CARACTERIZACION DE LAS MATERIAS PRIMAS

3.4.1 Experimental setup of PT connection

A steel PT connection is developed and experimentally validated in (Kim and Christopoulos 2008). The connection incorporates bolt-stressed FEDs consisting of stainless steel interfaces and new non asbestos (NAO) break lining pads as the energy dissipating mechanism. Figure 2.27 shows details of the FEDs. PT high strength steel bars are running parallel to the beam to provide self-centering capability to the system. Figure 3.17 shows a photo of the experimental setup in (Kim and Christopoulos 2008). Results from cyclic tests on the PT connection are presented in (Kim and Christopoulos 2008). The tests have been developed up to high drifts, where local buckling on the beam has been occurred.

The column and beams were W360x509 and W610x113 sections, respectively. Both members were 350W steel with nominal yield strength of 350MPa. Steel plates with 12mm thickness and nominal yield strength of 550MPa, were welded to the beam flanges and ended at 835mm from the column flange. Continuity plates, 20mm thick were welded to the inside of the column flanges and web to provide appropriate force flow from the beam to the column. Contact plates, 25mm thick (Grade 550) were placed between the beam and the column flange. Longitudinal stiffeners along the beam web have been inserted to prevent local buckling on the web. Roller supports have been inserted along the beams. Six 60-mm-diameter holes were introduced in the column flanges to accommodate the PT bars and the vertical movements expected during testing. Six 32-mm-diameter Dywidag high-strength bars with a nominal ultimate strength of 1,030 MPa were selected to provide the post-tensioning. Figure 3.18 shows a drawing depicting the experimental setup and Figure 2.19 shows the details of the PT connection. The initial tension of each PT bar was 200kN, which corresponded to 24% of their ultimate strength. The total friction developed by each FED was about 280kN. The loading protocol of AISC (2005) was applied on the specimen and the results are presented in (Kim and Christopoulos 2008).

102

Figure 3.17 Photo from the experimental setup (Kim and Christopoulos 2008)

Figure 3.18 Experimental setup developed (Kim and Christopoulos 2008)

103

3.4.2 FEM in Abaqus of PT connection

FEM models have been developed in Abaqus, simulating the aforementioned experimental setup. Figure 3.19 shows the model in Abaqus for the PT connection. The column, the longitudinal stiffeners, the anchor block the continuity plates and the beam apart from the areas where local buckling phenomena are expected to occur are modeled using C3D8R solid elements. The part of the beams after beam reinforcing plates, where local buckling is expected to occur (300mm along the beam axis) are modeled using solid elements with incompatible modes C3D8I and with a refined mesh. There are 16 elements along the length of the beam web, and 4 elements along the flanges length. There is one element along the thickness of the web and 2 elements along the thickness of the flanges and the reinforcing plates. The approximate global size of the beams elements is 75mm. The approximate global size of the washers elements is 15mm. The approximate global size of the column‟s elements is 50mm, while there are 12 elements along the perimeter of the holes. Anchor blocks have elements of 25mm global size. The web reinforcing plates have elements of 30mm global size. Figure 3.20 shows the mesh of the model for the PT connection.

Figure 3.19 FEM of the PT connection developed in Kim and Christopoulos (2008)

104

Figure 3.20 Mesh of the model for the PT connection developed in Kim and Christopoulos (2008)

An elasto-plastic law with isotropic hardening rule was specified for the steel material. PT bars were modeled using truss elements. The post tensioning was modeled using bolt load on the PT bars and by applying a certain shortening (Adjust length) on the truss elements in the first step of the analysis. The start and end points of the PT bars trusses are connected with the surfaces of the washers which are attached on the anchor blocks using tie constraints. The washers are connected to the anchor blocks using surface to surface contact with “hard contact” normal, and with friction coefficient 0.6 tangential behaviors. The beams are connected to the anchor blocks using tie constraints. Also, the beams are connected to the column using surface to surface contact with “hard contact” normal, and with friction coefficient 0.6 tangential behaviors. The FEDs were modeled using connector elements between the beams flanges and the column, where a perfectly plastic law was specified for the connector material. For the modelling of the column‟s support a beam element has been inserted under the column and has been restrained on the base in all the translational degrees of freedom but is free to rotate. The top point of the beam

105

element is constrained with the column‟s base using kinematic coupling constraint in all the degrees of freedom. The rollers supports on the beams have been modeled by applying restrains on the vertical translational degree of freedom in certain areas of the beams flanges.