Flat slabs represent one of the most common structural solutions for reinforced concrete building floors. Several experimental studies have been carried out with the aim to clarify flat slab seismic design rules. The state of the art concerns numerous isolated slab column connections, fewer single flat slab floors, and even fewer multi-storey flat slab frames. All of these tests – but one – were carried out on specimens to scale. No experimental campaign was carried out for building structure with primary walls and secondary frames. The main reason for this lack of tests on a real scale building structure lies is the limitation imposed by the dimensions; in fact, it is frequently impossible to build and to test such a system inside a laboratory. The related costs are another important factor, limiting this type of research. The SlabSTRESS project was developed to satisfy this need, and was proposed by four European universities (Polimi, Milano, EPFL Lausanne, UNOVA Lisbon and UTCB Bucharest) and carried out in collaboration with the European Commission’s Joint Research Centre (JRC) in Ispra (VA, Italy). The SlabSTRESS experimental campaign was performed at the ELSA (European Laboratory for Structural Assessment), Reaction Wall at the JRC in Ispra. The project was part of the transnational access activities of the SERA project. Forming a part of an extensive research project, this thesis proposes and applies a methodology for the analysis of measurement data to characterise and put into relation the global and local structural response. Moreover, the state of the art is analysed, and compared to the SlabSTRESS results, therefore to understand what part of the existing knowledge is confirmed and illustrate the innovative findings. The real-scale mock-up consisted of a reinforced concrete building with two ordinary thick slabs, supported by twelve columns on each floor, one of which was provided with transverse reinforcement. Two different longitudinal reinforcement layouts were used. The building was designed to consider the flat slabs as secondary members, with seismic resistant walls as primary ductile elements. The structure was designed to bear gravity loads and capable of reaching a lateral relative inter-storey drift compatible with moderate-high seismic lateral action. This design procedure was inspired by the method used in the North American codes for the seismic design of flat slabs. The primary seismic walls were not actually built in the structure but were virtual elements, simulated by the sub-structuring allowed by the pseudo-dynamic method used in the host lab. A specific design was made for the RC columns, with a structural steel stub at the centre of these members. The latter were required for measuring internal forces and bending moment in each slab-column connection. The first phase of the test programme simulated two levels of seismic activity: the service limit state (SLS) and the ultimate limit state (ULS). Seismic loading was applied using the pseudo-dynamic technique with linear sub-structuring. The seismic response of large-size specimens can be obtained by the combination of the experimental restoring forces with the analytical inertia and seismic-equivalent forces. Successively, during the second phase two cyclic tests were carried out to test the floors for a combination of gravity and lateral cyclic loading of increasing amplitude, up to near-failure conditions. Prior to the second cyclic test some of the damaged connections were strengthened and then were tested again. Both the global and the local measurements were acquired. Each column on both floors was provided with an ad hoc system designed to allow the internal forces and bending moment to be measured. A large number of sensors were used in this experimental campaign: 80 inclinometers, 48 extensometers and 192 strain-gauges, which permitted a series of data to be collected. All the sensors had been previously calibrated to ensure the precision of the measurement system. A new controller and data acquisition system, called ElsaREC, was developed by the ELSA laboratory team and used here for the first time, permitting a systematic calibration procedure for all the sensors adopted in the experiment at ELSA. Using the main statics and kinematics parameters, two schemes were chosen to represent the structure from both a global and a local point of view. Two kinematics parameters, one displacement for each floor and two static parameters, one force for each floor, described the global system. These parameters were acquired by the load cells applied to the actuators, while displacement transducers applied on two reference frames near the structure recorded the corresponding displacements. The local system to capture the behaviour of each slab-column connection was more complex. The slab and the column rotations represented the local kinematic parameters that were recorded by the inclinometers, and the slab cracks opening was recorded by the displacement transducers. Local kinematic measurements were acquired in the surrounding area of the slab on eight out of the twelve columns. The three connection (internal, edge and corner) were analysed with four, three and two instrumented sides respectively. Each side of the slab-column connection was equipped with an inclinometer to measure slab rotation along the East-West direction and a set of displacement transducers to measure crack opening through an increase of the slab thickness. Using the acquired slab rotation data and the column of each slab-column