This thesis concerns the study of an innovative moment resisting connection device to be adopted in the so-called moment resisting steel frames. Following previous seismic events, such as Northridge in 1994, many steel structures characterized by welded connections between beams and columns experienced extended damages and diffused brittle failures, making the facilities unserviceable and economically disadvantageous to repair. The answers to this problem is represented by a large number of studies, where several proposals for different devices, as well as special technical solutions, are available. However, the main idea developed among all the proposals is represented by the identification of a portion of the beam to be devoted to the onset of plastic strains, known in literature as Reduced Beam Section (RBS). In this way, this selected part is the only one of the beam to host the plastic strains, while the other portions behave elastically. The most representative example of this approach is the so-called dogbone proposed by Plumier. It reduces the strength of a specific cross-section by simply trimming off part of the flanges. This cut can follow different shapes, but the most common is the parabolic one. The dogbone certainly is effective, but on the other hand, it shows the side effect of increasing the flexural and torsional deformability of the element where it is realized, involving an additional percentage of interstorey drifts from 4% to 7%, as confirmed in international standard. The device studied in this work presents the possibility to independently fix the flexural stiffness and the strength, and it is called Limited Resistance Plastic Device (LRPD). The LRPD is made up of three different I-shaped parts: two outer portions and an inner one. The two outer portions are equal to each other and show a flange thickness greater than the inner one. Moreover, the device is characterized by symmetry with respect to three orthogonal barycentric planes, and the flanges of all portions possess a unique common medium plane. The web thickness is the same in all the portions, and it is assumed to be equal to the web thickness of the connected beam element. The described geometry is useful to guarantee the desired strength reduction, permitting the full development of the plastic strains within the inner portion, and at the same time it allows the device to exhibit independently the preassigned bending stiffness. The described features are obtained through the solution of an optimal design problem based on the minimization of a suitably chosen objective function while respecting appropriately assigned linear and nonlinear constraints. The chosen objective function is the volume of the device, while the constraints are related to the special device geometry, to the imposed resistance and stiffness, and, furthermore, to ensure that the device fully plasticizes avoiding undesired local buckling effects. The cited optimal design problem has been solved by means of a nonlinear programming tool such as the Matlab Optimization Toolbox. In addition, in this thesis, an approximate iterative procedure has been developed to obtain the geometric characteristics of LRPD without a specific numerical tool. The device was validated using various FEM simulations, with the goal of verifying the behaviour in relation to the length of the inner part to ensure the expected onset of the plastic deformations. The results confirmed that the LRPD behaves as expected and allowed to define a minimum length of the inner part in dependence of commercial profile (HE of IPE one). To check the affordability of the LRPD in real structures compared to its competitors, many examples of frames equipped with LRPD or dogbone subjected to static and dynamic loads prescribed by international standards were analyzed. The same frames without any device were analyzed under the same loading conditions, and the results were compared to those of the same frames equipped with dogbones and LRPD. It was observed that the presence of LRPD overcomes the greater deformability of the frames equipped with dogbones, allowing the interstorey drifts to be almost unaltered. Some efforts were devoted also to the study of the strength domains of the common I-shape cross sections, not only in terms of N-M interaction but also considering the shear contribution. Furthermore, the N-M domains were investigated taking into account the uncertainties that commonly characterize the device’s constitutive parameters, allowing to approximately quantify the effective limit domains. Another important topic focused on in the thesis was the realization of a breach in a masonry wall, and the consequent reinforcement of the panel by means of a steel frame. In this case, the goal is typically to recreate the element’s original stiffness while ignoring the variation in strength. It was demonstrated that by installing the LRPD in the reinforcement frame it is possible to adequately recreate the characteristics of the original panel. Finally, an experimental campaign was carried out. The obtained results showed the full validity of the LRPD, serving as a real benchmark of the reliability of the proposed model.
(2023). An innovative moment resisting steel connection: optimal design formulations, practical applications and experimental tests.
