Momentcurvature Relation for a Beam Using Shell Finite Element Model In many practical applications it is advantageous to use a hierarchical modeling approach. In general, it involves using a higherorder mathematical model (e.g., solids, shells) to obtain an accurate response for a subsystem and then embedding this information into a simplified model (e.g., beams, springs) and using this simplified model for analysis of the entire structure. In this News, we present an application of this approach to a civil engineering problem. The aim is to model the response of a large structure mainly made of unsymmetric Ibeam steel girders with horizontal and vertical stiffeners that undergo elastoplastic deformation and also experience local buckling of the lower flange and parts of the stiffeners. A simple beam element model can’t capture all these complex behaviors mainly due to the kinematic assumptions of the beam. Hence, we use a detailed MITC4 shell finite element model to study the behavior of a representative cell of the beam (Figure 1) and to establish its momentcurvature response. Later, we can embed this generalized response into the momentcurvature beam element in ADINA for analysis of the entire structure. A similar approach was adopted in the simulation of buckling of bridge braces where shell elements were used to construct the generalized forcedisplacement response of the beams. Such an approach has several advantages over using shell/solid elements for modeling the entire structure:
Figure 1 shows the schematic of the model. The left and right ends of the shell model are each constrained to the centroid of the corresponding crosssection using rigid links and each are subject to equal and opposite incremental rotations up to 0.02 rad. Note that due to symmetry, we could have used only one half of the model.
In both animations above, the elastoplastic behavior of the material is taken into account. In the left animation, the displacements are assumed to be small. Hence, no local buckling is detected while in the second analysis (animation on the right) nonlinear kinematics are included and as can be seen, towards the end of loading, the lower flange and some of the stiffeners experience buckling. Note that to highlight local buckling, displacements are magnified in both cases. The momentrotation responses of both models are depicted in Figures 2 and 3 below. Figure 2 represents the behavior when kinematic nonlinearity is not taken into account. The response consists of 3 distinct phases — I:linear response (material is still elastic), II:partial plasticity of the section, III:section becomes fully plastic and there is a plateau in the response with small slope due to the hardening.
Figure 3 represents the behavior when kinematic nonlinearity is included. It consists of 4 distinct phases — I, II, and III are similar to the above while IV is where the local buckling happens and the moment bearing capacity of the beam drops as the end rotations increase beyond a critical value.
Of course, there are other ways for obtaining momentcurvature response of a beam. One can use physical experiments or even use introductory strength of materialtype calculations when the local buckling effect is not important and the crosssection is relatively simple. Another example of such analyses in civil engineering applications is encountered in modeling steel beam/column connections in buildings. There, a detailed finite element model involving solids or shell elements can be used to study the elastoplastic response of these connections. Later, the result can be used in defining nonlinear rotational springs in the frameelement model of the building for further analyses of the entire structure. The momentcurvature beam element in ADINA provides a powerful tool for modeling slender structures, especially when nonlinear elastic or elastoplastic response is involved, local buckling is of interest or the crosssection of the beam is rather complex. ADINA Primer Problem 14 gives a tutorial on using momentcurvature beams in ADINA.
