Large roof steel structures are usually stiff enough to limit aeroelastic phenomena but flexible enough to be sensitive to the wind buffeting actions, so that moderate and frequent wind actions can lead to accumulation of fatigue damage and potential in-service structural failure. The literature does not present any case study on the fatigue assessment of these structures, and the major codes and standards lack of guidance on the problem. In that framework, the research presents a procedure for assessing the wind-induced fatigue of large roof steel structures and the application of the same for the fatigue assessment of the new AEK Stadium in Athens. The procedure involves the modelling of the wind climate conditions, the simulation of the time and space variant pressure fields, the computation of local stress cycles based on the aerodynamic response, and the estimation of the fatigue damage. The probabilistic study of the wind climate conditions allows the definition of the number of hours of occurrence of each mean wind speed and direction. A critical observation of the results enables the designer to identify a priori a limited number of “fatigue-relevant directions”. The time and space variant pressure fields is simulated based on the results of Wind Tunnel Tests on rigid models. The hypothesis of Reynolds number independence allows the time and pressure scaling. It has been demonstrated that the scaling is not required to be performed for each wind speed since it always exists a so-called “quasi-static wind speed”, below which no dynamic amplification occurs. Its value, compared to the design value, gives a measure of the fatigue sensitivity of a structure since, at least for the case study, the absence of dynamic amplification is traduced into an absence of fatigue damage. Because the quasi-static wind speed depends on the high frequency components of the fluctuating pressures, its accurate evaluation introduces a new requirement on the velocity scale adopted in the WTTs. The complexity of the structural joints of these structures requires that the link between structural response and local stress is obtained with the combined use of global and local finite element models, made possible by additional “interface nodes”. Then, the local stress time histories are obtained by combining the displacements time histories at the interface nodes with a so-called “influence vector”, namely the stress for unit displacements at the virtual strain-gauge position. This approach allows the use of local FE models merely for the computation of the influence vector, thus allowing more refined mesh with an increased accuracy. Finally, the stress time histories are processed with a Rainflow algorithm and the total damage computed following the Miner’s cumulative rule. One of the major problems related to the fatigue assessment of large roof steel structures concerns the need of massive calculations since, potentially, all the details of a structure should be checked. Therefore, a dedicated software has been created which performs almost automatically the fatigue assessment for the whole set of details of a structure, only requiring the wind climate measurements and the pressure coefficients recorded during the WTTs. Finally, the automation of the procedure has allowed to perform a comprehensive sensitivity analysis on the main influencing parameters.
La coperture in acciaio di grande luce, sebbene sufficientemente rigide da limitare gli effetti aeroelastici, sono molto sensibili alle azioni dinamiche del vento, e di conseguenza, venti moderati e frequenti possono generare un accumulo di danno a fatica e potenziali collassi in condizioni di esercizio. La letteratura scientifica non presenta alcun caso studio che riguardi la verifica a fatica indotta da vento per le grandi coperture in acciaio, e le principali normative peccano di linee guida per affrontare il problema. In questo contesto, la ricerca propone una procedura per verificare la fatica indotta da vento per grandi coperture in acciaio ed una applicazione al nuovo stadio AEK di Atene. La procedura coinvolge la modellazione delle condizioni climatiche, la simulazione del carico del vento, il calcolo di una tensione locale a partire dalla risposta aerodinamica, ed infine, il calcolo del danno a fatica. Lo studio probabilistico delle condizioni climatiche porta alla definizione del numero di ore di occorrenza di ciascuna velocità e direzione del vento. Un’analisi critica dei risultati permette l’identificazione, a priori, di un numero limitato di “direzioni rilevanti a fatica”. La simulazione del carico del vento viene svolta mediante l’uso di prove in galleria del vento su modelli rigidi. L’ipotesi di indipendenza dal numero di Reynolds permette la scalatura delle pressioni e del tempo. È stato dimostrato che non serve simulare le azioni del vento per tutte le velocità in quanto esiste una “velocità quasi-statica” al di sotto del quale non vi è più l’amplificazione dinamica. Il suo valore, paragonato a quello massimo di progetto, fornisce una misura della sensibilità a fatica di una struttura. Il caso studio ha mostrato, infatti, che l’assenza di amplificazione si traduce in un’assenza di danno a fatica. Poiché tale velocità dipende dalle componenti di fluttuazione delle pressioni ad alta frequenza, la sua valutazione accurata ha messo in luce un nuovo requisito sulla scelta della scala di velocità in galleria del vento. Data la complessità delle connessioni di grandi