DYNAMIC CHARACTERIZATION OF THE BASILICA OF S. MARIA DI COLLEMAGGIO AFTER THE EARTHQUAKE OF 2009 BY MEANS OF OPERATIONAL MODAL ANALYSIS

Transcription

1 DYNAMIC CHARACTERIZATION OF THE BASILICA OF S. MARIA DI COLLEMAGGIO AFTER THE EARTHQUAKE OF 2009 BY MEANS OF OPERATIONAL MODAL ANALYSIS 17 th International Conference on Computational Methods and Experimental Measurements 5 th 7 th May, Opatija (Croatia)

2 SCENARIO SANTA MARIA DI COLLEMAGGIO IS ONE OF THE MOST ATTRACTIVE CHURCHES IN CENTRAL ITALY. SPLENDID EXAMPLE OF ROMANESQUE ARCHITECTURE IN ABRUZZO, WAS BUILT AT THE END OF '200 ON A HILL IN THE CITY OF L'AQUILA. ITS PRESENT SHAPE IS THE RESULT OF MANY CHANGES, MOST OF WHICH AS A RESULT OF THE STRONG EARTHQUAKES THAT OCCURRED THROUGHOUT THE CENTURIES. IT HAS, INDEED, SUFFERED MANY INSULTS INCLUDING THE DEVASTATING EARTHQUAKE OF 1703: THE LONGITUDINAL AISLES, RECONFIGURED RECENTLY, MANIFESTED A GOOD BEHAVIOR, WHILE THE ENTIRE PORTION OF THE TRANSEPT WAS THE SUBJECT OF WIDESPREAD COLLAPSE. ON APRIL 6 TH 2009, THE EARTHQUAKE THAT PARTIALLY DESTROYED THE CITY OF L'AQUILA HAS CAUSED THE TOTAL COLLAPSE OF ALL ELEVATED STRUCTURES IN THE TRANSEPT OF THE BASILICA.

3 EARTHQUAKE EFFECTS IT IS MANIFESTED IN THE COLLAPSE OF THE WALL OF THE TRIUMPHAL ARCH, OF TERMINAL PILLARS OF THE NAVE, OF THE VAULTED STRUCTURES AND THOSE OF COVERAGE, RESULTING IN A PILE OF RUBBLE INSIDE THE BOX WALLS, REMAINED BROADLY INTACT AND COMPLETE UP TO THE TOP PORTION.

4 PROVISIONAL INTERVENTIONS UNDERPINNING OF ARCHES OF THE CENTRAL NAVE; POLYETHYLENE TAPE RIMMING OF PILLARS; TRANSVERSE AND LONGITUDINAL CHAINS ALONG THE NAVE AND IN THE APSE; CARBON FIBER PREPREG TAPES TO MASONRY ELEMENTS, WHERE FRACTURES OCCURRED; TRUSSES TO SUPPORT THE APSE VAULTS.

5 DESCRIPTION OF INTERVENTION REMAINING BOX WALLS SUBDIVIDED INTO 5 SUBSTRUCURES OPERATIONAL MODAL TESTING IMPACT TESTING TESTS PERFORMED ON DIFFERENT DAYS DATA RELATED BY MEANS OF COMMON REFERENCE POINTS.

6 SENSOR POSITIONING SUBSTRUCTURE 1 SUBSTRUCTURE 2 SUBSTRUCTURE 3 SUBSTRUCTURE 4 SUBSTRUCTURE 5 THE SUBSTRUCTURE 5 SHOWS THE MAIN FAÇADE

7 SENSORS, ACQUISITION SYSTEMS AND DEVICES AT A GLANCE

8 KINEMATICS MODELLING FACADE SUBSTRUCTURE 2

9 MEASUREMENTS APPARATUS ARRAY OF 8 ACCELEROMETERS CONNECTED VIA CABLE TO A DYNAMIC ACQUSITION SYSTEM 2 DATA ACQUISITION SYSTEMS, 8 CHANNELS EACH, GPS SYNCHRONIZED ON ABSOLUTE TIME SYNCHRONIZATION AS EXCELLENT SOLUTION FOR 2 OR MANY LINES, REFERRED TO GROUP OF SENSORS, SIMOULTANEOUS SMAPLING SINGLE SCANNING SYSTEM WOULD BE NEEDED LONG CABLE (LOSS OF QUALITY, DIFFICULTY OF INSTALLATION)

10 DATA ACQUISITION SYSTEMS PCB 483C05 POWER SUPPLY AND ICP SIGNAL CONDITIONER IMC CS7800/1 ACQUISITION SYSTEM VOLTAGE SIGNAL 8 CHANNELS WITH SIMULTANEOUS SAMPLING 24 BIT INTERNAL A/D CONVERTER ELECTRICAL NOISE LESS THEN 5nV RMS SPECTRUM ANALYZER EQUIPPED WITH A DSP LAPTOP PC PENTIUM FAMILY GARMIN GPS ANTENNA LMS SCADAS MOBILE SCR011 ACQUISITION SYSTEM V8-E 8 SIMULTANEOUS CHANNELS FOR VOLTAGE AND IEPE SIGNALS 24 BIT A/D CONVERTER SAMPLING UP TO 204,8 khz DYNAMIC RANGE OF 150 db SFR HP HANDHELD COMPUTER LMS GPS ANTENNA

