IMPLEMENTATION OF THE EC 8.3: ASSESSMENT AND INTERVENTIONS IN EARTHQUAKE PRONE AREAS

1. IMPLEMENTATION OF THE EC 8.3: ASSESSMENT AND INTERVENTIONS IN EARTHQUAKE PRONE AREAS Athens - April 12, 2013 Performance basedassessment of ancientmasonrybuildings. Outcome of the Europeanproject PERPETUATE Sergio Lagomarsino University of Genoa, Italy sergio.lagomarsino@unige.it PERPETUATE PERformance-basedaPproach to EarthquakeproTection of cUlturAlheriTage in European and mediterraneancountries

2. Earthquake protection of cultural heritage PERformance-based aPproach to Earthquake proTection of cUlturAlheriTage in European and mediterranean countries Main objectives of the project: Development of European Guidelines for the evaluation and mitigation of seismic risk to cultural heritage assets. Both architectonic assets (historic buildings; macroelements) and artistic assets (frescos, stucco-works, statues, pinnacles, battlements, banisters, balconies …) will be considered. Only masonry structures will be considered. www.perpetuate.eu

3. Partners • The Consortium consists of: • 6 Universities (Genoa, Thessaloniki, Athens, Ljubljana, Bath, Algiers) • 2 Public/Research Institutions (ENEA, Italy; BRGM, France) • 3 SMEs from Slovenia(ZRMK) and Italy (CENACOLO, PHASE).

4. Basic principles of PERPETUATE procedure • The protection of cultural heritage needs an improvement in methods of analysis and assessment procedures, rather than the development of new intervention techniques. • A reliable assessment procedure is the main tool to respect the principle of “minimum intervention” under the constraint of structural safety. • The displacement-basedapproachfor vulnerabilityassessment of cultural heritageassets and design of interventionsisadoptedas standard method of analysis. • Nonlinear models are necessary. Nonlinear static (pushover) analyses are considered as the main tool for the application of the assessment procedure. Nonlinear dynamic analyses are considered as an alternative tool, only for certain types of assets. • If, after a reliable seismic assessment, it comes out the monument is not safe, then its retrofitting is unavoidable, first of all in order to preserve its life along the time and also for the safety of occupants.

5. Basic principles of PERPETUATE procedure The outcome of the assessment is the maximum Intensity Measure (e.g. PGA) compatible with the fulfilment of each performance level that has to be considered.

6. Basic principles of PERPETUATE procedure • The format of the assessment proposed by PERPETUATE guidelines is deterministic, except for the occurrence of the earthquake, as well as in all codes and recommendations worldwide adopted at present. • However, it is well know many uncertainties, aleatory and epistemic, affect the assessment of an existing masonry building, with reference to: • a) the characteristics of seismic input (duration, frequency content, etc.) • b) the reliability of mechanical models • c) the material parameters • d) the incomplete knowledge of the construction • PERPETUATE takes into account probabilistic aspects in some steps of the procedure: • acceptance criteria for the definition of PLs • sensitivity analysis for drawing the protocol of in-situ investigations and defining the Confidence Factors

7. Basic steps of PERPETUATE procedure

8. Classification of architectonic assets PERPETUATE considers a classification of architectonicassetswhichisuseful for addressing the choice of propermechanicalmodelsto be adopted for the assessment.

9. Classification of architectonic assets It is functional to model main seismic behaviour of buildings

10. Classification of architectonic assets BOX-TYPE STRUCTURES (vertical walls and horizontal floors) A1 Palaces A2 Castles A3 Religious houses A4 Caravansaries

11. Classification of architectonic assets WIDE HALLS WITHOUT INTERMEDIATE FLOORS (macroelements) B6 Hammam B1 Churches B2 Mosques

12. Classification of architectonic assets SLENDER MASONRY STRUCTURES C1 Towers C2 Bell Towers C3 Minarets C4 Lighthouses

13. Classification of architectonic assets ARCHED AND VAULTED STRUCTURES D1 Triumphal arches D4 Cloisters

14. Classification of architectonic assets MASSIVE MASONRY CONSTRUCTIONS E2 Defensive city walls E1 Fortress

15. Classification of architectonic assets DRY BLOCKS STRUCTURES F3 Obelisks F1 Columns F2 Trilithes

16. Classification of architectonic assets AGGREGATED BUILDINGS IN HISTORICAL CENTRES Navelli, L’Aquila, Italy Skofja Loka, Slovenia

