PGA=0.023g), Via California Bridge  No. Record PGA Tn SaBASE / SaFF SvBASE / SvFF SdBASE / SdFF (g) (sec) 1 Borrego Springs 17Jul2010 0.006 0.372 2.35 2.13 2.36 2 Calexico 04Apr2010 0.020 0.396 2.93 3.56 2.93 3 Chino Hills 29Jul2008 0.023 0.381 1.92 1.75 1.92 Table 6.7.    Spectral ratios for the identified natural transverse period, Via California Bridge          Figure 6.14. Spectral ratios for the three analyzed earthquakes, Via California Bridge  57  The following remarks summarize the observed characteristic of the column base / free field spectral response evaluation: ? The spectral response at the column base is amplified if compared with the free field for the natural period in all cases.  ? For both Calexico 04Apr2010 and Chino Hills 29Jul2008 ground motions the amplification of the response at the column base was found for almost the entire spectrum. This starts from some apparent threshold period lower than the fundamental period and moves towards larger values. The Borrego Springs 17Jul2010 ?very low intensity shaking? shows that condition for the values that surround the natural period. ? The shapes of the spectra at the column base and free field are similar for all cases. The Chino Hills 29Jul2008 ground motion shows a second peak in the spectra for a period value larger than the natural period. This is also reflected in the spectral ratio. ? The spectral ratio plots of Borrego Springs 17Jul2010 and Calexico 04Apr2010 ground motions show very similar shapes.  ? The peak in the spectral ratio plots corresponds to the identified natural transverse period.   6.1.2.4 Fourier spectra analysis  The following main trends are observed in the Fourier spectra, shown in Figures 6.15 to 6.17: ? The free field records from Borrego Springs 17Jul2010 and Chino Hills 29Jul2008 show a similar distribution of peaks in the Fourier amplitude. For both records the predominant peak of amplitude in the signal recorded at the column base corresponds to the fundamental frequency. 58  ? The free field record from the Calexico 04Apr2010 ground motion shows a predominant peak of amplitude at F=0.63Hz. The same peak is noticed in the column base record, in addition to the peak that corresponds to the fundamental frequency. ? The records from the column base in all cases show higher amplitudes from the lower frequencies up to the region of the natural frequency. The opposite effect is shown for the rest of the spectrum.        Figure 6.15. Fourier spectra for Borrego Springs 07Jul2010 (PGA=0.006g), Via California Bridge        Figure 6.16. Fourier spectra for Calexico 04Apr2010 (PGA=0.020g), Via California Bridge  59         Figure 6.17. Fourier spectra for Chino Hills 29Jul2008 (PGA=0.023g), Via California Bridge  6.1.2.5 Free field record vs. spectral analysis results comparison  Table 6.8 and Figure 6.18 summarize the results for all the studied ground motions from this bridge.  The following trends are observed in the data: ? For both Borrego Springs 17Jul2010 and Chino Hills 29Jul2008 the period of the ground motion is very close to the fundamental period of the structure. ? The Calexico 04Apr2010 case shows the highest spectral response ratio, even though it has the largest period difference of all cases. The Fourier spectrum for the free field record shows a predominant peak of amplitude at a frequency of 0.63Hz, and the system identification results for this bridge associate the transverse translation mode to a frequency of 0.62Hz. From this it can be inferred that the peak in the spectral ratio is most likely associated with a coupled modal response. The significantly greater amplitudes observed in the Fourier spectra for this record when compared with the other cases is also noteworthy. ? Higher peak Fourier amplitudes seem to be related with higher response ratios. 60  Ground Motion Identified Fundamental Period [Tn (sec)] Free Field Record Spectral Acc.  Ratio Mean Period [ Tm (sec) ] Peak Amplitude Point Fourier Amplitude for Tn Location PGA (g) Period [ Tp (sec) ] Fourier Amplitude Borrego Springs 17Jul2010 0.006 0.372 0.552 0.602 3.854 1.470 2.351 Calexico 04Apr2010 0.020 0.396 1.160 1.575 37.637 7.655 2.928 Chino Hills 29Jul2008 0.023 0.381 0.366 0.431 10.615 8.145 1.919 Table 6.8.    Free field record properties vs. spectral analysis results, Via California Bridge.                 Figure 6.18. Free field record properties vs. spectral analysis results, Via California Bridge  61  6.1.3 Highway 395 Bridge 6.1.3.1 Description of the ground motions  All the ground motions analyzed in this case have low intensity shaking. Their key characteristics are listed in Table 6.9. No. Epicenter Location / Date Epicentral Distance (Km) PGA (g) PGV (cm/s) PGD (cm) 7 Qualeys Camp 18Sep2004 48.7 0.014 0.390 0.070 8 Toms Place 26Nov2006 16.1 0.015 0.250 0.010 9 Mammoth Lakes 12Jun2007 12.8 0.051 0.830 0.100 Table 6.9.    Ground motions included in the analysis of the Highway 395 Bridge              Figure 6.19. Displacement orbit plots of the ground motions included in the analysis, Highway 395 Bridge 62  Figure 6.19 shows the displacement orbit plots for all the ground motions and their relative orientation to the bridge. The plots corresponding to Qualeys Camp 18Sep2004 and Mammoth Lakes 12Jun2007 seem to be oriented in the transverse direction of the bridge. The Toms Place 26Nov2006 ground motion show important shaking in the direction perpendicular to the skewed bents.   6.1.3.2 Modal identification results  The identified modal parameters correspond to the first mode in the transverse direction.  No. Ground Motion PGA (g) Frequency (Hz) Period (sec) Period Variation (%) Damping Ratio (%) Damping R. Variation (%) 7 Qualeys Camp 18Sep2004 0.014 5.379 0.186 - 0.56 - 8 Toms Place 26Nov2006 0.015 5.202 0.192 3.4 0.54 -3.6 9 Mammoth Lakes 12Jun2007 0.051 5.342 0.187 0.7 0.36 -35.7 Table 6.10.    Identified fundamental mode parameters in the transverse direction for the Highway 395 Bridge  There is low variability in the identified fundamental transverse frequency. The values of identified damping ratio seem to be related to the intensity of shaking.  6.1.3.3 Response spectra analysis  The resulting spectra for the three ground motions analyzed in this case are shown below.  63       Figure 6.20. Response spectra for Qualeys Camp 18Sep2004 (PGA=0.014g), Highway 395 Bridge       Figure 6.21. Response spectra for Toms Place 26Nov2006 (PGA=0.015g), Highway 395 Bridge       Figure 6.22. Response spectra for Mammoth Lakes 12Jun2007 (PGA=0.051g), Highway 395 Bridge  The resulting spectral ratios for the natural are summarized in Table 6.11, and shown in Figure 6.23 .  64   No. Record PGA Tn SaBASE / SaFF SvBASE / SvFF SdBASE / SdFF (g) (sec) 7 Qualeys Camp 18Sep2004 0.014 0.186 1.51 1.70 1.51 8 Toms Place 26Nov2006 0.015 0.192 2.76 2.82 2.76 9 Mammoth Lakes 12Jun2007 0.051 0.187 1.59 1.37 1.60 Table 6.11.    Spectral ratios for the identified natural transverse period, Highway 395 Bridge          Figure 6.23. Spectral ratios for the three analyzed earthquakes, Highway 395 Bridge  The following remarks summarize the observed characteristic of the column base / free field spectral response evaluation: ? The spectral ratios for the natural period show amplification at the column base in all cases.  ? For both the Toms Place 26Nov2006 and Mammoth Lakes 12Jun2007 ground motions the amplification of the response at the column base was found for almost the entire spectrum. The amplification starts from a threshold value below the natural period and extends for 65  almost all larger periods studied. The Qualeys Camp 18Sep2004 record shows amplification in the region of the natural period. ? The shapes of the spectra at the column base and free field are similar for all cases. The Toms Place 26Nov2006 ground motion shows amplification factors significantly larger than the other cases.  ? The main peak in the spectral ratio plots correspond to the identified natural transverse period for all cases.   6.1.3.4 Fourier spectra analysis  Figures 6.24 to 6.26 show the Fourier spectrum plots for the analyzed records in this case. The following main trends are observed in the spectra: ? The free field records from Qualeys Camp 18Sep2004 and Mammoth Lakes 12Jun2007 do not have a predominant peak of amplitude. The highest amplitude peak in the column base record is associated with the fundamental transverse frequency in all cases. The column base record from Qualeys Camp 18Sep2004 shows an important peak at F=4.33Hz for the column base record. ? The frequency of the peak amplitude point is similar to the mean frequency only for the Qualeys Camp 18Sep2004. ? The records from the column base in all cases show higher amplitudes from the lower frequencies up to the region of the natural frequency. The opposite effect is shown for the rest of the spectrum.  