The issue of seismic assessment of existing RC buildings has been extensively studied in the last few years and the international reference framework, both with regard to the scientific research and the development of technical codes, is very wide. Nevertheless, there are still a lot of challenging questions about the definition of reliable numerical models and methods of analysis, which are strongly affected by many uncertainty sources (knowledge of structural details, material properties, seismic input; accuracy and reliability of capacity models and discretization strategies). The management of these issues, especially in view of practice-oriented applications, requires the availability of effective strategies, so to allow a probabilistic assessment approach that can be relatively accessible in terms of implementation hurdle the computational time. After an extensive background about the approaches to vulnerability assessment proposed by recent scientific literature and technical codes, the dissertation discusses the critical aspects related to some assumptions commonly adopted in the seismic modelling of existing RC buildings, with the aim of proposing proper sanitization strategies, which can be particularly useful in view of practical applications. As a first issue, the influence on the global response of alternative modelling assumptions for secondary structural elements such as slabs is investigated. The usual hypothesis of rigid floor is assessed by performing a sensitivity analysis based on several parameters, which are particularly significant for the structural response evaluation. Then, based on the results of the analyses, a numerical procedure for modelling the floor system is proposed, defining an orthotropic equivalent shell element capable to simulate the in-plan stiffness of the floor. The methodology actually increases the computational efforts, but has the significant advantage of avoiding aprioristic assumptions about the floor stiffness. An application of the method to the numerical modelling of existing RC buildings is then proposed, by appraising the variation of results in comparison with alternative models for considering in-plan stiffness (namely, equivalent strut models). Lastly, the application possibilities of the proposed procedure are appraised, by presenting a number of examples. As an additional effect, the presence of infill panels is considered, in the perspective of retrofit solutions. More specifically, the possibility of increasing the capacity to horizontal actions by reinforcing the infilled frames or by introducing additional RC shear walls on the building perimeter is appraised. The second issue addressed in the dissertation is the definition of the most effective methodology to be used for identifying the structural response both in the elastic and inelastic field. After a review of the nonlinear methods of analysis provided by the scientific literature, both static and dynamic, the dissertation presents some applications of the pushover method, which is by far the most popular choice of practitioners. Firstly, an application of conventional pushover analysis is performed on a set of ideal buildings, with the aim of appraising the role of the control node position. Anyway, as highlighted by current technical laws (Italian building code and Eurocode 8), nonlinear static procedure cannot be always applied in its conventional formulation. In particular, some limitations arise in the presence of structural irregularities or in the cases where higher modes have a strong influence. With the aim to bridge these gaps, a solution can be represented by non-conventional methods as multimodal or adaptive pushover analysis. With regard to this question, a simplified multimodal pushover procedure is proposed in the dissertation. The main advantage of the proposal is represented by the easiness of application, thanks to the adoption of a single load profile in the computation, which is moreover an approach very familiar to practitioners. For assessing the reliability of the procedure, it is tested on a real case study characterized by relevant dynamic irregularity and a consistent inhomogeneity of in-situ materials. The final part of the dissertation is devoted to the possibility of extensively bringing the concepts at the base of Performance Based Earthquake Engineering (PBEE) to a wider audience of users, considering that this method has a high scientific relevance for the assessment of existing RC buildings. Generally, the application of PBEE needs a specialist knowledge about probability theories and about nonlinear modelling and analysis, which are skills not always common among practitioners. With the aim of reducing these obstacles, a methodology of nonlinear dynamic analysis is proposed, which consists in an application of the multi-stripe analysis on numerical models implemented through a commercial software. In particular, the new procedure, called Few Stripe Analysis (FSA), is applied on a sample of 15 existing RC school buildings (located in the province of Foggia, Southern Italy) and the results, in terms of damage states, are compared with the ones obtained from SPO2FRAG software, an userfriendly tool able to compute the fragility curves starting from pushover curves. Finally, a new simplified modelling procedure for estimating the global response of existing RC buildings is presented. It is able to produce 3D reduced-order models (characterized by very few degrees of freedom) starting from the geometrical and mechanical features of the case study. The main advantage of the present approach is to account for the effects predictable with MDoF models, but with low analysis time and computational efforts, with elevate convergence capacity, typical of the SDoF models. The performance of this simplified numerical modelling procedure has been tested by the application on the previously mentioned sample of school buildings and comparing the results, in terms of structural response, damage states and confidence levels, with the ones previously obtained from the application of FSA. The relevance and perspective impact of the research work here presented should be seen in the wider field of the vulnerability analysis of the building stock at the regional scale, which is a crucial issue for the scientific community and for the civil society. Governments and administrations are invested with the difficult task of providing mitigation strategies for the seismic risk for a very wide and inhomogeneous portfolio of buildings and the economic resources are often very limited. Therefore, the development of methods for estimating the vulnerability with limited data has been a subject of intense research activity. The framework that is depicted in the dissertation can provide a tool potentially very impactful, since it could allow, by the exploitation of the 3D Reduced Order Models combined with FSA, to overcome the well-known limitations of empirical vulnerability approaches in favor of mechanical based methods managed in a full probabilistic framework.