17. From classification to mechanical models ARCHITECTONIC CLASSES A B C D E F MODELS CLASSES CORRELATION CCLM - Continuum ConstitutiveLawsmodels SEM - StructuralElementmodels DIM – Discrete Interface models MBM – Macro Block models

18. Classification and modelling Some types of assets(can be studied by a global 3D model, while in othercasesitisnecessary to develop more thanone model, even of differenttypes. Moreover, the assessmentrequirestakingintoaccontthe possibleactivation of localmechanisms.

19. Safety and conservation requirements Performance Levels PERFORMANCE LEVELS USE and HUMAN LIFE ARTISTIC ASSETS BUILDING CONSERVATION EC8 part 3 NEAR COLLAPSE SIGNIFICANT BUT RESTORABLE DAMAGE LOSS PREVENTION LIFE SAFETY IMMEDIATE OCCUPANCY RESTORABLE DAMAGE DAMAGE LIMITATION NEAR INTEGRITY OPERATIONAL NO DAMAGE

20. Safety and conservation requirements Performance Levels & correlation with damage PERFORMANCE LEVELS DAMAGE LEVEL DAMAGE LEVEL USE and HUMAN LIFE ARTISTIC ASSETS BUILDING CONSERVATION 4 NEAR COLLAPSE SIGNIFICANT BUT RESTORABLE DAMAGE LOSS PREVENTION 3 3 LIFE SAFETY STRUCTURAL ELEMENT – LOCAL SCALE GLOBAL SCALE – WHOLE ASSET IMMEDIATE OCCUPANCY RESTORABLE DAMAGE DAMAGE LIMITATION 2 2 NEAR INTEGRITY 1 1 OPERATIONAL NO DAMAGE

21. Safety and conservation requirements Performance and damage Levels & Acceptance Criteria

22. Safety and conservation requirements Performance Levels & corresponding Target Seismic Demand Levels

23. Safety and conservation requirements Performance Levels & corresponding Target Seismic Demand Levels IMPORTANCE FACTORS EC8.3 - 225 years

24. Seismic hazard A proper definition of the seismic demand is addressed by the information and choices that have been assumed in the first step (Classification). • Intensity Measure (IM): it depends on the Class of the architectonic asset. • Seismic Input can be described by: • in case of nonlinear static analyses, an Acceleration-Displacement Response Spectrum (ADRS), completely defined for the specific site of the building under investigation as a function the Intensity Measure (IM); • in case of non linear dynamic analyses, a proper set of time-histories, selected from real recorded accelerograms, obtained through numerical models of the seismic source and the propagation to the site or artificially generated, in order to be compatible with a target response spectrum (this last option is questionable).

25. Seismic hazard ProbabilisticSeismicHazardAssessment e.g. PGA & PGD

26. Seismic hazard For architectonic assets that require the adoption of an IM representative of the ADRS in the long periods range (e.g. Class F), or more than one IM (Vector-Valued PSHA), the shape of the response spectrum must change, in order to be compatible with the information provided by GMPEs for short a long periods

27. Seismic hazard Floorspectra (for the assessment of localmechanisms in the higherlevels)

28. As built information In this sub-step, geometrical, technological and mechanical features of the asset are analysed in depth, with the aim of defining the structural model of the building and related artistic assets. Although it is necessary to investigate in detail the construction, in case of ancient masonry buildings it is necessary to consider that: in cultural heritage assets, due to conservation requirements, the impact of investigations should be minimized; in seismic assessment of an existing building epistemic uncertainties due to the incomplete knowledge and reliability of models add up to aleatory uncertainties, in particular related to material parameters. The approach adopted in PERPETUATE is based on the use of Confidence Factors.