66        Figure 6.24. Fourier spectra for Qualeys Camp 18Sep2004 (PGA=0.014g), Highway 395 Bridge        Figure 6.25. Fourier spectra for Toms Place 26Nov2006 (PGA=0.015g), Highway 395 Bridge        Figure 6.26. Fourier spectra for Mammoth Lakes 12Jun2007 (PGA=0.051g),. Highway 395 Bridge 67  6.1.3.5 Free field record vs. spectral analysis results comparison  A comparison of results shows the following characteristics: ? The Qualeys Camp 18Sep2004 and Mammoth Lakes 12Jun2007 cases show higher spectral ratio for lower period differences. ? The Toms Place 26Nov2006 case shows the highest spectral response ratio, even though the Mammoth Lakes 12Jun2007 case has a smaller period difference and higher peak Fourier amplitude. The Fourier spectrum for the column base record shows a peak of amplitude at F=4.33Hz. The system identification process results associate a frequency of 4.30Hz to the translational mode of the system in the direction perpendicular to the skewed bents. The orbit plots for this ground motion also show a predominant direction of shaking approximately perpendicular to the bents. From these observations it is possible to infer that the peak in the spectral ratio is most likely associated with a coupled modal response, where both the fundamental transverse mode and the translational mode define the response. As in the case of the Via California Bridge described in Section 6.1.2 previously.  ? Ground Motion Identified Fundamental Period [Tn (sec)] Free Field Record Spectral Acc.  Ratio Mean Period [ Tm (sec) ] Peak Amplitude Point Fourier Amplitude for Tn Location PGA (g) Period [ Tp (sec) ] Fourier Amplitude Qualeys Camp 18Sep2004 0.014 0.186 0.260 0.307 3.172 1.945 1.506 Toms Place 26Nov2006 0.015 0.192 0.134 0.092 3.332 0.806 2.765 Mammoth Lakes 12Jun2007 0.051 0.187 0.175 0.145 7.253 1.742 1.587 Table 6.12.    Free field record properties vs. spectral analysis results, Highway 395 Bridge.   68                 Figure 6.27. Free field record properties vs. spectral analysis results, Highway 395  6.1.4 Alviso Overpass (Bridges K and L) 6.1.4.1 Description of the ground motions  The only ground motion available in this station corresponds to low intensity shaking. Its characteristics are listed below.  69  No. Epicenter Location / Date Epicentral Distance (Km) PGA (g) PGV (cm/s) PGD (cm) 10-11 Gilroy 13May2002 60.0 0.053 0.660 0.074 Table 6.13.    Ground motion included in the analysis of the Alviso Overpass (Bridges K and L)             Figure 6.28. Displacement orbit plot of the ground motion included in the analysis. Alviso Overpass (Bridges K and L)  The displacement orbit plot of the Gilroy 13May2002 ground motion does not show a predominant shaking orientation. There is a slight trend of shaking in the direction of Channel 1, which is rotated about 45 degrees to the longitudinal axis of the bridge.    70  6.1.4.2 Modal identification results  The identified modal parameters correspond to the first mode in the transverse direction. The results are summarized as follows:  Bridge No. Ground Motion PGA (g) Frequency (Hz) Fundamental Period (sec) Damping Ratio (%) Alviso Overpass (K) 10 Gilroy 13May2002 0.019 1.636 0.611 4.22 Alviso Overpass (L) 11 Gilroy 13May2002 0.019 2.568 0.389 5.5 Table 6.14. Identified fundamental modal parameters in the transverse direction for the Alviso Overpass (Bridges K and L)  The identified natural transverse frequency and damping ratio are higher for Bridge L. This is expected considering the substructure characteristics both bridges.  6.1.4.3 Response spectra analysis  The resulting spectral ratios for the natural are summarized in table 6.15. No. Bridge Record PGA Tn SaBASE / SaFF SvBASE / SvFF SdBASE / SdFF (g) (sec) 10 Alviso Overpass (K) Gilroy 13052002 0.019 0.611 1.46 1.50 1.46 11 Alviso Overpass (L) Gilroy 13052002 0.019 0.389 3.16 3.21 3.16 Table 6.15.   Spectral ratios for the identified natural transverse period. Highway 395 Bridge  The following remarks summarize the observed characteristic of the column base / free field spectral response evaluation: 71  ? The spectral ratios for the natural period show amplification at the column base in all cases.  ? The records from Bridge K show peak response of the column base in a period significantly lower than the identified natural period, for both acceleration and velocity spectra. Both peaks are also reported in the displacement spectra, but the one corresponding to the natural period is slightly higher. The peak spectral response at the center of the deck corresponds to the identified frequency (see appendix). ? The amplification of the response at the column base was found for almost the entire spectrum in both bridges. The peak spectral ratio is not matching the fundamental frequency, as it happens with the spectral response.  The response spectra and spectral ratios plots for both bridges are shown in Figures 6.29 to 6.32.      Figure 6.29. Response spectra corresponding for Gilroy 13May2002 (PGA=0.019g), Alviso Overpass Bridge K      Figure 6.30. Response spectra for Gilroy 13May2002 (PGA=0.019g), Alviso Overpass Bridge L 72         Figure 6.31. Spectral ratios for the Gilroy 13May2002 earthquake (PGA=0.019g), Alviso Overpass (Bridge K)        Figure 6.32. Spectral ratios for the Gilroy 13May2002 earthquake (PGA=0.019g), Alviso Overpass (Bridge L)  6.1.4.4 Fourier spectra analysis  The Fourier spectra in these cases ?shown in Figures 6.33 and 6.34? present the following main characteristics: 73  ? The signals at the column base show peaks of Fourier amplitude at several frequencies. The peaks close to 3.75Hz, 4.25Hz and 5.50Hz were found in the signal at the column base of both bridges. ? The plots from Bridge K show the higher peak of Fourier amplitude close to 3.75Hz, which does not correspond to the natural identified frequency.  ? The signal recorded at the column base in Bridge L has its larger peak of Fourier amplitude at the identified natural transverse frequency.        Figure 6.33. Fourier spectra for Gilroy 13May2002 (PGA=0.019g), Alviso Overpass (Bridge K)        Figure 6.34. Fourier spectra for Gilroy 13May2002 (PGA=0.019g), Alviso Overpass (Bridge L) 74   The peaks of amplitude found in both spectra can be associated to the response of the similar foundation soil of the contiguous structures. This influence is bigger for the bridge with less transverse stiffness, where the fundamental response of the superstructure is not transferring more energy than the soil response at the foundation level. All these inferences are tied to the low intensity condition of the ground motion in this case, considering the elastic behavior that would be expected.  6.1.4.5 Free field record vs. spectral analysis results comparison  The following Table 6.16 and Figure 6.35 summarize the results for these bridges. The results can be compared since the structures are contiguous bridges that have similar plan geometry and height, with the number of columns in the bents as the main difference between them.  Bridge / Ground Motion Identified Fundamental Period [Tn (sec)] Free Field Record Spectral Acc.  Ratio Mean Period [ Tm (sec) ] Peak Amplitude Point Fourier Amplitude for Tn Bridge /  GM Location & Date PGA (g) Period [ Tp (sec) ] Fourier Amplitude Bridge K /  Gilroy 13May2002 0.014 0.611 0.340 0.258 4.230 1.880 2.351 Bridge L /  Gilroy 13May2002 0.014 0.389 0.340 0.258 4.230 2.980 2.928 Table 6.16.    Free field record properties vs. spectral analysis results. Alviso Overpass, Bridges K and L.  The results indicate that the fundamental period of Bridge L is very close to both period of peak amplitude and mean period of the free field record. This bridge shows the higher spectral ratio, which agrees with the predominant trend of the previous cases of study, where higher spectral 75  ratios are associated to low differences between the fundamental period of the structure and the period that represents the free field record.              Figure 6.35. Free field record properties vs. spectral analysis results. Alviso Overpass (Bridges K and L)  6.1.5 Murray Road Bridge 6.1.5.1 Description of the ground motions  The ground motion recorded at this station is described as follows.  No. Epicenter Location / Date Epicentral Distance (Km) PGA (g) PGV (cm/s) PGD (cm) 12 Ferndale 09Jan2010 64.7 0.077 13.120 3.149 Table 6.17.    Ground motion included in the analysis of the Murray Road Bridge 76   Figure 6.36 shows the displacement orbit plot for the ground motions and its relative orientation to the bridge. The orbit shows two main shaking directions, one of them is longitudinal to the bridge and the other one is rotated about 45? to the longitudinal axis.                  Figure 6.36. Displacement orbit plot of the ground motion included in the analysis, Murray Road Bridge   77  6.1.5.2 Modal identification results  The identified modal parameters listed below correspond to the first mode in the transverse direction.  No. Ground Motion PGA (g) Frequency (Hz) Period (sec) Damping Ratio (%) 12 Ferndale 09Jan2010 0.077 6.581 0.152 0.35 Table 6.18.    Identified modal parameters in the transverse direction for the Murray Road Bridge  6.1.5.3 Response spectra analysis  The response spectra and spectral ratios for the recorded ground motions are listed in Table 6.19 and plotted in Figures 6.37 to 6.38.  The following list summarizes the remarks in this case: ? The spectral ratios for the natural period show slight de-amplification at the column base.  ? The peak of the spectral response at the column base does not correspond to the natural identified period. ? Most of the periods greater than the first-mode value show amplification or similar value in the spectral response.  