11 ACCELEROMETERS IEPE SEISMIC TYPE ACCELEROMETERS (HIGH-SENSITIVITY, LOW NOISE AND CAPABLE OF SAMPLING AT LOW FREQUENCIES) HAVE BEEN UTILIZED

12 OPERATIONAL MODAL ANALYSIS Creation and filtering of crosspower from TH Reference selection SSI extraction method (not used) Merging of two sets with common references Validation phase Geometry creation Selection of useful crosspower set Frequency domain extraction method Synt&measured confrontation AUTOMAC

13 OUTPUT-ONLY FREQUENCY DOMAIN MODEL OUTPUT SPECTRA AS PRIMARY DATA HALF SPECTRA OBTAINED BY USING ONLY THE CORRELATIONS WITH A POSITIVE LAG L w R0 j k t Syy ( ) wk Rke 0 2 k 1 RELATION BETWEEN HALF SPECTRA AND THE FULL SPECTRA H S ( ) S ( ) ( S ( )) yy yy WHITE NOISE INPUT ASSUMPTION yy S yy ( j ) n i 1 i g i j i * i g * i j * i THE OUTPUT CORRELATIONS ARE EQUIVALENT TO IMPULSE RESPONSE

14 ALGORiTHM OF EXTRACTION OF MODAL SHAPES FURTHER POLYREFERENCE EVOLUTION OF THE LSCF ESTIMATION METHOD 17 th International Conference on Computational Methods and Experimental Measurements 5 th 7 th May, Opatija (Croatia)

15 RESULTS SUBSTRUCTURE 2 COMPARISON ON THE BASIS OF THE INDEX OF MODAL CORRELATION

16 RESULTS SUBSTRUCTURE 5 (FACADE) IN FACT, THE FAÇADE IS LAID ASIDE ORTHOGONAL TO THE DIRECTION OF THE EARTHQUAKE

17 COMPARISON WITH IMPACT HAMMER RESULTS FOR THE FIRST THREE SUBSTRUCTURES IT WAS ALSO MADE AN IMPULSIVE MODAL ANALYSIS, WITH AN INSTRUMENTED HAMMER.

18 RESULTS SUMMARY TABLE FOR SUBSTRUCTURES FROM 1 TO 4, FLEXURAL FIRST MODE BETWEEN 1,02 AND 1,04 HZ, FOR HIGHER MODES THERE ARE SIGNIFICANT DIFFERENCES Substruct. 1 Holy Door wall Substruct. 2 East external wall Substruct. 3 West colonnade Substruct. 4 Transept side Substruct. 5 Facade Impact Test Global f [Hz] z (%) f [Hz] z (%) f [ Hz] z (%) f [Hz] z (%) f [H z] z (%) f [Hz] z (%) Mode 1 1,03 1,11 1,03 1,61 1,04 0,94 1,02 1,49 2,62 1,15 1,05 2,59 Mode 2 1,07 0,73 1,44 1,49 1,09 1,29 2,07 1,14 4,28 0,47 1,42 3,02 Mode 3 2,08 1,34 2,07 1,23 1,98 0,47 2,77 0,70 5,79 1,25 2,10 1,63 Mode 4 2,58 0,42 2,29 1,00 2,65 0,75 3,61 0,44 7,50 0,32 2,66 3,32 HIGHER MODES Mode 5 3,07 0,67 4,71 0,91 2,76 0,64 4,70 0,39 8,30 1,10 3,36 2,06 THE RESULTS OF IMPACT HAMMER COMPARABLE WITH THOSE OF OMA. REDUCTION OF THE STIFFNESS VARYING FROM 20% TO 30% WITH REF. TO PREVIOUS SURVEYS

19 CONCLUSIONS DETERMINATION OF CHARACTERISTICS MODES AND DYNAMIC PARAMETERS FOR ALL THE SUBTRUTCTURES INVESTIGATED MONOLITHIC BEHAVIOR DEMONSTRATED WITH THE SAME FIRST FREQUENCY FOR FOUR OF THE SUBSTRUCTURES INVESTIGATED GOOD ENSEMBLE BEHAVIOR WITH VALUES SLIGHTLY LOWER THAN BEFORE THE FAÇADE HAS UNDERGONE THE GREATEST LOOSENING METHOD WHICH IS THE MOST SUITABLE, EXPEDITIOUS AND EFFECTIVE FOR THIS TYPE OF MANUFACTURED MASONRY POWERFUL, VERSATILE AND SOPHISTICATED SOFTWARE ALGORITHM BASED ON THE EXTRACTION OF MODE SHAPES MEASURING CHAIN OF EXCELLENT PERFORMANCE

20 THIS WORK WAS DONE BY ESSEBI LTD IN COLLABORATION WITH THE CONSORZIO PER LA SPERIMENTAZIONE EDILIZIA OF THE UNIVERSITY OF L'AQUILA. A HEARTFELT THANKS GOES TO THE ENI OIL COMPANY WITHOUT THE FINANCING OF WHICH THIS WORK WOULD NOT HAVE BEEN ABLE TO PLAY. THANKS TO ALL THE ATTENDEES FOR COMING.