29. As built information Maindrawbacks Approachused by codes Usually the reaching of a certain knowledge level implies an almost homogeneous degree of accuracy to be reached on all the different knowledge aspects (material, geometry, constructive details). The confidence factor is conventionally applied only to the parameter which is assumed to mainly affect the structural response. This parameter (or set of parameters) is proposed by codes a priori: usually, in fact, it coincides with strength mechanical parameters (only some codes – such as the ASCE/SEI 41-06 – proposes to apply it to the drift values, only in case of deformed controlled modes). The value of the confidence factor is conventionally proposed by codes only as a function of the reached knowledge level. Since in general different parameters may differently (more or less significantly) affect the structural behaviour, it seems more reasonable to allows a calibration of the improvement of knowledge required as a function of the degree of sensitivity of the response. In some cases, a lower level of knowledge has been accepted because no remarkable building details have been analysed notwithstanding these abovementioned information could not be so relevant in terms of seismic safety. In general, as a function of the structure examined, this conventional assumption could not be the best one. The actual variability of from time to time examined parameters which may affect the response of the building is not considered.

30. As built information The idea is that investigation program should be based on a preliminary sensitivity analysisaimed to: identify the main parameters to be investigated; define proper Confidence Factors (CFs), to be used for the assessment in order to take into account uncertainties. The identification of main parameters influencing the structural response of the asset allows to finalize the investigation to few important points (thus reducing costs and time) and to reduce the number of destructive tests. The calibration of CFs on the basis of sensitivity analyses instead of a-priori assumptions (as usually done for standard buildings) provides more reliable models and results.

31. As built information: sensitivity analysis

32. As built information: sensitivity analysis • Notes on steps d) related to the execution of sensititvity analysis Examples of sensitivity analyses results SENSITIVITY TO DRIFT SENSITIVITY TO MECHANICAL PARAMTERS SENSITIVITY TO MASSES

33. As built information Sensitivity to random variables Sensitivity to epistemicuncertainties Performance Level

34. As built information In sensitivity analyses both statistical uncertainties, treated by proper random variables, and epistemic uncertainties, treated by a logic tree approach, are considered.

35. As built information: sensitivity analysis Steps of the proposed procedure: Achievement of a “basic” knowledge level to preliminarlyidentify the suitable model (or models) to be adopted for the seismic assessment. Identification of all parameters or set of parameters (in case of parameters dependent on each other) affecting the model. For each parameter, identification of the more rational range of variation. Execution of the sensitivity analysis onto selected parameters in order to evaluate how much each one really affects the seismic behaviour of the examined building. Non linear static analysis is assumed as the standard one. Attribution of a “sensitivity class” (low, medium or high), on the basis of the post processing of results provided from the sensitivity analysis, for each parameter. Plan of the investigation and testing by using the results obtained from steps d) and e). Execution of investigations and tests; Definition of the confidence factor, possible updating of the mean value of parameters (on the basis of tests/investigations results) and final definition of parameters to be used in models for the seismic assessment.

36. As built information: sensitivity analysis • Notes on steps h) related to the definition of confidence factor According to the common approaches based on the use of confidence factor , it seems reasonable to apply CF to the “main parameter” which affect the structural seismic response. In the case of PERPETUATE procedure, it is possible choosing the “main parameter” among those associated to a high sensitivity class and not conventionally a priori. Regarding the value to be adopted for CF it has to take into account : the actual variability of the parameter which will be applied to (by considering the related fk value); the “remaining” uncertainties associated to the incomplete knowledge process (since it could be not possible to reach the maximum knowledge level on all parameters ).

37. As built information: sensitivity analysis • Notes on steps h) related to the definition of confidence factor It is necessary to increase the knowledge level for parameters that have a higher sensitivity class. • Hence CF is computed as: where fc corresponds to fkassociated to the k-th parameter assumed as reference as “main parameter” that affects the seismic response.