No. Record PGA Tn SaBASE / SaFF SvBASE / SvFF SdBASE / SdFF (g) (sec) 12 Ferndale 09Jan2010 0.077 0.152 0.94 0.82 0.95 Table 6.19.    Spectral ratios for the identified natural transverse period. Murray Road Bridge   78       Figure 6.37. Response spectra for Ferndale 09Jan2010 (PGA=0.077g), Murray Road Bridge          Figure 6.38. Spectral ratios for the analyzed earthquake, Murray Road Bridge  6.1.5.4 Fourier spectra analysis  The Fourier spectra ?included in Figure 6.36?  show the following main trends: ? There is a predominant peak of amplitude at F = 0.62Hz for both free field and column base records.  ? The peaks of amplitude in both signals are distributed in a similar way. 79  ? There is no peak of amplitude in the column-base signal at the fundamental frequency or its surroundings.        Figure 6.39. Fourier spectra for Ferndale 09Jan2010 (PGA=0.077g), Murray Road Bridge  6.1.5.5 Free field record vs. spectral analysis results comparison  The comparison between parameters and results is summarized in Table 6.20. The signal recorded at the column base does not show a predominant peak at the fundamental period. This seems to indicate that the movement at the column base is more influenced by the surrounding and foundation soil than by the seismic response of the superstructure, considering the similarities between both signals in frequency domain.  Ground Motion Identified Fundamental Period [Tn (sec)] Free Field Record Spectral Acc.  Ratio Mean Period [ Tm (sec) ] Peak Amplitude Point Fourier Amplitude for Tn Location PGA (g) Period [ Tp (sec) ] Fourier Amplitude Ferndale 09Jan2010 0.077 0.152 0.899 1.606 51.383 6.898 0.943 Table 6.20. Free field record properties vs. spectral analysis results, Murray Road Bridge.  80  6.1.6 Summary of results for comparative evaluation The fundamental periods and spectral ratios for all cases are included in the detailed analysis are listed in the following table:  No. Bridge Record PGA Tn SaBASE / SaFF SvBASE / SvFF SdBASE / SdFF (g) (sec) 1 Meloland Overpass Calexico 22May2010 0.031 0.259 0.87 0.85 0.86 2 Meloland Overpass Calexico 30Dec2009 0.174 0.288 0.66 0.61 0.66 3 Meloland Overpass Calexico 04Apr2010 0.213 0.273 0.71 0.66 0.71 4 Via California Borrego Springs 17Jul2010 0.006 0.372 2.35 2.13 2.36 5 Via California Calexico 04Apr2010 0.020 0.396 2.93 3.56 2.93 6 Via California Chinohillls 29Jul2008 0.023 0.381 1.92 1.75 1.92 7 Highway 395 QualeysCamp 18Sep2004 0.014 0.186 1.51 1.70 1.51 8 Highway 395 Toms Place 26Nov2006 0.015 0.192 2.76 2.82 2.76 9 Highway 395 Mammoth Lakes 12Jun2007 0.051 0.187 1.59 1.37 1.60 10 Alviso Overpass (K) Gilroy 13May2002 0.019 0.611 1.46 1.50 1.46 11 Alviso Overpass (L) Gilroy 13May2002 0.019 0.389 3.16 3.21 3.16 12 Murray Road Bridge Ferndale 09Jan2010 0.077 0.152 0.94 0.82 0.95 Table 6.21.   Spectral response ratios for acceleration (Sa), velocity (Sv) and displacement (Sd) corresponding to the identified fundamental transverse period (Tn) for all the studied cases  The main findings from all the cases of study are listed in Table 6.22.  Meloland Overpass System Identification The fundamental period tends to be lower for low intensity shaking, without a clear trend with respect to PGA. The damping ratio values show important variations. Higher damping ratios seem to be related to higher period variations. Response Spectra De-amplification of the spectral response at the fundamental period for all cases. The peak of the response spectra at the column base is not located at the fundamental period, which is related to the effect of the abutment-embankment-bridge interaction. Fourier Spectra The Fourier amplitudes at the column base are lower than the free field for most frequencies, starting from approximately the fundamental frequency to higher values. Free field vs. Spectral Analysis The spectral response ratios and the period differences have similar trend if compared with the PGA, for all the ground motions. The larger spectral response ratios seem to be related with smaller period differences and higher peak Fourier amplitudes.  81  Via California Bridge System Identification Low variability of Tn and no clear trend with respect to PGA. Higher variability for the damping ratio values.  Response Spectra The spectral response at the column base is amplified at the fundamental period and most part of the spectrum for all cases. The peak of the response spectra at the column base is located at the fundamental period. Fourier Spectra The records from the column base show higher amplitudes than the one from free field, in a range that goes from the lower frequencies to the region of the fundamental frequency, and opposite effect for the rest of the spectrum. Free field vs. Spectral Analysis Two records show the trend of higher spectral response ratios for lower period differences. The third earthquake shows the higher response ratio even though it has the larger period difference of all cases. This is associated to a coupled fundamental-translational modal response. The peak Fourier amplitude and the spectral ratios of all the records follow the same trend if compared with the PGA. Highway 395 Bridge System Identification Low variability and no clear trend with respect to PGA. High variation between damping ratio values.  Response Spectra Amplification of spectral response at the column base in all cases. One ground motion shows significantly larger amplification factors than the other cases, which is most likely as a result of the orientation of shaking. The main peak in the spectral ratio plots corresponds to the identified natural transverse period for all cases. Fourier Spectra For all cases the signal from the column base show higher Fourier amplitudes than the free field record from the lower frequencies to the surroundings of the fundamental frequency. The frequency of the highest peak is always lower than the mean frequency. The highest peak of amplitude in the column base record corresponds to the fundamental transverse frequency in all cases, and the Toms Place 26Nov2006 record shows a second important peak. Free field vs. Spectral Analysis The trend of higher spectral ratios for smaller period differences was noticed for two earthquakes. The third record shows the highest spectral response ratio of all cases, and does not correspond to the smaller period difference. This is associated to a coupled fundamental-translational modal response induced by the direction of shaking in this particular case. Alviso Overpass ? Bridges K and L System Identification Fundamental Periods of Tn=0.611sec for Bridge K and Tn=0.389 sec for Bridge L. Higher damping ratio identified for the stiffer bridge. Response Spectra Amplification of the spectral response at the column base in both cases, for almost the entire spectrum. For Bridge K, the record from the column base shows its peak spectral acceleration at a period significantly lower than the fundamental period. The record from the column base of Bridge L has its peak is at the fundamental period. Fourier Spectra The plots from Bridge K show the higher peak amplitude close to 3.75Hz, which does not correspond to the fundamental frequency. The signal recorded at the column base of Bridge L has its larger energy peak at the fundamental frequency. Three peaks of Fourier amplitude were found in the spectra from the column base of both bridges.  Free field vs. Spectral Analysis The peaks of energy found in both spectra can be associated to the response of the foundation soil of the contiguous structures. The motion of the column base of the bridge with less transverse stiffness seems to be very influenced by the foundation soil.   82  Murray Road Bridge System Identification Identified fundamental period Tn=6.581sec for the single available record in this station. Response Spectra The spectral ratio for the natural period show slight de-amplification at the column base. The peak of the spectral response at the column base does not correspond to Tn. Fourier Spectra Predominant peak of amplitude at F = 0.62Hz for both free field and column base records. There is no peak of amplitude in the column-base signal at the fundamental frequency or its surroundings. Free field vs. Spectral Analysis The evaluation suggests that the movement at the column base is more influenced by the surrounding and foundation soil than by the seismic response of the superstructure, considering the similarities between both signals in frequency domain. The sensitivity analysis of free field record parameters is not possible due to lack of records. Table 6.22. Summary of observations and findings from the evaluation of response spectra  6.2 Time-frequency analysis  This section is dedicated to the evaluation of the frequency content of the recorded signals during the ground motion. The results are separated into five sections. As in Section 6.1, there will be one for each analyzed bridge, where a single ground motion is presented to exemplify the evaluation process. The summary and findings for all the ground motions are included in the last section. The time-frequency spectra are a result of Wavelet analysis as detailed in Chapter 3 and correspond to signals at the free field, column base, and center of the deck. They each show the Time-Integral Squared Amplitude for the Power Spectral Density (PSD-TISA) in a colored scale on a 2D contour plot.   