connection, it was possible to derive the relative rotation of the slab with respect to the column. Bending moment and shear were acquired from load cells located at the centre of each column on each floor, specially designed for this experimental campaign. The unbalanced moment for each connection was then calculated. The recorded signals were processed in order to characterise both the global behaviour and the local behaviour of slab-column connections. All the global analyses were performed using the inter-storey drift ratio as main coordinates to represent the behaviour of the structure. At local level the inter-storey drift ratio was correlated to local measurements, columns and slabs rotations. Lastly described, the procedure followed to obtain conversion factors that allow the inter-storey drift ratio to be converted to the column-slab rotation as a function of the slab-column connections typology is described. The global and local response are analysed using force deformation relations, stiffness deteriorations diagrams, energy dissipation and ductility analysis. The local response is further investigated using the measurements of slab thickness variations, related to cracking; together with the crack patterns, these were used to study the different failure modes. The results obtained were then compared to the state of the art. Some results confirm findings obtained from individual connection and scaled frame tests. Other aspects are innovative. The tests showed global deformation capacity of the structure up to 2.5% global drift ratio with the 1st floor without shear reinforcement and 6% with some strengthening. This confirms global tests of scaled floor and frame specimens reaching deformation capacities in the range 3-6%. The connection response confirms the deformation capacity results of the data bases collected in the literature, for specimens without and with shear reinforcement. failure modes similar to those shown in the literature were observed. The innovative results are several. The response of a building structure with a flat slab frame and primary seismic resistant walls was verified experimentally for the design SLS and ULS seismic action. very limited damage was inflicted to the flat slab frame, because of the limited drift ratio allowed by the primary seismic system. A full-scale flat slab frame was tested for the first time up to global near failure conditions (i.e. 16% lateral load bearing resistance loss). This showed that the structure could bear the gravity loads at the conclusion of the test without progressive collapse, and highlighted the efficiency of single integrity reinforcement details. The local response in a multi-storey frame was measured with an innovative measurement system, the connection response is thus studied under the real boundary conditions and the internal force redistribution in the whole structure. The response was measured for different types of connections, i.e. interior, edge and corner, the data on the latter two not being very numerous in the literature. The efficiency of a strengthening system for flat slab connections in a floor was shown, with redistribution of internal forces and moments. The information collected in this study represents the basis for further understanding of flat slab response to combined gravity and lateral loading, the calibration of a model can take advantage of the measurements carried out on a complete full scale structure. This experimental work and its developments path the way to new and more advantage design rules, in particular for the European Codes.
I solai a piastra rappresentano una delle soluzioni strutturali più comuni per la costruzione di edifici in cemento armato. Con lo scopo di chiarire le regole di progettazione sismica di questa soluzione costruttiva, in letteratura è possibile trovare numerosi studi riguardanti le singole connessioni piastra-pilastro, un numero minore riguardanti singoli solai e ancora meno per solai a più piani. Tutti i test appena citati, ad eccezione di uno, sono stati eseguiti su campioni in scala ridotta e inoltre è stata condotta alcuna campagna sperimentale per strutture con muri primari e telai secondari. La ragione principale di questa mancanza di prove su una struttura in scala reale è il limite imposto dalle dimensioni; infatti, spesso è impossibile costruire e provare un’intera struttura all'interno di un laboratorio. Un altro fattore limitante per questo tipo di ricerca, sono i costi elevati. Per rispondere a queste esigenze, quattro università europee (Polimi Milano, EPFL Lausanne, UNOVA Lisbon e UTCB Bucharest) in collaborazione con il Joint Research Centre (JRC) della Commissione Europea a Ispra (VA, Italia), hanno sviluppato il progetto SlabSTRESS. La campagna sperimentale SlabSTRESS è stata condotta ad ELSA (European Laboratory for Structural Assessment), laboratorio del JRC ad Ispra. Il progetto faceva parte delle attività di accesso transnazionale del progetto SERA. Questa tesi, parte dell’ampio progetto di ricerca SlabSTRESS, propone e applica una metodologia di analisi dei dati di misura per caratterizzare e mettere in relazione la risposta strutturale globale e locale. Inoltre, un’attenta analisi dello stato dell'arte ha permesso il confronto con i dati di SlabSTRESS, confermando alcuni dei risultati già noti e illustrando quelli innovativi. Il provino è costituito da un edificio in cemento armato in scala reale con due solette ordinarie di grosso spessore, sorrette da dodici colonne per piano, di cui una dotata di rinforzo