Search results

Accidental loadings due to blast or impact may easily cause failure of the elements that are exposed or located in the vicinity of the hazard, leading in some cases, to the progressive collapse of the whole structure; therefore, assessment of the structural over strength is critical for structural engineers to ensure a certain level of security and validate alternative unloading paths. The T-stub model is used to describe the behaviour of components i) “column flange in bending” and ii) “endplate in bending” usually present in a beam-to-column bending resistant connection [1]. These components are responsible for the behaviour in the tension zone of connections, being able to provide ductility to a connection; thus, proper characterization of T-stub behaviour under impact loading is crucial. In this paper, a 3D finite element model exploring the behaviour of a welded T-stub with flange thickness of 10 mm (T-10) (Fig. 1) is validated against experimental results from: i) one quasi-static loading (reference case) (grey dotted line Fig. 2 and Fig. 4); ii) and two rapidly applied dynamic loadings according to the gas pressure in the chamber: a. 120 Bar (Impact #1 - T10-D120-160 - Fig. 2); and b. 160 Bar (Impact #1 - T10-D160 - Fig. 4) [2]. The steel grade of the T-stub is S355 and the bolts M20 grade 8.8 are fully threaded. The dynamic loading simulations take into account the elevated strain rate effects in the stress enhancement, based on dynamic increase factors, following the Johnson-Cook material model [3]. The dynamic loadings are applied as a boundary condition in the “pull out surface” (Fig. 1) considering the Tstub’s transient displacement responses obtained from experimental tests; maximum displacement values are reached in approximately 0.08 sec. The accuracy of the numerical force-displacement predictions for both quasi-static and dynamic loading schemes confirms that the Johnson-Cook material model used, provide accurate stress enhancement to describe the behaviour of bolted steel connections subject to impact loadings. From Fig. 2 and Fig. 4, it can be observed that the elastic stiffness remains unchanged for all loading schemes: ki = 180 kN/m, as the steel’s elastic modulus introduced in the numerical models are the same for both quasi-static and dynamic situations; moreover, the strain rates developed are similar for both dynamic loading (Fig. 3), inducing the same dynamic increase factors for the stress enhancement; the F-δ flows are therefore, similar for both numerical dynamic responses but with different failure displacements. Plastic resistances of the T-stub: FRd,quasi-static = 161 kN and FRd,120 Bar = FRd,160 Bar = 195 kN; corresponding to an enhancement of +21% of the plastic resistance due to the elevated strain rate effects. Fig. 3 illustrates the pattern of the strain rate (ER), ranging from 1/s to 3/s in the plastic hinge developed next to the weld toe, corresponding to a DIFs of 1.27 and 1.31. Furthermore, comparison of the equivalent plastic strain (PEEQ) pattern for both loading situations, shows that two plastic hinges are developed per flange leg, consistently with the plastic failure mode type 1 predicted by the Eurocode 3, part 1.8 [1]. However, in the dynamic case, the plastic hinges are slightly underdeveloped and higher strains are required in the bolt to meet the same deformation level.

Current paper presents a finite element analyses for the characterization of the nonlinear behaviour of bolted t-stub component subject to impact loading followed by fire. The proposed numerical model has previously validated against experimental results under monotonic static loading at ambient and elevated temperatures (Ribeiro et al., 2013). 3D solid and contact elements from the finite element package Abaqus are used to perform the structural model. The temperature dependent material properties, the geometrical and material nonlinearities (including the strain rate sensitivity) were taken into account to predict the failure of the t-stub. A parametric study was conducted to to provide insight into the overall behavior, namely their stiffness, resistance, ductility and failure modes due to the effects of dynamic loading followed by fire.