Il problema della verifica di vulnerabilità sismica di edifici esistenti in calcestruzzo armato è stato oggetto negli ultimi anni di studi approfonditi, che hanno favorito lo sviluppo di un quadro di riferimento internazionale sul tema molto ampio, sia dal punto di vista della ricerca scientifica che da quello delle normative tecniche vigenti. Tuttavia, sono ancora molte le questioni irrisolte a riguardo di temi come la modellazione numerica e i metodi di analisi sismica, fasi fortemente influenzate da continue fonti di incertezza (conoscenza dei dettagli geometrici e strutturali, proprietà dei materiali, input sismico, accuratezza e affidabilità di modelli di capacità e strategie di discretizzazione). Ai fini di una valutazione affidabile delle prestazioni sismiche, tali problematiche richiedono lo sviluppo di strategie di modellazione e analisi innovative ed efficaci, soprattutto da un punto di vista di una accurata valutazione probabilistica e con uno sguardo attento alla pratica progettuale, dove la facilità di implementazione e i tempi di calcolo assumono un’importanza prioritaria. Dopo un’estesa ricerca bibliografica degli approcci proposti e utilizzati per effettuare verifiche di vulnerabilità sismica di edifici esistenti in calcestruzzo armato, proposti dalla letteratura scientifica e dalle normative tecniche vigenti, nella tesi sono stati discussi inizialmente alcuni aspetti critici di modellazione, relativi alle consuete ipotesi semplificative adottate. Nella fattispecie, l’influenza dell’ipotesi di piano rigido, con riferimento agli elementi strutturali secondari come il solaio, è stata analizzata, con l’obiettivo di proporre un’idonea strategia efficiente di modellazione per una pratica applicazione, rivolta a ricercatori e professionisti. Ciò stante, un’analisi iniziale di sensibilità è stata condotta, investigando quali parametri influenzano significativamente la risposta sismica globale della tipologia di edifici in oggetto. Sulla base dei risultati ottenuti, una nuova procedura numerica di modellazione dell’impalcato è stata proposta, atta a definire una piastra ortotropa equivalente capace di simulare la reale rigidezza nel piano, per azioni orizzontali. La metodologia adottata, nonostante incrementi lo sforzo computazionale dell’analisi, ha il vantaggio di evitare le assunzioni aprioristiche sulla rigidezza dell’impalcato. Al fine di validare quanto proposto, il metodo è stato applicato ad un edificio esistente in calcestruzzo armato, valutando i risultati e comparandoli con altre metodologie proposte dalla letteratura scientifica per considerare il comportamento nel piano dell’impalcato, come quella a puntoni equivalenti. Infine, è stata valutata la possibilità di applicare la procedura nei casi in cui si considera l’influenza delle tamponature esterne e successivamente, in una prospettiva di miglioramento o adeguamento sismico dell’edificio. In quest’ultimi casi, la prestazione dell’edificio alle azioni orizzontali è stata migliorata, mediante l’uso di tamponature rinforzate e mediante l’inserimento di pareti in calcestruzzo armato sul perimetro dell’edificio. Per quanto riguarda la fase di analisi sismica, stabilire quale sia la metodologia più efficace per identificare la risposta strutturale in campo elastico e inelastico assume una grande importanza, considerando soprattutto la vasta casistica di procedure proposte dalla letteratura scientifica e dalle normative tecniche vigenti. A valle di un’estesa valutazione di quest’ultime, con particolare attenzione ai metodi di analisi non lineari, sia statici che dinamici, la dissertazione presenta alcune applicazioni di analisi statiche non lineari, metodo che rappresenta la prima scelta da parte dei professionisti. Inizialmente, un’applicazione di analisi statica non lineare convenzionale è stata condotta su un campione di edifici esistenti ideali in calcestruzzo armato, con l’obiettivo di verificare il ruolo del nodo di controllo. Tuttavia, come già evidenziato dalle normative tecniche vigenti (Normativa Tecnica Italiana e Eurocodice 8), le procedure di analisi statica non lineare