trasversale. Sono stati utilizzati due diversi layout di rinforzo longitudinale. L'edificio è stato progettato per considerare i solai piani come elementi secondari, con le pareti antisismiche come elementi primari duttili. Inoltre, la struttura è stata progettata per sopportare carichi gravitazionali e in grado di raggiungere una deriva laterale relativa di interpiano compatibile con un'azione sismica laterale alta o medio-alta. La progettazione è stata basata sul metodo utilizzato nei codici nordamericani per la progettazione sismica di solai a piastre. I muri sismici primari non sono stati realizzati realmente nella struttura ma erano elementi virtuali, simulati grazie al metodo pseudo-dinamico utilizzato nel laboratorio ELSA (test pseudo-dinamico con sottostrutturazione). Un progetto specifico è stato realizzato per le colonne, con un componente in acciaio strutturale al centro di questi elementi. Questi ultimi erano necessari per misurare le forze interne e il momento flettente in ciascuna connessione lastra-pilastro. La prima fase dei test ha simulato due livelli di attività sismica: lo stato limite di servizio (SLS) e lo stato limite ultimo (ULS). Il carico sismico è stato applicato utilizzando la tecnica pseudo-dinamica con sottostrutture lineari. Successivamente, durante la seconda fase, sono state eseguite due prove cicliche per testare i solai con una determinata combinazione di carichi gravitazionali e con carico ciclico laterale di ampiezza crescente, fino a condizioni prossime al cedimento. Prima della seconda prova ciclica, alcune delle connessioni danneggiate sono state rinforzate e poi sono state nuovamente testate. Sono state acquisite sia misurazioni globali che locali. Come già accennato, ogni colonna (su entrambi i piani) è stata dotata di un sistema ad hoc progettato per consentire la misurazione delle forze interne e del momento flettente. In questa campagna sperimentale sono stati utilizzati un gran numero di sensori: 80 inclinometri, 48 sensori di spostamento e 192 estensimetri, che hanno consentito di raccogliere un numero considerevole dati. Tutti i sensori sono stati preventivamente calibrati per garantire la precisione e accuratezza del sistema di misura. Un nuovo controllore e sistema di acquisizione dati, chiamato ElsaREC, è stato sviluppato dal team del laboratorio ELSA e utilizzato qui per la prima volta, consentendo una procedura di calibrazione sistematica per tutti i sensori adottati nell'esperimento. Utilizzando i principali parametri statici e cinematici, sono stati scelti due schemi per rappresentare la struttura da un punto di vista globale e locale. il sistema globale è descritto da due parametri cinematici, uno spostamento per ogni piano e due parametri statici, una forza per ogni piano. Questi ultimi parametri sono stati acquisiti dalle celle di carico applicate agli attuatori, mentre i corrispondenti spostamenti sono stati acquisiti tramite trasduttori di spostamento (applicati su due sistemi di riferimento vicini alla struttura. Il sistema locale per catturare il comportamento di ciascuna connessione solaio-pilastro e’ più complesso. Le rotazioni della soletta e della colonna rappresentano i parametri cinematici locali acquisiti tramite inclinometri, e l'apertura delle fessure della soletta registrata da trasduttori di spostamento. Le misure cinematiche locali sono state acquisite nell'area circostante la piastra su otto delle dodici colonne. I tre collegamenti (interno, bordo e angolo) sono stati analizzati rispettivamente con quattro, tre e due lati strumentati. Su ciascun lato della connessione lastra-colonna è stato inserito un inclinometro per misurare la rotazione del solaio lungo la direzione Est-Ovest e una serie di trasduttori di spostamento per misurare l'apertura della fessura attraverso la misura dello spessore del solaio. Utilizzando i dati acquisiti di rotazione del solaio e del pilastro di ogni collegamento solaio-pilastro, è stato possibile ricavare la rotazione relativa del solaio rispetto al pilastro. I momenti flettenti e i tagli sono stati acquisiti dalle celle di carico situate al centro di ogni colonna su ogni piano. È stato quindi calcolato il momento sbilanciato per ciascuna connessione. I segnali registrati sono stati elaborati al fine di caratterizzare sia il comportamento globale che il comportamento locale dei collegamenti solaio-pilastro. Tutte le analisi globali sono state eseguite utilizzando il rapporto di deriva di interpiano come coordinate principali per rappresentare il comportamento della struttura. A livello locale il rapporto di deriva di interpiano è stato correlato alle misurazioni locali, rotazioni di colonne e piastre. Infine si è descritta la procedura seguita per ottenere i fattori di conversione che consentono di passare dal rapporto di deriva di interpiano alla rotazione pilastro-solaio in funzione della tipologia dei collegamenti solaio-pilastro. La risposta globale e locale è stata analizzata utilizzando la perdita della rigidezza, la dissipazione di energia e l’analisi della duttilità. Le diverse modalità di rottura sono state studiate grazie all’analisi delle fotografie eseguite durante e alla fine di ogni test. I risultati ottenuti sono stati confrontati con lo stato dell'arte. Alcuni risultati hanno confermato i dati ottenuti dai test su singole connessioni e solai in scala, molti altri risultati invece rappresentano un’innovazione. I test hanno mostrato una capacità di deformazione globale della struttura fino al 2,5% di rapporto di deriva globale con il 1° piano senza rinforzo a taglio