Each plot shows the identified fundamental frequency as a horizontal white line, and an icon with the location of the recording instrument. The time history plot of the signal is located at the bottom.  83  6.2.1 Meloland Overpass  The frequencies of the peaks of PSD in the spectra and the identified frequencies are compared in Table 6.23. Figures 6.40 ? 6.42 show the time-frequency spectra corresponding to the Calexico 22May2010 ground motion.  Epicenter  Location / Date PGA Fundamental Frequency Frequency of Peak PSD-TISA Freq. Ratios Period Ratios  (g) Fn (Hz) Fw (Hz) Fw/Fn Tw/Tn Calexico 22May2010 0.031 3.86 3.03 0.78 1.27 Calexico 30Dec2009 0.174 3.47 2.59 0.75 1.33 Calexico 04Apr2010 0.213 3.66 2.54 0.70 1.43 Table 6.23. Comparison between observed frequencies of peak PSD-TISA at the center of the deck and identified fundamental frequencies, Meloland Overpass           Figure 6.40. Time-frequency spectrum of the signal at the free field. Calexico 22May2010 earthquake (PGA=0.031g), Meloland Overpass 84            Figure 6.41. Time-frequency spectrum of the signal at the column base. Calexico 22May2010 earthquake (PGA=0.031g), Meloland Overpass           Figure 6.42. Time-frequency spectrum of the signal at the center of the deck. Calexico 22May2010 earthquake (PGA=0.031g), Meloland Overpass 85  The following remarks are related to the observed behavior: ? There is a de-amplification of the PSD at the column base if compared with the free field signal. ? The spectra that correspond to the column base signal show clear similarities to the other records, with several peaks of PSD that were noticed in both the free field and deck signals. This indicates the strong influence of the soil behavior in the dynamic response of the structural system, as expected for integral bridges.  ? The peaks of PSD-TISA in the wavelet spectra from the deck correspond to frequencies significantly lower than those identified for the structure. This reduction in the frequency ratios is larger for higher intensities of shaking.           Figure 6.43. Frequency Domain Decomposition chart and modal shapes associated to frequencies of interest for Calexico 22May2010 earthquake (PGA=0.031g), Meloland Overpass  86  Figure 6.43 includes the results from the Frequency Domain Decomposition process, including the modal shapes associated to the frequency of the peak PSD-TISA (Fp=3.03Hz) and the fundamental identified frequency (Fn=3.86Hz) for this case. The mode that corresponds to the frequency of peak PSD reveals a coupled behavior, where the fundamental transverse mode and the pure translation of the deck occur at the same time. This can be explained based on the condition of integral bridge, where the fixing situation to the abutment-embankment system plays a crucial role, allowing the embankments to define the overall behavior of the system. This condition was observed in all the recorded motions from this station, and was also addressed in the findings from the response spectra evaluation.  Previous research conducted by Wilson & Tan (1990)  identified the natural transverse frequency of the approach embankments for the Meloland Overpass based on the Imperial Valley 1979 ground motion. The resulting frequency is close to 2.5Hz for the initial part of the record, which helps to understand the reduction in the natural frequency of the system shown in this document. The doctoral dissertation presented by Carvajal (2011) contains further information regarding this type of bridges.  6.2.2 Via California bridge The example set of spectra shown in this section correspond to the Chino Hills 29Jul2008 ground motion (PGA = 0.023g).     87            Figure 6.44. Time-frequency spectrum of the signal at free field. Chino Hills 29Jul2008 earthquake (PGA=0.023g). Via California Bridge.           Figure 6.45. Time-frequency spectrum of the signal at column base. Chino Hills 29Jul2008 earthquake (PGA=0.023g). Via California Bridge 88            Figure 6.46. Time-frequency spectrum of the signal at the center of the deck. Chino Hills 29Jul2008 earthquake (PGA=0.023g). Via California Bridge           Figure 6.47. Time-frequency spectrum of the signal at the right end of the deck. Chino Hills 29Jul2008 earthquake (PGA=0.023g). Via California Bridge. 89  The following remarks summarize the noticed behavior in this case, and can be related to all the analyzed ground motions: ? There is a clear amplification of PSD at the column base when compared with signal recorded at the free field. ? The noticeable peaks in the signal at column base and center of the deck correspond to the identified natural frequency in all cases. ? The peaks in the free field motion are weakly reflected at the column base. The record at the center of deck does not show peaks corresponding to those in the free field signal. ? The records from the extreme points of the deck ?near the abutments? reflect part of the peaks shown in the input signal, with amplified PSD.  The similarities in the frequency content between the free field and the abutment records are related to the fixing condition of the bridge to the abutment (rocker bearings). These observations are further evidence of the amplification effect previously reported in the response spectra analysis, and might be related to the low intensity of the motions.  6.2.3 Highway 395 bridge The example set of spectra shown in Figures 6.48 to 6.51 correspond to the Mammoth Lakes 12Jun2007 ground motion (PGA = 0.051g). Table 6.25 summarize the observed frequencies.    90  Epicenter  Location / Date PGA Natural Frequency Frequency of Peak PSD-TISA Freq. Ratios Period Ratios  (g) Fn (Hz) Fw (Hz) Fw/Fn Tw/Tn Qualeys Camp 18Sep2004 0.014 5.38 5.50 1.02 0.98 TomsPlace 26Nov2006 0.015 5.20 5.18 1.00 1.00 Mammoth Lakes 12Jun2007 0.051 5.34 5.35 1.00 1.00 Table 6.24.    Comparison between observed frequencies of Peak PSD-TISA at the center of the deck and identified fundamental frequencies, Highway 395 Bridge            Figure 6.48. Time-frequency spectrum of the signal at free field. Mammoth Lakes 12Jun2007 earthquake (PGA=0.051g). Highway 395 Bridge       91            Figure 6.49. Time-frequency spectrum of the signal at column base. Mammoth Lakes 12Jun2007 earthquake (PGA=0.051g). Highway 395 Bridge          Figure 6.50. Time-frequency spectrum of the signal at the center of the deck. Mammoth Lakes 12Jun2007 earthquake (PGA=0.051g). Highway 395 Bridge. 92           Figure 6.51. Time-frequency spectrum of the signal at the right end of the deck. Mammoth Lakes 12Jun2007 earthquake (PGA=0.051g). Highway 395 Bridge.  The following characteristics describe the observed behavior, and can be related to all the analyzed ground motions: ? The peaks of PSD for the signal at column base and center of the deck correspond to the identified natural frequency in all cases. ? The PSD at the column base is amplified with respect to the free field signal, for frequencies up to approximately the fundamental identified frequency, and de-amplified for the higher frequencies. ? The record at the center of deck does not reflect the peaks of PSD observed in the free field signal. ? The records from the extreme points of the deck ?near the abutments? show similar peaks of PSD w.r.t. the free field signal. 93  As it happened with the previous bridge, the rocker bearings that connect the deck to the abutments cause the similarities in the frequency content between the free field and the abutment records. The amplification noticed here was also reported in the response spectra analysis, and may be related to the low intensity of the motions.  6.2.4 Alviso Overpass (Bridges K and L) The spectra included in this section correspond to the Gilroy 13May2002 ground motion (PGA=0.019g) for both Bridge K and Bridge L. The set of spectra from Bridge K is presented first as follows. These bridges are located parallel and close to each other, and the signal recorded at the free field location is the same for both structures.           Figure 6.52. Time-frequency spectrum of the signal at free field. Gilroy 13May2002 earthquake (PGA=0.019g). Alviso Overpass (Bridges K and L).  94           Figure 6.53. Time-frequency spectrum of the signal at column base. Gilroy 13May2002 earthquake (PGA=0.019g). Alviso Overpass (Bridge K).           Figure 6.54. Time-frequency spectrum of the signal at the center of the deck. Gilroy 13May2002 earthquake (PGA=0.019g). Alviso Overpass (Bridge K). 95            Figure 6.55. Time-frequency spectrum of the signal at the left end of the deck. Gilroy 13May2002 earthquake (PGA=0.019g). Alviso Overpass (Bridge K).  Epicenter  Location / Date PGA Natural Frequency Frequency of Peak PSD-TISA Freq. Ratios Period Ratios  (g) Fn (Hz) Fw (Hz) Fw/Fn Tw/Tn Gilroy 13May2002 0.019 1.64 1.70 1.04 0.96 Table 6.25.    Comparison between observed frequencies of Peak PSD-TISA at the center of the deck and identified fundamental frequencies. Alviso Overpass (Bridge K)  The evaluation of the previous spectra for Bridge K leads to the next remarks: ? The signal at the center of the deck shows its larger peak of PSD matching the identified fundamental frequency. The peaks observed in the free field signal are not reflected. 