non possono essere sempre applicate, a causa di alcune limitazioni dovute alle caratteristiche dell’edificio analizzato, come le irregolarità e la forte influenza dei modi superiori. Con l’obiettivo di proporre una strategia che possa colmare i limiti sopraelencati, una possibile soluzione è rappresentata dai metodi non convenzionali come le analisi statiche non lineari multimodali o adattive. A questo proposito, una procedura semplificata di analisi statica non lineare multimodale è stata proposta. La peculiarità di tale metodologia è dovuta ad un algoritmo capace di fornire un singolo profilo di carico, facilmente implementabile nelle stesse modalità di un’analisi convenzionale. Al fine di verificare l’efficienza del metodo, quest’ultimo è stato applicato ad un edificio esistente in calcestruzzo armato, caratterizzato da irregolarità dinamiche e da elevata inomogeneità dei materiali in situ. Nella parte finale della tesi, è stata analizzata la possibilità di implementare i concetti alla base del Performance Based Earthquake Engineering (PBEE), metodo di elevata rilevanza scientifica, per applicazioni pratiche nella verifica di vulnerabilità sismica di edifici in calcestruzzo armato. Generalmente, l’applicazione del PBEE richiede conoscenze specifiche circa le teorie della probabilità e competenze specialistiche nel campo della modellazione e analisi non lineare, qualità non sempre comuni tra i professionisti. Con l’obiettivo di ridurre i sopramenzionati ostacoli, una metodologia di analisi dinamica non lineare è stata proposta, consistente in un’applicazione del metodo “multi stripe analysis” su modelli numerici redatti con programmi di calcolo commerciali. Nella fattispecie, la nuova procedura, chiamata “Few Stripe Analysis” (FSA) è stata applicata e testata su un campione di 15 edifici scolastici esistenti in calcestruzzo armato (nella provincia di Foggia, Sud Italia) e i risultati ottenuti, in termini di stato di danno e curve di fragilità, sono stati confrontati con quelli ottenuti utilizzando il programma di calcolo SPO2FRAG. Quest’ultimo consente di calcolare curve di fragilità, partendo da curve di capacità ottenute da analisi statiche non lineari. Infine, una nuova procedura di modellazione per valutare la risposta sismica globale di edifici in calcestruzzo armato è stata proposta. In particolare, la metodologia consente di produrre modelli 3D ad ordine ridotto (caratterizzati da pochi gradi di libertà), partendo dalle caratteristiche geometriche e meccaniche di un edificio esistente. Il vantaggio principale del presente approccio è quello di cogliere molti degli effetti predicibili con un MDoF, ma con bassi tempi di calcolo e analisi e elevata capacità di convergenza, caratteristiche tipiche dei modelli SDoF. L’efficienza di questi modelli semplificati è stata testata sul campione di edifici esistenti sopramenzionato e i risultati, in termini di risposta strutturale, stato di danno e livello di confidenza, sono stati confrontati con quelli ottenuti precedentemente dall’applicazione della metodologia FSA. La rilevanza e l’impatto futuro del lavoro di ricerca presentato può essere valutato in una prospettiva più ampia e relativa ad un’analisi di vulnerabilità del patrimonio costruito a scala territoriale, che risulta essere attualmente un aspetto critico sia per la comunità scientifica che per le autorità governative. Infatti quest’ultime hanno il difficile compito di proporre strategie di mitigazione del rischio sismico per un ampio e disomogeneo patrimonio strutturale, ma con risorse economiche spesso molto limitate. Pertanto, lo sviluppo di metodologie per la stima della vulnerabilità basata su dati limitati è un tema soggetto ad intense attività di ricerca. Le proposte presentate nella tesi possono fornire un potenziale strumento di analisi di grande utilità, in quanto potrebbero consentire, attraverso l’uso dei modelli 3D ad ordine ridotto combinati con la metodologia FSA, di superare le ben note limitazioni mostrate dagli approcci empirici, a favore di metodi meccanici, utilizzati in un quadro completo di analisi probabilistica.