38. As built information: sensitivity analysis Measure of the sensitivity of the structural performance at n-thfactors Measure of the sensitivity of the structural performance at k-th parameter

39. Modelling and verification procedures Verification procedures adopted:

40. Modelling and verification procedures METHODS OF ANALYSIS AND ASSESSMENT PROCEDURES Pushoveranalyses and CapacitySpectrum Method STANDARD METHOD Nonlineardynamicanalysis (IDA) MOREACCURATE METHOD – APPLICABLE WITH A REASONABLE COMPUTATIONAL EFFORT ONLY FOR SOME CLASSES Linear elasticanalysis and assessmentbased on behaviourfactor (q) ONLY IN CASE OF VERY COMPLEX ASSETS

41. Modelling and verification procedures ASSESSMENT PROCEDURES AND ARCHITECTONIC ASSET CLASSES Homogeneous and coherentcriteria are adopted for the development of the assessmentprocedureswhichhave to be applied to the differentclasses of assetsdefined in D4 (from A to F) In particularthreemaincasesmay be identified: • Assetcomposed by a single macroelement; • Complexassetsdescribedby a single capacity curve; • Complexassetsdescribed by a Ncapacitycurves Case 3) Case 1) Case 2)

42. Modelling and verification procedures ASSESSMENT PROCEDURES AND ARCHITECTONIC ASSET CLASSES A B C D E F

43. Assets composed by a single macroelement SUMMARY OF THE PROCEDURE: • Pushoveranalysis • Definition of PLs on the pushover curve • Capacitycurve – Equivalent SDOF • Seismicdemand and hazard curve (IntensityMeasure - IM) • Computationof the maximum IM value (IMPLi) compatible with the givenperfomancelevel (PLi) • Comparisonof IMPLiwith the corresponding target value IM*PLi • Use of seismicassessmentoutcome for the rehabilitationdecisions

44. Assets composed by a single macroelement NON LINEAR KINEMATIC ANALYSIS MACRO BLOCK MODELS (MBM) Seismicforcesproportional to massesapplied in eachblock NON LINEAR STATIC ANALYSIS CONTINUUM CONSTITUTIVE LAWS MODELS (CCLM) AND STRUCTURAL ELEMENT MODELS (SEM) Seismicforcesproportional to the first mode • PUSHOVER ANALYSIS F and D CLASSES CLASS C

45. Assets composed by a single macroelement NON LINEAR KINEMATIC ANALYSIS MBM models F and D classes NON LINEAR STATIC ANALYSIS CCLM and SEM models C class 2. Definition of PLs on the pushover curve

46. Assets composed by a single macroelement NON LINEAR KINEMATIC ANALYSIS MBM models It is assumed as a fundamental shape the block displacements of the kinematism NON LINEAR STATIC ANALYSIS CCLM and SEM models It is based on the fundamental modal shape 3. Capacity curve – Equivalent SDOF d=u/G ay=Vy /m*G

47. Assets composed by a single macroelement • Seismicdemandis an Acceleration DisplacementResponseSpectra (ADRS). • Itmay be defined (for a givensoilcondition): • by a set of parameters (PGA, TC, TB, F0, …..) and analyticalfunctions; • by a set of values (T, Sa) • An IntensityMeasure (IM) has to be definedasrepresentative for the ProbabilisticSeismicHazardAssessment (PSHA) • In case of more IM  Vector-valued PSHA • In general, the Peak Ground Acceleration (PGA) isassumedas IM • The IM varies with the returnperiod (TR) (thatis with the annualprobability of occurrence 4. Seismicdemand and hazard curve (IntensityMeasure - IM) By a set of parameters and analytical functions By a set of values

48. Assets composed by a single macroelement • PERPETUATE procedure needs, for eachPLi, the evaluation of IMPLi: to this end, itissufficient to define the dampingcoefficient (xPLi) and period (TPLi) • Use of CapacitySpectrum Method (overdampedspectra) 5. Maximum IM value (IMPLi) compatible with the givenperfomancelevel (PLi) Through cyclic pushover Through analitycal laws proposed in literature Law calibrated on experimental campaigns on full scale masonry building (S2 Project)

49. Assets composed by a single macroelement 5. Maximum IM value (IMPLi) compatible with the givenperfomancelevel (PLi) Sa Sd (T, IM ,x) IMPLi Sd0 (T,x) 1 TPLi dPLi Sd

50. Assets composed by a single macroelement • By using the hazard curve the returnperiodTR,PLicorresponding to IMPLimay be evaluated • The assessmentconsists in the comparison of TR,Pli with T*R,Pli. • Itis positive ifTR,Pli > T*R,Pli. 6. Comparison of IMPLiwith the corresponding target value IM*PLi IMPLi lRPLiT RPLi