96  ? The records from the column base and the extreme points of the deck ?near the abutments? reflect the peaks of PSD noticed in the free field signal with amplification at F = 3.75Hz. The fundamental frequency does not show any PSD peak at these locations.  Table 6.27 includes the results for Bridge L. The spectra for this case are shown in Figures 6.56 to 6.58.   Epicenter  Location / Date PGA Natural Frequency Frequency of Peak PSD-TISA Freq. Ratios Period Ratios  (g) Fn (Hz) Fw (Hz) Fw/Fn Tw/Tn Gilroy 13May2002 0.019 2.57 2.62 1.02 0.98 Table 6.26.    Comparison between observed frequencies of Peak PSD-TISA at the center of the deck and identified natural frequencies. Alviso Overpass (Bridge L)           Figure 6.56. Time-frequency spectrum of the signal at column base. Gilroy 13May2002 earthquake (PGA=0.019g). Alviso Overpass (Bridge L). 97           Figure 6.57. Time-frequency spectrum of the signal at the center of the deck. Gilroy 13May2002 earthquake (PGA=0.019g). Alviso Overpass (Bridge L).           Figure 6.58. Time-frequency spectrum of the signal at the left end of the deck. Gilroy 13May2002 earthquake (PGA=0.019g). Alviso Overpass (Bridge L). 98  The Bridge L case shows the following characteristics: ? There is amplification of the peaks of PSD corresponding to the fundamental frequency at the column base if compared with the free field record. ? The peak of PSD observed in the signal at the center matches the identified fundamental frequency. The frequency content of the input signal is almost not reflected. ? The records from the column base and the extreme points of the deck reflect the frequency content of the free field signal. There is an amplification of the PSD at F=3.75Hz for the column base record. These records also show a peak of energy matching the identified natural frequency, more accentuated at the column base.  The evaluation of these cases ?where both bridges are affected by similar input motions? highlights the importance of the structural properties in the soil-structure effects, as it was noticed before in the response spectra analysis. Since the Bridge L has a stiffer structure in the transverse direction, the inertial soil-structure interaction effects in that case are more important than in the other bridge, therefore the modification of the motion at the column base is more noticeable. This explains why the signal recorded at the column base in Bridge K is similar to the signal from free field, whereas in Bridge L is similar to the signal from the deck.  These inferences are tied to the low level of shaking, since the non-linear behavior of the soil and structural system is less possible.  99  6.2.5 Murray Road Bridge The records from this bridge correspond to the Ferndale 09Jan2010 ground motion (PGA = 0.077g). The results are presented in Table 6.28 and Figures 6.59 to 6.62.  Epicenter  Location / Date PGA Natural Frequency Frequency of Peak PSD-TISA Freq. Ratios Period Ratios  (g) Fn (Hz) Fw (Hz) Fw/Fn Tw/Tn Ferndale 09Jan2010 0.077 6.59 6.80 1.03 0.97 Table 6.27.    Comparison between observed frequencies of Peak PSD-TISA at the center of the deck and identified natural frequencies. Murray Road Bridge.           Figure 6.59. Time-frequency spectrum of the signal at free field. Ferndale 09Jan2010 earthquake (PGA=0.077g). Murray Road Bridge    100            Figure 6.60. Time-frequency spectrum of the signal at column base. Ferndale 09Jan2010 earthquake (PGA=0.077g). Murray Road Bridge           Figure 6.61. Time-frequency spectrum of the signal at the center of the deck. Ferndale 09Jan2010 earthquake (PGA=0.077g). Murray Road Bridge 101            Figure 6.62. Time-frequency spectrum of the signal at the left extreme of the deck. Ferndale 09Jan2010 earthquake (PGA=0.077g). Murray Road Bridge  The next observations are related to this case: ? The peaks of PSD noticed in the records from the column base and the extreme points of the deck ?near the abutments? are located on similar time-frequency coordinates. The peak PSD noticed in the record from free field corresponding to F = 3.75Hz is also observed with less PSD at the column base. The fundamental frequency does not show any energy peak at the column base. ? The signal at the center of the deck shows a main peak of PSD matching the identified fundamental frequency. The frequency content of the input signal is barely reflected.  102  As it happened previously for Alviso Overpass (Bridge K), this case shows a negligible inertial effect of the superstructure on the motion at the column base. In the same way, this condition can be related to a structural system that is not stiff enough to produce an important influence on the column base motion. These inferences are also tied to a low intensity of shaking, since the non-linear behavior of the soil and structural system is less possible.  6.2.6 Summary of results for time-frequency analysis On the whole there is good agreement between the frequency values identified natural frequency and the peak of PSD in the signal at the center of the deck, with the Meloland Overpass case as the single exception. That bridge shows peaks of PSD corresponding to a frequency value lower than the identified fundamental frequency of the bridge. The fact that this reduction in frequency is more important for higher intensities is further evidence of the effect of the soil in the system, as it was noticed in the analysis previously described in this document. Tables 6.29 and 6.30 summarize the results and findings for all cases.  Bridge Epicenter Location / Date PGA Natural Frequency Peak PSD-TISA Frequency Freq. Ratios Period Ratios   (g) Fn (Hz) Fw (Hz) Fw/Fn Tw/Tn Meloland Overpass Calexico 22May2010 0.031 3.86 3.03 0.78 1.27 Meloland Overpass Calexico 30Dec2009 0.174 3.47 2.59 0.75 1.33 Meloland Overpass Calexico 04Apr2010 0.213 3.66 2.54 0.70 1.43 Via California Borrego Springs 07Jul2010 0.006 2.69 2.60 0.97 1.03 Via California Calexico 04Apr2010 0.020 2.52 2.55 1.01 0.99 Via California Chino Hills 29Jul2008 0.023 2.62 2.65 1.01 0.99 Hwy 395 (Lake Crowley) Qualeys Camp 18Sep2004 0.014 5.38 5.50 1.02 0.98 Hwy 395 (Lake Crowley) TomsPlace 26Nov2006 0.015 5.20 5.18 1.00 1.00 Hwy 395 (Lake Crowley) Mammoth Lakes 12Jun2007 0.051 5.34 5.35 1.00 1.00 Alviso Overpass (Bridge K) Gilroy 13May2002 0.019 1.64 1.70 1.04 0.96 Alviso Overpass (Bridge L) Gilroy 13May2002 0.019 2.57 2.62 1.02 0.98 Murray Road Bridge Ferndale 09Jan2010 0.077 6.59 6.80 1.03 0.97 Table 6.28.    Summary of results for time-frequency analysis. 103  Bridge Main observations Meloland Overpass ? De-amplification of the PSD at the column base with respect to the free field record.  ? The peak PSD is located at a smaller frequency than the fundamental, which is related to a coupled mode shape where the embankments induce translation of the system.  ? The fact that this frequency reduction (or period enlargement) is more important for higher intensities evidences the importance of the SSI effects. Via California ? Amplification of the PSD of the signals at the column base if compared with the free field record.  ? The signals recorded near the abutments reflect the frequency content of the signals recorded at free field. Hwy 395 (Lake Crowley) ? Amplification of the PSD of the signals at the column base if compared with the free field record, for the fundamental frequency.  ? The signals recorded near the abutments and at free field show similar frequency content. Alviso Overpass (Bridges K and L) ? Since the Bridge L has a stiffer structure in the transverse direction, the inertial soil-structure interaction effects in that case are more important than in the other bridge, therefore the influence of the superstructure at the column base is more noticeable.  ? This explains why the frequency content of the column base in Bridge K is similar to the signal from free field, whereas in Bridge L is similar to the signal from the deck.   Murray Road Bridge ? The effect of the superstructure on the motion at the column base is negligible if compared with the free field record.  ? This condition can be related to a structural system that is not stiff enough to produce an important influence on the column base motion.  Table 6.29.   Summary of findings from the time-frequency analysis. 104  Chapter  7: Conclusions and recommendations  7.1 Summary  This study was focused on highlighting evidences of soil-structure interaction effects on the seismic response of instrumented bridges. The evaluation methodology was applied to a group of 6 bridges located in California. The set of records correspond to 12 ground motions. The data was provided by the California Earthquake Strong Motion Database.  The dynamic characteristics of the structural system were assessed first and summarized in Chapter 5, and the properties of the recorded earthquakes were detailed at the beginning of Chapter 6. Then the evaluation process was applied for all the cases, and the results were detailed for each case in Chapter 6.   