Risk management is a modern concept concerning the way to cope with natural and man-induced catastrophic events. Definitions of risk are reported and discussed and the particular relevance of Aeolian risk for a built environment is highlighted. Due to their strategic and neuralgic role, the importance of bridge structures as “elements at risk" is clear. It is also well known that fluid-structure interaction (aeroelasticity) can give rise to phenomena of major concern for the design of flexible bridges. In particular flutter can induce diverging oscillations and consequently bring to the collapse of the structure. This doctoral work focuses on the vulnerability assessment of flexible bridges with respect to flutter and two main contributions can be remarked. First, the Performance-Based Design approach is applied to the collapse limit state due to flutter, following the Pacific Earthquake Engineering Research Center (PEER) formulation. This risk-consistent design philosophy has been developed in the seismic engineering field and only recently some attempts have been made to adapt it to wind engineering applications but never taking into account aeroelastic phenomena. For the first time flutter derivatives are considered as random variables and the flutter problem is approached in a probabilistic way via Monte-Carlo simulations. A single-box girder deck is experimentally studied in the CRIACIV wind tunnel in order to make available data for this particular analysis. Interesting results are obtained both for this section model and for two rectangular cylinders. The second main contribution of this work is a sort of “pre-normative" study concerning flutter assessment, which could be useful to enhance the codes, as a measure of risk mitigation. The final goal is the set-up of a simplified method to estimate the flutter critical wind speed without performing wind-tunnel tests. Such a tool could be very useful for bridge engineers, especially concerning medium-span flexible bridges and/or pre-design stages. As a matter of fact, deep wind-tunnel investigations are expensive and time-consuming and, albeit absolutely necessary over all the design steps for long-span suspension bridges, they could be sometimes avoided or limited to the final validation stage for less important structures, for which aeroelastic phenomena are less concerning, even though they cannot be excluded a priori. In this context, the relationship between multimodal and bimodal approach to flutter is carefully analyzed and discussed, also with the support of two case studies, and then approximate formulas retaining only three aeroelastic functions are derived. This strong simpliflcation is validated on the basis of a wide range of structural and aerodynamic data, showing its extensive applicability. Finally, a relatively large number of flutter derivative sets are compared according to the definition of a few classes of deck cross-sectional geometry. These data include the trapezoidal single-box deck section with lateral cantilevers, whose experimental tests in the CRIACIV wind tunnel are described in details. The previously mentioned simplified formulas, reducing to three the flutter derivatives to be accounted for, make possible an attempt of generalization. Although this is only a first step towards this ambitious goal, it shows all the difficulties which have to be overcome but also highlights some interesting and promising results.

Flutter Vulnerability Assessment of Flexible Bridges – Wind-Tunnel Tests, Probabilistic Model, Analytical Investigation / Claudio Mannini. - STAMPA. - (2008), pp. 1-241.

### Flutter Vulnerability Assessment of Flexible Bridges – Wind-Tunnel Tests, Probabilistic Model, Analytical Investigation

#####
*MANNINI, CLAUDIO*

##### 2008

#### Abstract

Risk management is a modern concept concerning the way to cope with natural and man-induced catastrophic events. Definitions of risk are reported and discussed and the particular relevance of Aeolian risk for a built environment is highlighted. Due to their strategic and neuralgic role, the importance of bridge structures as “elements at risk" is clear. It is also well known that fluid-structure interaction (aeroelasticity) can give rise to phenomena of major concern for the design of flexible bridges. In particular flutter can induce diverging oscillations and consequently bring to the collapse of the structure. This doctoral work focuses on the vulnerability assessment of flexible bridges with respect to flutter and two main contributions can be remarked. First, the Performance-Based Design approach is applied to the collapse limit state due to flutter, following the Pacific Earthquake Engineering Research Center (PEER) formulation. This risk-consistent design philosophy has been developed in the seismic engineering field and only recently some attempts have been made to adapt it to wind engineering applications but never taking into account aeroelastic phenomena. For the first time flutter derivatives are considered as random variables and the flutter problem is approached in a probabilistic way via Monte-Carlo simulations. A single-box girder deck is experimentally studied in the CRIACIV wind tunnel in order to make available data for this particular analysis. Interesting results are obtained both for this section model and for two rectangular cylinders. The second main contribution of this work is a sort of “pre-normative" study concerning flutter assessment, which could be useful to enhance the codes, as a measure of risk mitigation. The final goal is the set-up of a simplified method to estimate the flutter critical wind speed without performing wind-tunnel tests. Such a tool could be very useful for bridge engineers, especially concerning medium-span flexible bridges and/or pre-design stages. As a matter of fact, deep wind-tunnel investigations are expensive and time-consuming and, albeit absolutely necessary over all the design steps for long-span suspension bridges, they could be sometimes avoided or limited to the final validation stage for less important structures, for which aeroelastic phenomena are less concerning, even though they cannot be excluded a priori. In this context, the relationship between multimodal and bimodal approach to flutter is carefully analyzed and discussed, also with the support of two case studies, and then approximate formulas retaining only three aeroelastic functions are derived. This strong simpliflcation is validated on the basis of a wide range of structural and aerodynamic data, showing its extensive applicability. Finally, a relatively large number of flutter derivative sets are compared according to the definition of a few classes of deck cross-sectional geometry. These data include the trapezoidal single-box deck section with lateral cantilevers, whose experimental tests in the CRIACIV wind tunnel are described in details. The previously mentioned simplified formulas, reducing to three the flutter derivatives to be accounted for, make possible an attempt of generalization. Although this is only a first step towards this ambitious goal, it shows all the difficulties which have to be overcome but also highlights some interesting and promising results.I documenti in FLORE sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.