Since the partial collapse of the Ronan Point apartment in London in 1968, requirements for the avoidance of disproportionate collapse are addressed in the design codes. Despite these requirements, the ability of steel connections to sustain large tensile forces whilst undergoing significant rotations has been questioned by recent studies and real evidences (as for example the collapse of the WTC towers in 2001). A point raised in FEMA’s report is: “Connection performance under impact loads… needs to be analytically understood and quantified for improved design capabilities and performance as critical components in structural frames”. Additionally, a recent report presented by Arup made the following recommendation (rec. nº 26): “… the strain rate enhancement of yield strengths in connections could still be important. It is recommended that research is undertaken to examine this effect using rate-sensitive material models”. Ductility of joints under accidental loadings, such as impact and fire are being investigated in the scope of the FCT project “IMPACTFIRE”, at the University of Coimbra. This paper presents and describes the results of a finite element model for the characterization of: i) the nonlinear behaviour of a bolted t-stub component under impact loading and ii) its strain-rate sensitivity. In order to identify relevant parameters that influence the dynamic behaviour of the t-stub, the effect of the loading magnitude, the effect of t-stub thickness and advantages of using implicit or explicit integration procedures are also studied.

O presente artigo apresenta uma avaliação numérica do comportamento de ligações aparafusadas com placa de extremidade, sujeitas a carregamentos de impacto. O modelo numérico foi previamente validado com os resultados de ensaios experimentais das mesmas ligações quando sujeitas a carregamentos estáticos. Este modelo foi desenvolvido no software ABAQUS, recorrendo a algoritmos de integração directa para a determinação da resposta da ligação quando solicitada por uma força instantânea. Dos resultados obtidos, concluiu-se que a ligação estudada quando sujeita a carregamentos dinâmicos apresenta uma maior capacidade resistente e que o seu modo de rotura tende para modos de roturas com menor ductilidade.

After the accidents occurred during the first decade of this millennium, such as the World Trade Center (2001), London (2005) and Madrid (2004), special attention was given to the study of robust structures subjected to different accidental loads. The World Trade Center attack highlighted troublesome weakness in connections, which exhibited poor performance caused by brittle failure. Structural details played a very significant behavioural role when the structure is subjected to impulsive loads [1]. Concerning the behaviour of steel joint, the literature presents several studies on steel connections under both static and cyclic loads [2, 3]; many results of these studies have contributed for improving the current standards, such as the Eurocode 3, part 1-8 [4]. However, only scarce information exists concerning the behaviour of these joints directly loaded by higher loading rates [5]. This paper is devoted to the report of an experimental programme on steel joints under impact loading, in particular to the assessment of T-stub response under tension. The T-stub is used to evaluate the behaviour of the tensile components that are responsible for the deformability of the joint, such as the end plate in bending. Firstly, the paper describes the features of an experimental system developed at the University of Coimbra, to apply high rates of loading; then, it presents the experimental campaign and the corresponding results. The test apparatus is defined by a rigid reaction frame fixed to a reaction slab and connected to a rigid “flying beam” (HEM 340, S355J2); the impact force is applied in this beam through a pneumatic driven cylinder ( = 125 mm). This “flying beam” consists in a second class lever pivot located at the opposite end of the cylinder location, and the tested specimen is subjected to the dynamic force at the middle of this beam. The pneumatic cylinder was designed to work with a maximum operating pressure of 30 MPa. During the impact tests, force, displacements, accelerations and strains are measured. Because this type of tests occurs in a very short time intervals (hundredths of a second), specific equipment with large sample rate are used. The experimental programme includes two impact tests on welded T-stubs: i) test T-10-D120-160 - rapidly applied loading of 120 Bar [12 MPa], followed by 160 Bar [16 MPa]; and ii) test T-10-D160 - rapidly applied loading of a single impact equal to 160 Bar [16 MPa]. The results of these tests are compared against reference quasi-static tests [6]. The T-stub geometry is defined by two plates, the flange and the web, both with 10 mm of thickness and welded by means of a continuous 45º fillet. The flange is bolted through two bolts M20, grade 8.8 fully threaded.

The work presented in this paper is part of an ongoing research project at the University of Coimbra IMPACTFIRE PTDC/ECM/110807/2009, which the main focus is the characterization of the behaviour of bolted steel connections subjected to accidental loads, such as impact and fire. Detailed description of the experimental parts developed, designed and fabricated at University of Coimbra, to carry out tests under high rates of loading is presented. This experimental part is operated by high pressure nitrogen comprising three main components: pneumatic reservoir, pneumatic cylinder and a rapidly opening valve, which allows the instantaneous nitrogen flow from the reservoir to the cylinder. Furthermore, the data acquisition system, the methodology for analysis of the results and the results of preliminary tests are also reported.