The assessment was done in two stages. The initial phase is a comparative evaluation of the seismic response of the structures. This included the identification of the modal properties followed by the comparison of spectral response estimated based on free field and column base records. The next stage of the analysis was intended to evaluate the changes in properties of the input and the structural response through time-frequency spectra, obtained by using wavelet transform techniques. The findings for all the 6 bridges and 12 earthquakes are related to the remarks of the previous step and the overall properties of the structural systems.   105  7.2 Conclusions  This method of analysis highlights several aspects that are related to the soil-structure interaction condition of the bridges. The main findings are summarized as follows: ? The spectral response analyses show a general trend of having amplification of the spectral response for low levels of shaking. The de-amplification effect can be found in both low and high intensity motions, but is more important for higher intensities. This is noticed in the Meloland Overpass case, an integral bridge where the dynamic response is highly affected by the embankments.  ? The natural period of the structure was deemed as an important factor used to evaluate the inertial aspect of the soil-structure interaction effect. This was clearly observed in the evaluation of the Alviso Overpass, where two bridges of similar length with similar foundation soil are subjected to the same input motion. In this case, the response at the column base of the stiffer bridge ?with lower period? was clearly affected by the dynamic behavior of the superstructure, based on the similarities in the time-frequency spectra of the signals. On the other hand, the superstructure of the bridge with more flexible bents ?and higher period? had almost no effect on the response at the column base. ? The amplification or deamplification of the response seems to be affected by the difference between the fundamental period of the structure and the mean period of the free field record. When those periods are close, the spectral ratio (column base versus free field) tends to be higher. This was also noticed when the period of the peak Fourier amplitude of the free field record was used for comparison. 106  ? The time-frequency analysis results show that the peak response at the deck corresponds to the natural transverse period for almost all cases. This does not occur in the Meloland Overpass case, where the dynamic behavior is highly affected by the approaching embankments. ? The soil-structure interaction effect on the seismic response of the structures has been clearly noticed in the studied cases, either by amplification or de-amplification of the response. Neglecting this condition could lead to design considerations that evidently differ from the actual behavior of the bridge.  7.3 Recommendations  This method can be applied to several bridges in order to produce enough data to allow a statistical parametric study regarding soil-structure interaction effects. The results from this investigation should be used to develop guidelines to evaluate the effects of soil-structure interaction on the seismic response of bridges.  The Enhanced Frequency Domain Decomposition method was applied in this research work as the modal identification technique. Another system identification technique may be used either to determine the dynamic properties or to confirm the values when needed. The Stochastic System Identification technique is a reliable option.   An important contribution to this method would be the inclusion of results from ambient vibration tests. These findings can be used to estimate the natural period of the bridge in a 107  condition in which the effect of the ground shaking on the response is negligible, which can be used as a reference to define how the soil-structure interaction affects the response under seismic excitation.   As an example, the Alviso Overpass bridges are a good option to study SSI effects in more detail, taking advantage of having two structures with the same input motion. Ambient vibration test could be done to estimate the dynamic properties of the structures, and compare them with the dynamic properties identified from the low intensity earthquake included in this report. This can also be used if other earthquakes are recorded in this station, especially if the intensity is higher, allowing to evaluate the effect of this parameter in the response.              108   Bibliography American Petroleum Institute. Recommended practice for planning, designing and constructing fixed offshore platforms - working stress design. Washington: API Publishing Services, 2007. ASCE. Seismic Rehabilitation of Existing Buildings. Standard ASCE/SEI 41-06. . Reston, Virginia: American Society of Civil Engineers, 2007. Bielak, Jacobo. "Dynamic behaviour of structures with embedded foundations." Earthquake engineering and structural dynamics 3, no. 3 (1974): 259-274. Boore, David M. "Effect of baseline corrections on displacements and response spectra for several recordings of the 1999 Chi-Chi, Taiwan, earthquake." Bulletin of the Seismological Society of America, no. 91 (2001): 1199-1211. Brincker, Rune, Carlos E. Ventura, and Palle Andersen. "Damping estimation by frequency domain decomposition." 19th International Modal Analysis Conference. 2001. 698-703. Brincker, Rune, Lingmi Zhang, and Palle Andersen. "Modal identification from ambient responses using frequency domain decomposition." Proceedings of the 18th international modal analysis conference. San Antonio, 2000. 625-630. Carvajal, Juan C. Seismic embankment-abutment-structure interaction of integral abutment bridges. Vancouver: University of British Columbia, 2011. Chopra, Anil K. Dynamics of Structures. 3rd. New Jersey: Prentice Hall, 2006. COSMOS. "Center for Engineering Strong Motion Data." Cosmos Format. August 2001. http://www.strongmotioncenter.org/NCESMD/reports/cosmos_format_1_20.pdf (accessed May 2013). 109  Finn, Liam. "Aspects of Soil-Structure interaction." Soil Structure Interaction Seminar. Vancouver: University of British Columbia, 2010. Gazioglu, Sal, and Michael O?Neill. "Evaluation of p-y relationships in cohesive soils." Analysis and Design of Pile Foundations. San Francisco: ASCE, 1984. 192-213. Kausel, Eduardo. "Early history of soil-structure interaction." Soil Dynamics and Earthquake Engineering 30 (2009): 822-832. Matlock, Hudson. "Correlations for design of laterally loaded piles in soft clay." 2nd Annual Offshore Technology Conference. Houston, 1970. Murchison , Jack, and Michael O?Neill. "Evaluation of p-y relationships in cohesionless soils." Analysis and Design of Pile Foundations. San Francisco: ASCE, 1984. 174-191. O'Neill, M. W., and J. M. Murchinson. An Evaluation of p-y Relationships in Sands. PRAC 82-41-1, Houston: University of Houston, 1983. Pandey, Bishnu, Carlos Ventura, and Liam Finn. "Modification of free-field motions by soil-foundation-structure interaction for shallow foundations." 15 World Conference on Earthquake Engineering. Lisbon, 2012. Rahmani, Amin, Mahdi Taiebat, W.D. Liam Finn, and Carlos E. Ventura. "Evaluation of p?y Curves Used in Practice for Seismic Analysis of Soil?Pile Interaction." GeoCongress 2012: State of the Art and Practice in Geotechnical Engineering. Lisbon: ASCE, 2012. 1780-1788. Rathje, Ellen, Fadi Faraj, Stephanie Russell, and Jonathan Bray. "Empirical Relationships for Frequency Content Parameters of Earthquake Ground Motions." Earthquake Spectra (Earthquake Engineering Research Institute) 20, no. 1 (2004): 119-144. 110  Reese, Lymon C., William R. Cox, and Francis D. Koop. "Field testing and analysis of laterally loaded piles in stiff clay." Offshore Technology Conference. Houston, 1975. 671?690. Stewart, Jonathan, Raymond Seed, and Gregory Fenves. Empirical Evaluation of Inertial Soil Structure Interaction Effects. Berkeley: University of California, 1998. Structural Vibrations Solutions. Frequency Domain Operational Modal Analysis. 2013. http://www.svibs.com/products/frequency_domain_modal_analysis.aspx#FDD (accessed August 2013). Torrence , Christopher, and Gilbert Compo. "A Practical Guide to Wavelet Analysis." Bulletin of the American Meteorological Society 79, no. 1 (1998): 61-78. Veletsos, Anestis, and Damodaran Nair. "Seismic interaction of structures on hysteretic foundations." Journal of the Structural Division 101, no. 1 (January 1975): 109-129. Wilson, John C., and Boon S. Tan. "Bridge abutments assessing their influence on earthquake response of Meloland Road Overpass." Edited by ASCE. Journal of Engineering Mechanics 116(8) (1990): 1838-1856.          111  Appendices  Appendix A  BRIDGE-SSI System manual  This document is intended to guide the user of the BRIDGE-SSI System in the following aspects: ? Database folders creation and uploading ? Using the BRIDGE-SSI stand-alone software  The following sections include detailed explanation of all the topics.  A.1 BRIDGE-SSI Database management  The online database stores files in ?.zip? format. Each file contains a folder that can be opened by the software after being extracted to a directory in the hard drive. The following sections describe the creation process for those files.  