An innovative moment resisting steel connection: optimal design formulations, practical applications and experimental tests
VAZZANO, Santo
2023-02-01
Abstract
This thesis concerns the study of an innovative moment resisting connection device to be adopted in the so-called moment resisting steel frames. Following previous seismic events, such as Northridge in 1994, many steel structures characterized by welded connections between beams and columns experienced extended damages and diffused brittle failures, making the facilities unserviceable and economically disadvantageous to repair. The answers to this problem is represented by a large number of studies, where several proposals for different devices, as well as special technical solutions, are available. However, the main idea developed among all the proposals is represented by the identification of a portion of the beam to be devoted to the onset of plastic strains, known in literature as Reduced Beam Section (RBS). In this way, this selected part is the only one of the beam to host the plastic strains, while the other portions behave elastically. The most representative example of this approach is the so-called dogbone proposed by Plumier. It reduces the strength of a specific cross-section by simply trimming off part of the flanges. This cut can follow different shapes, but the most common is the parabolic one. The dogbone certainly is effective, but on the other hand, it shows the side effect of increasing the flexural and torsional deformability of the element where it is realized, involving an additional percentage of interstorey drifts from 4% to 7%, as confirmed in international standard. The device studied in this work presents the possibility to independently fix the flexural stiffness and the strength, and it is called Limited Resistance Plastic Device (LRPD). The LRPD is made up of three different I-shaped parts: two outer portions and an inner one. The two outer portions are equal to each other and show a flange thickness greater than the inner one. Moreover, the device is characterized by symmetry with respect to three orthogonal barycentric planes, and the flanges of all portions possess a unique common medium plane. The web thickness is the same in all the portions, and it is assumed to be equal to the web thickness of the connected beam element. The described geometry is useful to guarantee the desired strength reduction, permitting the full development of the plastic strains within the inner portion, and at the same time it allows the device to exhibit independently the preassigned bending stiffness. The described features are obtained through the solution of an optimal design problem based on the minimization of a suitably chosen objective function while respecting appropriately assigned linear and nonlinear constraints. The chosen objective function is the volume of the device, while the constraints are related to the special device geometry, to the imposed resistance and stiffness, and, furthermore, to ensure that the device fully plasticizes avoiding undesired local buckling effects. The cited optimal design problem has been solved by means of a nonlinear programming tool such as the Matlab Optimization Toolbox. In addition, in this thesis, an approximate iterative procedure has been developed to obtain the geometric characteristics of LRPD without a specific numerical tool. The device was validated using various FEM simulations, with the goal of verifying the behaviour in relation to the length of the inner part to ensure the expected onset of the plastic deformations. The results confirmed that the LRPD behaves as expected and allowed to define a minimum length of the inner part in dependence of commercial profile (HE of IPE one). To check the affordability of the LRPD in real structures compared to its competitors, many examples of frames equipped with LRPD or dogbone subjected to static and dynamic loads prescribed by international standards were analyzed. The same frames without any device were analyzed under the same loading conditions, and the results were compared to those of the same frames equipped with dogbones and LRPD. It was observed that the presence of LRPD overcomes the greater deformability of the frames equipped with dogbones, allowing the interstorey drifts to be almost unaltered. Some efforts were devoted also to the study of the strength domains of the common I-shape cross sections, not only in terms of N-M interaction but also considering the shear contribution. Furthermore, the N-M domains were investigated taking into account the uncertainties that commonly characterize the device’s constitutive parameters, allowing to approximately quantify the effective limit domains. Another important topic focused on in the thesis was the realization of a breach in a masonry wall, and the consequent reinforcement of the panel by means of a steel frame. In this case, the goal is typically to recreate the element’s original stiffness while ignoring the variation in strength. It was demonstrated that by installing the LRPD in the reinforcement frame it is possible to adequately recreate the characteristics of the original panel. Finally, an experimental campaign was carried out. The obtained results showed the full validity of the LRPD, serving as a real benchmark of the reliability of the proposed model.File | Dimensione | Formato | |
---|---|---|---|
Thesis - Vazzano.pdf
accesso aperto
Tipologia:
Tesi di dottorato
Dimensione
5.56 MB
Formato
Adobe PDF
|
5.56 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.