e 6% con il rinforzo di alcune connessioni. Ciò conferma i dati ottenuti da test globali su provini in scala ridotta su singoli solai o telai che raggiungono capacità di deformazione comprese tra il 3 e il 6%. La risposta delle connessioni conferma i risultati relativi alla capacità di deformazione presenti in letteratura, per provini senza e con armatura a taglio. Sono state osservate modalità di danno simili a quelle mostrate in letteratura. Per le azioni sismiche di progetto SLS e ULS, a causa del limitato rapporto di deriva consentito dal sistema sismico primario, si sono osservati danni molto limitati. Per la prima volta è stato testato un telaio a piastre piane a grandezza reale fino alle condizioni globali di quasi cedimento (ovvero pari ad una perdita di resistenza del carico laterale del 16%). Ciò ha dimostrato che la struttura ha sopportato i carichi gravitazionali fino al termine della prova senza cedimenti progressivi ed ha evidenziato l'efficienza dei sistemi di rinforzo. Grazie alla misura della risposta locale, si è potuto studiare la risposta delle connessioni nelle reali condizioni considerando anche la redistribuzione interna delle forze nell’intera struttura. La presenza di diversi tipi di connessioni (interne, laterali e d’angolo), ha consentito di raccogliere dati su tipologie di connessioni poco presenti in letteratura. Questa tesi rappresenta la base per permettere di comprendere meglio la risposta di edifici a piastre soggetti al carico gravitazionale e laterale ciclico. Questo lavoro sperimentale ed i suoi sviluppi aprono la strada a nuove e più vantaggiose regole di progettazione, in particolare per gli Eurocodici.
Seismic response and strengthening of flat slab buildings
Netti, Teresa
2020/2021
Abstract
Flat slabs represent one of the most common structural solutions for reinforced concrete building floors. Several experimental studies have been carried out with the aim to clarify flat slab seismic design rules. The state of the art concerns numerous isolated slab column connections, fewer single flat slab floors, and even fewer multi-storey flat slab frames. All of these tests – but one – were carried out on specimens to scale. No experimental campaign was carried out for building structure with primary walls and secondary frames. The main reason for this lack of tests on a real scale building structure lies is the limitation imposed by the dimensions; in fact, it is frequently impossible to build and to test such a system inside a laboratory. The related costs are another important factor, limiting this type of research. The SlabSTRESS project was developed to satisfy this need, and was proposed by four European universities (Polimi, Milano, EPFL Lausanne, UNOVA Lisbon and UTCB Bucharest) and carried out in collaboration with the European Commission’s Joint Research Centre (JRC) in Ispra (VA, Italy). The SlabSTRESS experimental campaign was performed at the ELSA (European Laboratory for Structural Assessment), Reaction Wall at the JRC in Ispra. The project was part of the transnational access activities of the SERA project. Forming a part of an extensive research project, this thesis proposes and applies a methodology for the analysis of measurement data to characterise and put into relation the global and local structural response. Moreover, the state of the art is analysed, and compared to the SlabSTRESS results, therefore to understand what part of the existing knowledge is confirmed and illustrate the innovative findings. The real-scale mock-up consisted of a reinforced concrete building with two ordinary thick slabs, supported by twelve columns on each floor, one of which was provided with transverse reinforcement. Two different longitudinal reinforcement layouts were used. The building was designed to consider the flat slabs as secondary members, with seismic resistant walls as primary ductile elements. The structure was designed to bear gravity loads and capable of reaching a lateral relative inter-storey drift compatible with moderate-high seismic lateral action. This design procedure was inspired by the method used in the North American codes for the seismic design of flat slabs. The primary seismic walls were not actually built in the structure but were virtual elements, simulated by the sub-structuring allowed by the pseudo-dynamic method used in the host lab. A specific design was made for the RC columns, with a structural steel stub at the centre of these members. The latter were required for measuring internal forces and bending moment in each slab-column connection. The first phase of the test programme simulated two levels of seismic activity: the service limit state (SLS) and the ultimate limit state (ULS). Seismic loading was applied using the pseudo-dynamic technique with linear sub-structuring. The seismic response of large-size specimens can be obtained by the combination of the experimental restoring forces with the analytical inertia and seismic-equivalent forces. Successively, during the second phase two cyclic tests were carried out to test the floors for a combination of gravity and lateral cyclic loading of increasing amplitude, up to near-failure conditions. Prior to the second cyclic test some of the damaged connections were strengthened and then were tested again. Both the global and the local measurements were acquired. Each column on both floors was provided with an ad hoc system designed to allow the internal forces and bending moment to be measured. A large number of sensors were used in this experimental campaign: 