A.1.1. Folders creation  Each folder corresponds to a ground motion recorded in a particular bridge. It contains information from the structure and the ground motion, as well as the recorded signals from the instruments on the bridge. The items included in the folders are the bridge information file, the image files ?i.e. pictures, sensor layouts, drawings and soil profiles? and the recorded 112  acceleration values from the instruments on the bridge (in COSMOS ?.V2? format). Each file has to be prepared based on particular criteria before being included in the folder, which are explained in the following subsections.  ? Bridge Information File  The bridge information file contains the description of the bridge, its instrumentation and the overall characteristics of the ground motion. It must be created as a text file, organizing the data as it is shown in Table A.1.  Line # Description Example content 001 Bridge name I5/Via California 002 Location (City) Capistrano Beach 003 Location (State) California 004 Location (Country) United States of America 005 Station name/number CSMIP 13795 006 Station Manager California Geological Service 007-013 Superstructure?s description 6-cell concrete box girder. Cantilever abutments. 014-021 Substructure?s description Reinforced concrete columns. 2 columns per bent. 022-030 Foundation system?s description Spread footings. 031 No. Spans (span?s support configuration) 6 (Simply supported) 032 Travel surface configuration Deck 033 Plan shape Straight 034 Bents orientation (perpendicular/skewed) Skewed (variable) 035 Structural system ? beam/girder type Deck beam 036 Structural system ? truss type Pratt 037 Structural system ? type of arch Two-hinged 038 Structural system ? type of suspension Cable-stayed 039 Earthquake ? Name (Location - Date) CerroPrietoEvent1 11Feb2008 040 Earthquake ? Year 2008 041 Earthquake ? Magnitude 5.4 042 Earthquake ? Epicentral distance (Km) 45.8 043 Google maps link for the bridge?s location https://maps.google.ca/?ll=33.4659,-117.667&spn=0.153103,0.338173&t=m&z=16 044-300 Channels number-description 01. Deck ? Transversal  Table A.1.    Bridge Information File content 113  An Excel file was prepared to facilitate the preparation of the information file. It can be downloaded from the BRIDGE-SSI Website (http://bridgessi.civil.ubc.ca/Aboutus.html). The file should be used as explained below: ? Fill the table with the earthquake overall characteristics in the ?1.Earthquake Info? tab (see figure A.1) ? Check the corresponding boxes of the classification types in the ?2.Bridge Classification? tab (see figure A.2) ? Write the bridge characteristics in the yellow-filled cells in the ?3.Bridge-Earthquake Form? tab (see figure A.3) ? Review the information in the ?4.Final TXT File? tab, then select the blue cells and copy the information (see figure A.4) ? Create the bridge information file and paste the copied cells inside. The file must have the same name than the bridge, saved as text file. Example: ?Meloland overpass.txt?.      Figure A.1. ?1. Earthquake Info? tab in the Bridge Information Creator file     114           Figure A.2. ?2. Bridge Classification? tab in the Bridge Information Creator file            Figure A.3. ?3. Bridge-Earthquake Form? tab in the Bridge Information Creator file  115         Figure A.4. ?4.Final TXT File? tab in the Bridge Information Creator file  ? Image Files  The bridge pictures, sensor layouts, structural/foundation drawings, and soil profiles are shown to the user as images in the software interface.  Type Extension Name coding Name examples Recommended width  (pixels) Bridge map png [Station#] 1-Map.png 01336 1-Map.png 600 Bridge pictures JPG [Station#] 2-Images-[Counter].png 01336 2-Images-01.JPG 89732 2-Images-04.JPG 640 Drawings gif [Station#]  3-Drawings-[Counter].gif 01336 3-Drawings-01.gif 01336 3-Drawings-03.gif 1280* Soil profile jpg [Station#]  4-Soil-[Counter].jpg 01336 4-Soil-01.jpg 13795 4-Soil-04.jpg 1280* Sensors layouts GIF [Station#]  5-Sensor-[Counter].GIF 01336 5-Sensor-01.GIF 1024* * The best option would be the smaller size that allows proper reading of the text inside the image  Table A.2    Bridge Information File content  Table A.1.2. shows the naming-formatting requirements for each type of image. The uppercase/lowercase format of the file extension must be followed as it is presented in the list. 116  The recommended width values allow better visualization in the software interface. Figures A.5 to A.8 show examples of image files.           Figure A.5. Example bridge map          Figure A.6. Example bridge picture 117           Figure A.7. Title block of an example bridge drawings            Figure A.8. Example bridge drawings 118  ? Channel files  The record files contain the measured acceleration values. They have to be prepared following the standards from the COSMOS Strong Motion Data Format (COSMOS 2001) and saved with the ?.V2? extension. Figure A.1.9 shows an example of a record file.                 Figure A.9. Sample header of a record file in COSMOS format  119  Each channel has to be saved in a different file. Thus the file has to be named based on the channel recorded in that file, following the numbering criterion shown in the sensor layout. The example file shown in the figure was named ?CHAN017.V2?.  The header has to be prepared (or modified) ensuring that the information corresponds to the channels shown in the sensors layout for that recording instrument. In the example header shown in Figure A.9 the following information has to be checked/edited: ? Channel number at the beginning of Line 8 should correspond to the sensors layout coding. In the example the channel information shows ?Channel 1: Up? followed by ?(Sta Chn: 17)?, meaning that the actual channel number is 17 for this case. Thus the first part has to be edited to show: ?Channel 17:? (the ?Up? description can be removed). Note that the string had the word ?Channel? followed by 2 spaces and a digit (?Channel__1:?), and now one of the spaces is removed to fit the two digits number (?Channel_17:?).  ? Number of points and sampling interval (in seconds) must be included following the format shown in line 47.  ? Folder preparation and testing  The folders must include all the files previously listed, prepared as it was detailed (see Figure A.10). It is important to notice that the format has to be kept in that particular way, otherwise the software might not be able to read part of the information.   120  The folder name should follow this code: [Station #]-[Counter](Year). For example, a folder for the Meloland Overpass bridge (Station No. 01336), which corresponds to the 5th record (sorted by date) from an earthquake recorded in 2008 would have the name ?01336-05(2008)?.              Figure A.10. BRIDGE-SSI Database folder example  The prepared folder has to be tested using the BRIDGE-SSI Software. The following steps must be completed without problems to consider the folder ready-to-use: ? Open the folder ? Check the bridge information ? Check the ability to show the images (map, pictures, sensor layout, drawings, soil profile) 121  ? Check the list of channels to be sure that all the recorded channels are included ? Load three records and check their time history plots  The final step is to create a compressed ?.zip? file that contains the tested folder, which should have the same name than the folder (e.g. ?01336-05(2008).zip?). This is the file that will be uploaded to the database.  ? Files uploading  The process to upload the prepared compressed folders to the database includes three main steps: ? Upload the file to the server ? Enter the characteristics of the uploaded record in the table ? Associate the uploaded file with its corresponding line in the database table  The detailed steps were included in the software manual, available at the UBC Earthquake Engineering Research Facility (contact: http://www.civil.ubc.ca/about/facilities/eerf.php).        122  A.2 BRIDGE-SSI Software user manual  A.2.1. Installation and setup  The BRIDGE-SSI software was developed as a stand-alone application that can be used under any OS that allows running the Java Virtual Machine (JMV), which can be downloaded from the Java website (http://java.com/en/download/index.jsp)  The software is available at the BRIDGE-SSI System Website (http://bridgessi.civil.ubc.ca/). The link allows downloading a ?.zip? compressed file that contains the folder with the software, which must be extracted. The folder contains the setup file that must be executed to complete the installation process.           Figure A.11. Initial screen and File menu of the BRIDGE-SSI Software 123  A.2.2. Opening a folder  After downloading the compressed file from the BRIDGE-SSI website, it is recommended to save the extracted folder close to the root directory (e.g. C:\Bridge Records\Via California\). The information is loaded in the software by using the ?File Opener? command under the ?File? menu, shown in Figure A.2.1.          Figure A.12. ?File Opener? window of the BRIDGE-SSI Software  The folder to be loaded has to be selected by a single click on its name before clicking on the ?Open? button (see figure A.12 above). A double click on any folder will show the content instead of loading it.   124  A.2.3. Bridge information  The ?Bridges? section ?shown in Figures A.13 to A.15? is displayed after opening a folder.          