coperture in acciaio, il legame tra la risposta aerodinamica e le tensioni locali è stato ottenuto combinando il modello globale della struttura con quelli locali dei collegamenti mediante una serie di “nodi di interfaccia”. Ciò ha permesso di ottenere la storia temporale di una tensione in un punto sulla base degli spostamenti dei nodi di interfaccia e di un “vettore di influenza”, il quale, rappresenta la tensione per unità di spostamento alla posizione dello strain-gauge virtuale. Poiché l’approccio prevede che i modelli locali siano utilizzati esclusivamente per il calcolo del vettore di influenza, la mesh può essere infittita a vantaggio di una maggiore accuratezza dei risultati. La storia temporale della tensione viene poi processata con il metodo di Rainflow e, infine, il calcolo a fatica viene svolto utilizzando la legge di cumulo di Miner. Una delle maggiori problematiche legate alla verifica a fatica di grandi coperture in acciaio è il calcolo massivo legato all’elevato numero di dettagli di queste strutture. Questo ha spinto alla creazione di un software dedicato il quale permette di svolgere in maniera pressoché automatica la verifica a fatica di grande coperture in acciaio, solo sulla base dei dati climatici del vento e delle pressioni ricavate dalle prove in galleria del vento. Infine, l’automazione della procedura ha permesso di svolgere un’approfondita analisi di sensibilità riguardante i principali aspetti critici del fenomeno.
A procedure for assessing the wind-induced fatigue for large roof steel structures: application to the case of the new AEK Stadium in Athens
ROSSI, ROBERTO
2017/2018
Abstract
Large roof steel structures are usually stiff enough to limit aeroelastic phenomena but flexible enough to be sensitive to the wind buffeting actions, so that moderate and frequent wind actions can lead to accumulation of fatigue damage and potential in-service structural failure. The literature does not present any case study on the fatigue assessment of these structures, and the major codes and standards lack of guidance on the problem. In that framework, the research presents a procedure for assessing the wind-induced fatigue of large roof steel structures and the application of the same for the fatigue assessment of the new AEK Stadium in Athens. The procedure involves the modelling of the wind climate conditions, the simulation of the time and space variant pressure fields, the computation of local stress cycles based on the aerodynamic response, and the estimation of the fatigue damage. The probabilistic study of the wind climate conditions allows the definition of the number of hours of occurrence of each mean wind speed and direction. A critical observation of the results enables the designer to identify a priori a limited number of “fatigue-relevant directions”. The time and space variant pressure fields is simulated based on the results of Wind Tunnel Tests on rigid models. The hypothesis of Reynolds number independence allows the time and pressure scaling. It has been demonstrated that the scaling is not required to be performed for each wind speed since it always exists a so-called “quasi-static wind speed”, below which no dynamic amplification occurs. Its value, compared to the design value, gives a measure of the fatigue sensitivity of a structure since, at least for the case study, the absence of dynamic amplification is traduced into an absence of fatigue damage. Because the quasi-static wind speed depends on the high frequency components of the fluctuating pressures, its accurate evaluation introduces a new requirement on the velocity scale adopted in the WTTs. The complexity of the structural joints of these structures requires that the link between structural response and local stress is obtained with the combined use of global and local finite element models, made possible by additional “interface nodes”. Then, the local stress time histories are obtained by combining the displacements time histories at the interface nodes with a so-called “influence vector”, namely the stress for unit displacements at the virtual strain-gauge position. This approach allows the use of local FE models merely for the computation of the influence vector, thus allowing more refined mesh with an increased accuracy. Finally, the stress time histories are processed with a Rainflow algorithm and the total damage computed following the Miner’s cumulative rule. One of the major problems related to the fatigue assessment of large roof steel structures concerns the need of massive calculations since, potentially, all the details of a structure should be checked. Therefore, a dedicated software has been created which performs almost automatically the fatigue assessment for the whole set of details of a structure, only requiring the wind climate measurements and the pressure coefficients recorded during the WTTs. Finally, the automation of the procedure has allowed to perform a comprehensive sensitivity analysis on the main influencing parameters.| File | Dimensione | Formato | |
|---|---|---|---|
|
A procedure for assessing the wind-induced fatigue for large roof structures_application to the case of the new AEK stadium in Athens_2.pdf
non accessibile
Descrizione: A procedure for assessing the wind-induced fatigue for large roof steel structures: application to the case of the new AEK stadium in Athens
Dimensione
15.93 MB
Formato
Adobe PDF
|
15.93 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/10589/146437