Finite Element Response Sensitivity Analysis of Steel-Concrete Composite Beams with Deformable Shear Connection

Abstract

The behavior of steel-concrete composite beams is strongly influenced by the type of shear connection between the steel beam and the concrete slab. For accurate analytical predictions, the structural model must account for the interlayer slip between these two components. In numerous engineering applications (e.g., in the fields of structural optimization, structural reliability analysis and finite element model updating), accurate response sensitivity calculations are needed as much as the corresponding response simulation results. This paper focuses on a procedure for response sensitivity analysis of steel-concrete composite structures using displacement-based locking-free frame elements including deformable shear connection with fiber discretization of the cross-section. Realistic cyclic uniaxial constitutive laws are adopted for the steel and concrete materials as well as for the shear connection. The finite element response sensitivity analysis is performed according to the Direct Differentiation Method (DDM).The concrete and shear connection material models as well as the static condensation procedure at the element level are extended for response sensitivity computations. Two steel-concrete composite structures for which experimental test results are available in the literature are used as realistic testbeds for response and response sensitivity analysis. These benchmark structures consist of a non-symmetric, two-span continuous beam subjected to monotonic loading, and a frame sub-assemblage under cyclic loading. The new analytical derivations for response sensi- 3

Finite element structural response sensitivity and reliability analyses using smooth versus non-smooth material constitutive models

“…(45), where the identity between continuum and consistent tangent moduli for uniaxial constitutive models is used. Other researchers (Haukaas and Der Kiureghian, 2004) found that continuity of finite element response sensitivities can be obtained by using smooth multi-axial constitutive models.…”

Finite element response sensitivity analysis using three‐field mixed formulation: general theory and application to frame structures

“…2). The procedure for response sensitivity calculation in presence of static condensation has been previously derived by the authors and can be found elsewhere [32]. This finite element was proven to provide accurate response simulations and to be superior in the evaluation of local quantities (e.g., section deformations, section stress resultants, shear force distribution at the steel-concrete interface, etc.)…”

“…The obtained three-field mixed monolithic frame element is characterized by exact distributions of section stress resultants (as for force-based frame elements [19]), while the assumed displacement fields have the same form as for standard displacement-based frame elements. The above useful feature allows to assemble easily monolithic and composite frame elements in frame models, representing correctly the connections between steel columns and steel beams or reinforced concrete columns and concrete slabs [32].…”

“…The values of the material constitutive parameters are given in Table 1. Reference [41] as well as Reference [32] also provide comparisons between analytical predictions and experimental results for the response of steel-concrete composite beams and frame structures modeled using a displacement-based frame element with deformable shear connection.…”

“…In fact, the concrete never reaches its peak strength and therefore the response sensitivities with respect to ε f and f f are equal to zero for the entire pushover and particular attention has been given in choosing the lower value of parameter perturbation so as to avoid numerical problems related to round-off errors ("step-size dilemma", see [5,6,8,32,42]). Convergence studies for other response quantities and other material parameters have also been performed, giving similar results in terms of convergence of forward FDM computations to DDM sensitivities for decreasing parameter perturbation values [41].…”