Figure A.13. Bridge basic information tab of the BRIDGE-SSI Software          Figure A.14. Bridge?s drawings tab of the BRIDGE-SSI Software 125                        Figure A.15. Some of the examples of bridge classification types used in the BRIDGE-SSI Software 126  The bridge basic information interface ?Figure A.2.3? includes the location map and allows selecting one of the available pictures to be shown on screen. The bridge?s structure and foundation system are described next, followed by the classification of the bridge. The classification system used in this software ?shown in Figure A.2.5? is based on the information available at the Pghbridges Project website (http://pghbridges.com/basics.htm).  The next tabs in this section display the drawings ?Figure A.2.4? and soil profile images, allowing selecting and showing on screen one drawing at a time.  A.2.4. Loading acceleration record files  The recorded signals must be loaded and processed one by one. The channels are listed in the Time History section, shown in Figure A.16.          Figure A.16. List of channels in the Time-History section of the BRIDGE-SSI Software 127  The records are loaded following the steps listed below: a) Choose the channel to be loaded. You can see the channels location and recording directions in the sensors layout shown in the ?Records? section. b) Go to the ?Time-History? section and select the channel from the list, as shown in Figure A.2.6. c) Define the input parameters. The content of the record file is shown in the Input File window (see Figure A.17). The time step (in seconds), the first line and the last line of acceleration values are automatically read from the file. The predefined scaling factor is 1.0. Any of those parameters can be modified by the user. Click on the ?OK? button to load the channel. d) Define the parameters for signal processing. Baseline correction and filtering can be applied if needed. The window that shows the parameters is accessed by clicking on the ?Signal processing options? button. The changes in the acceleration, velocity and displacement time-history plots can be seen by clicking on their corresponding tab while the signal processing window is open. e) Generate the response spectra. This is done by clicking on the ?Refresh? button in the ?Response Spectra? section (Figure A.18) f) Generate the plots in the Time-Domain section. Enter to the section and define the limits of the significant duration interval in the ?Arias Intensity? tab ?Figure A.19? for ?Channel A?. Click on the ?Refresh? button. This has to be done in the ?Channel B? tab after loading the second record and in the ?Channel C? tab after loading the third one. g) Repeat the process for the second and third channel to be loaded in the software.   128           Figure A.17. ?Input File Parameters? window in the Time-History section of the BRIDGE-SSI Software           Figure A.18. ?Refresh? button in the ?Response Spectra? section of the BRIDGE-SSI Software   129           Figure A.19. ?Refresh? button in the ?Arias Intensity? section of the BRIDGE-SSI Software  A.2.5. Time-history plots and data  The ?Time-History? section contains the acceleration, velocity and displacement data after processing, as shown in Figure A.20.   A.2.6. Response spectra plots and data  The response spectra section includes the data corresponding to acceleration, velocity, displacement, pseudo-acceleration and pseudo-velocity responses.   The acceleration tab is shown in Figure A.21. The left side includes the calculation parameters of the spectra, and the maximum period and abscissas axis variable ?i.e. Period or Frequency? to be 130  shown in the plots. The ?Spectra? refresh-results button must be pressed after any change in the parameters.           Figure A.20. ?Time-History? section of the BRIDGE-SSI Software          Figure A.21. ?Response Spectra? section of the BRIDGE-SSI Software  131  A.2.7. Fourier and Power Spectra  Figure A.22 shows the Fourier Amplitude tab in the Frequency Domain section of the software. Channels A, B and C correspond to the first, second and third recorded signals ?respectively? that were selected in the time-history section. The Power Spectra tab has the same structure.          Figure A.22. ?Frequency Domain? section of the BRIDGE-SSI Software  A.2.8. SSI effects section  This part is intended to use the information from the signal processing results to highlight the effects of soil-structure interaction condition on the seismic response of the bridge. This first version of the software includes a sub-section dedicated to spectral response ratios shown in Figure A.23.   132           Figure A.23. ?SSI Effect? section of the BRIDGE-SSI Software  Each curve in the acceleration tab shows the ratio of spectral acceleration between two channels. If channels A, B and C correspond to the first, second and third recorded signals selected in the time-history section, the first curve corresponds to the ratio Channel A / Channel B, the second curve to Channel A / Channel C, and the third curve to Channel B / Channel C.   A value of 1.0 in the spectral ratio for any period T means that the spectral response calculated from those records is the same at that particular period. This threshold is shown in the graph as a yellow line to help interpreting the plots.   Other factors that might be useful to understand the SSI effects can be added in further versions as new tabs, based on the research results and plans of the research group.  133  A.2.9. Ground motion parameters (Arias Intensity, Bracketed Duration and Directionality)   These parameters were included in the Time-Domain section of the software, in separate tabs. Channels A, B and C correspond to the first, second and third recorded signals selected in the time-history section for all the tabs.  The first tab ?shown in Figure A.24? corresponds to the ?Arias Intensity? parameter. The time history, Arias Intensity and Accumulated Percentage of Total Arias Intensity are shown in both data table and graphs formats. The resulting significant duration is included above the data tables. The interval can be modified and the screen refreshed to show the new results.           Figure A.24. ?Arias Intensity? tab in the ?Time-Domain? section of the BRIDGE-SSI Software  134  The second tab ?see Figure A.25? corresponds to the Bracketed Duration parameter. The plots and data used to calculate this parameter are shown on screen, with the resulting duration. The threshold can be modified and the screen refreshed to show the new results.          Figure A.25. ?Bracketed Duration? tab in the ?Time-Domain? section of the BRIDGE-SSI Software          Figure A.26. ?Directionallity? tab in the ?Time-Domain? section of the BRIDGE-SSI Software 135  The third tab corresponds to the directionality plots, as shown in Figure A.26. The ?Refresh? button has to be pressed to show in the plots. A decimation factor can be set to reduce the points if needed.  A.2.10. Data and plots managing  The plots can be formatted, copied and/or saved as image files. A right click on the graphic will show the graphic properties, allowing changing it.  The data can be selected in the tables and copied either to a spreadsheet or as text-format data.               136  Appendix B  Response spectra plots The following Figures contain the response spectra for all the analyzed cases.                     Figure B.1. Response Spectra for Borrego Springs 17Jul2010 (PGA=0.012g), Meloland Overpass 137                       Figure B.2. Response Spectra for Calexico 20Nov2008 (PGA=0.017g), Meloland Overpass  138                      Figure B.3. Response Spectra for Calexico 22May2010 (PGA=0.031g), Meloland Overpass   139                      Figure B.4. Response Spectra for Calexico 30Dec2009 (PGA=0.174g), Meloland Overpass   140                      Figure B.5. Response Spectra for Calexico 04Apr2010 (PGA=0.213g), Meloland Overpass   141                      Figure B.6. Response Spectra for Borrego Springs 17Jul2010 (PGA=0.006g), Via California Bridge   142                      Figure B.7.  Response Spectra for Calexico 04Apr2010 (PGA=0.020g), Via California Bridge   143                      Figure B.8. Response Spectra for Chino Hills 29Jul2008 (PGA=0.023g), Via California Bridge   144                       Figure B.9. Response Spectra for Qualeys Camp 18Sep2004 (PGA=0.014g), Highway 395 Bridge  145                       Figure B.10. Response Spectra for Toms Place 26Nov2006 (PGA=0.015g), Highway 395 Bridge  146                       Figure B.11. Response Spectra for Mammoth Lakes 12Jun2007 (PGA=0.051g), Highway 395 Bridge  147                       Figure B.12.  Response Spectra for Gilroy 13May2002 (PGA=0.014g), Alviso Overpass (Bridge K)  148                       Figure B.13. Response Spectra for Gilroy 13May2002 (PGA=0.014g), Alviso Overpass (Bridge L)  149                       Figure B.14.  Response Spectra for Ferndale 09Jan2010 (PGA=0.077g), Murray Road Bridge  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0074210/manifest

Comment

Related Items