80 inclinometers, 48 extensometers and 192 strain-gauges, which permitted a series of data to be collected. All the sensors had been previously calibrated to ensure the precision of the measurement system. A new controller and data acquisition system, called ElsaREC, was developed by the ELSA laboratory team and used here for the first time, permitting a systematic calibration procedure for all the sensors adopted in the experiment at ELSA. Using the main statics and kinematics parameters, two schemes were chosen to represent the structure from both a global and a local point of view. Two kinematics parameters, one displacement for each floor and two static parameters, one force for each floor, described the global system. These parameters were acquired by the load cells applied to the actuators, while displacement transducers applied on two reference frames near the structure recorded the corresponding displacements. The local system to capture the behaviour of each slab-column connection was more complex. The slab and the column rotations represented the local kinematic parameters that were recorded by the inclinometers, and the slab cracks opening was recorded by the displacement transducers. Local kinematic measurements were acquired in the surrounding area of the slab on eight out of the twelve columns. The three connection (internal, edge and corner) were analysed with four, three and two instrumented sides respectively. Each side of the slab-column connection was equipped with an inclinometer to measure slab rotation along the East-West direction and a set of displacement transducers to measure crack opening through an increase of the slab thickness. Using the acquired slab rotation data and the column of each slab-column connection, it was possible to derive the relative rotation of the slab with respect to the column. Bending moment and shear were acquired from load cells located at the centre of each column on each floor, specially designed for this experimental campaign. The unbalanced moment for each connection was then calculated. The recorded signals were processed in order to characterise both the global behaviour and the local behaviour of slab-column connections. All the global analyses were performed using the inter-storey drift ratio as main coordinates to represent the behaviour of the structure. At local level the inter-storey drift ratio was correlated to local measurements, columns and slabs rotations. Lastly described, the procedure followed to obtain conversion factors that allow the inter-storey drift ratio to be converted to the column-slab rotation as a function of the slab-column connections typology is described. The global and local response are analysed using force deformation relations, stiffness deteriorations diagrams, energy dissipation and ductility analysis. The local response is further investigated using the measurements of slab thickness variations, related to cracking; together with the crack patterns, these were used to study the different failure modes. The results obtained were then compared to the state of the art. Some results confirm findings obtained from individual connection and scaled frame tests. Other aspects are innovative. The tests showed global deformation capacity of the structure up to 2.5% global drift ratio with the 1st floor without shear reinforcement and 6% with some strengthening. This confirms global tests of scaled floor and frame specimens reaching deformation capacities in the range 3-6%. The connection response confirms the deformation capacity results of the data bases collected in the literature, for specimens without and with shear reinforcement. failure modes similar to those shown in the literature were observed. The innovative results are several. The response of a building structure with a flat slab frame and primary seismic resistant walls was verified experimentally for the design SLS and ULS seismic action. very limited damage was inflicted to the flat slab frame, because of the limited drift ratio allowed by the primary seismic system. A full-scale flat slab frame was tested for the first time up to global near failure conditions (i.e. 16% lateral load bearing resistance loss). This showed that the structure could bear the gravity loads at the conclusion of the test without progressive collapse, and highlighted the efficiency of single integrity reinforcement details. The local response in a multi-storey frame was measured with an innovative measurement system, the connection response is thus studied under the real boundary conditions and the internal force redistribution in the whole structure. The response was measured for different types of connections, i.e. interior, edge and corner, the data on the latter two not being very numerous in the literature. The efficiency of a strengthening system for flat slab connections in a floor was shown, with redistribution of internal forces and moments. The information collected in this study represents the basis for further understanding of flat slab response to combined gravity and lateral loading, the calibration of a model can take advantage of the measurements carried out on a complete full scale structure. This experimental work and its developments path the way to new and more advantage design rules, in particular for the European Codes.| File | Dimensione | Formato | |
|---|---|---|---|
|
TeresaNetti.pdf
accessibile in internet per tutti
Dimensione
67.91 MB
Formato
Adobe PDF
|
67.91 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/174003