Advanced strategies for the seismic assessment of existing RC moment-frame buildings: appraisal of modelling assumptions and development of parsimonious PBEE-based methods of analysis / Ruggieri, Sergio. - ELETTRONICO. - (2019). [10.60576/poliba/iris/ruggieri-sergio_phd2019]
Advanced strategies for the seismic assessment of existing RC moment-frame buildings: appraisal of modelling assumptions and development of parsimonious PBEE-based methods of analysis
Ruggieri, Sergio
2019-01-01
Abstract
The issue of seismic assessment of existing RC buildings has been extensively studied in the last few years and the international reference framework, both with regard to the scientific research and the development of technical codes, is very wide. Nevertheless, there are still a lot of challenging questions about the definition of reliable numerical models and methods of analysis, which are strongly affected by many uncertainty sources (knowledge of structural details, material properties, seismic input; accuracy and reliability of capacity models and discretization strategies). The management of these issues, especially in view of practice-oriented applications, requires the availability of effective strategies, so to allow a probabilistic assessment approach that can be relatively accessible in terms of implementation hurdle the computational time. After an extensive background about the approaches to vulnerability assessment proposed by recent scientific literature and technical codes, the dissertation discusses the critical aspects related to some assumptions commonly adopted in the seismic modelling of existing RC buildings, with the aim of proposing proper sanitization strategies, which can be particularly useful in view of practical applications. As a first issue, the influence on the global response of alternative modelling assumptions for secondary structural elements such as slabs is investigated. The usual hypothesis of rigid floor is assessed by performing a sensitivity analysis based on several parameters, which are particularly significant for the structural response evaluation. Then, based on the results of the analyses, a numerical procedure for modelling the floor system is proposed, defining an orthotropic equivalent shell element capable to simulate the in-plan stiffness of the floor. The methodology actually increases the computational efforts, but has the significant advantage of avoiding aprioristic assumptions about the floor stiffness. An application of the method to the numerical modelling of existing RC buildings is then proposed, by appraising the variation of results in comparison with alternative models for considering in-plan stiffness (namely, equivalent strut models). Lastly, the application possibilities of the proposed procedure are appraised, by presenting a number of examples. As an additional effect, the presence of infill panels is considered, in the perspective of retrofit solutions. More specifically, the possibility of increasing the capacity to horizontal actions by reinforcing the infilled frames or by introducing additional RC shear walls on the building perimeter is appraised. The second issue addressed in the dissertation is the definition of the most effective methodology to be used for identifying the structural response both in the elastic and inelastic field. After a review of the nonlinear methods of analysis provided by the scientific literature, both static and dynamic, the dissertation presents some applications of the pushover method, which is by far the most popular choice of practitioners. Firstly, an application of conventional pushover analysis is performed on a set of ideal buildings, with the aim of appraising the role of the control node position. Anyway, as highlighted by current technical laws (Italian building code and Eurocode 8), nonlinear static procedure cannot be always applied in its conventional formulation. In particular, some limitations arise in the presence of structural irregularities or in the cases where higher modes have a strong influence. With the aim to bridge these gaps, a solution can be represented by non-conventional methods as multimodal or adaptive pushover analysis. With regard to this question, a simplified multimodal pushover procedure is proposed in the dissertation. The main advantage of the proposal is represented by the easiness of application, thanks to the adoption of a single load profile in the computation, which is moreover an approach very familiar to practitioners. For assessing the reliability of the procedure, it is tested on a real case study characterized by relevant dynamic irregularity and a consistent inhomogeneity of in-situ materials. The final part of the dissertation is devoted to the possibility of extensively bringing the concepts at the base of Performance Based Earthquake Engineering (PBEE) to a wider audience of users, considering that this method has a high scientific relevance for the assessment of existing RC buildings. Generally, the application of PBEE needs a specialist knowledge about probability theories and about nonlinear modelling and analysis, which are skills not always common among practitioners. With the aim of reducing these obstacles, a methodology of nonlinear dynamic analysis is proposed, which consists in an application of the multi-stripe analysis on numerical models implemented through a commercial software. In particular, the new procedure, called Few Stripe Analysis (FSA), is applied on a sample of 15 existing RC school buildings (located in the province of Foggia, Southern Italy) and the results, in terms of damage states, are compared with the ones obtained from SPO2FRAG software, an userfriendly tool able to compute the fragility curves starting from pushover curves. Finally, a new simplified modelling procedure for estimating the global response of existing RC buildings is presented. It is able to produce 3D reduced-order models (characterized by very few degrees of freedom) starting from the geometrical and mechanical features of the case study. The main advantage of the present approach is to account for the effects predictable with MDoF models, but with low analysis time and computational efforts, with elevate convergence capacity, typical of the SDoF models. The performance of this simplified numerical modelling procedure has been tested by the application on the previously mentioned sample of school buildings and comparing the results, in terms of structural response, damage states and confidence levels, with the ones previously obtained from the application of FSA. The relevance and perspective impact of the research work here presented should be seen in the wider field of the vulnerability analysis of the building stock at the regional scale, which is a crucial issue for the scientific community and for the civil society. Governments and administrations are invested with the difficult task of providing mitigation strategies for the seismic risk for a very wide and inhomogeneous portfolio of buildings and the economic resources are often very limited. Therefore, the development of methods for estimating the vulnerability with limited data has been a subject of intense research activity. The framework that is depicted in the dissertation can provide a tool potentially very impactful, since it could allow, by the exploitation of the 3D Reduced Order Models combined with FSA, to overcome the well-known limitations of empirical vulnerability approaches in favor of mechanical based methods managed in a full probabilistic framework.File | Dimensione | Formato | |
---|---|---|---|
PhD_thesis_SERGIO_RUGGIERI.pdf
accesso aperto
Descrizione: Tesi di dottorato completa di frontespizio
Tipologia:
Tesi di dottorato
Licenza:
Creative commons
Dimensione
79.39 MB
Formato
Adobe PDF
|
79.39 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.