PSI - Issue 78

Gaetano Della Corte et al. / Procedia Structural Integrity 78 (2026) 199–206

201

assuming an unstiffened base plate can lead to significant underestimation of the stiffness and the resistance of the connections. Since plate stiffeners can be of significant aid in designing steel-to-concrete connections subjected to combinations of large axial forces and bending moments, extending the component method and verifying it for stiffened plates is a topic worthy of investigation. Besides, the calculation rules in EN 1993-1-8 were written assuming a proportional load path, that is a proportional increase of all the loads acting on the connection. Though, in principle, considering non-proportional load paths require relatively simple formal adjustments to the model equations, the ability of the method to predict the connection response for such alternative load paths is also worthy of investigation, especially when the formal changes are combined with the more substantial topological change of a stiffened base plate. Clearly, column base plate connections can be taken as the starting point for the investigation. In fact, column base plate connections comprise all the features encountered when dealing with connections between a steel exoskeleton and a concrete frame. The EN 1993-1-8 rules are provided for both tension and compression axial forces. In the case of a simultaneous bending moment acting on the connection, the code provides distinct rules for the two cases of “small” or “large” force eccentricity: for a small eccentricity equilibrium is reached with only compressive contact stresses or only tension anchor forces, whereas for a large eccentricity the connecting plate is partially compressed and tension forces exist in the anchors. The case of a large tension axial force is of less significance for the purposes of this study. Indeed, in the case of no compressive contact stress, the presence of stiffeners could be easily dealt with using the rules already established by the EN-1993-1-8 to characterize equivalent T-stubs. Therefore, this study considered the case of a compressive axial force with large eccentricity. The scope was to investigate how the plate stiffeners affect the location of the neutral axis and, consequently, the response of the connection. For I-section column base connections with unstiffened plates, EN 1993-1-8 assumes that the resultant of compression contact stresses between a connecting plate and the concrete is applied at the center of the column flange in compression. In fact, the column flange acts like a vertical stiffener for the base plate, thus determining a concentration of contact stresses below that column flange. This approach neglects the contribution from the column web in transferring compressive contact stresses, looking for a simplification of the calculation procedure that is justified by the fact that the corresponding column web compressive force is close to the column centroid, thus providing a smaller contribution to the moment equilibrium. However, if plate stiffeners extending beyond the column flange are used, then their contribution cannot be neglected, because they provide additional compressive forces with large lever arms. Therefore, the simple model with a single spring in compression (unstiffened plate) must be substituted with a model having multiple springs corresponding to the vertical stiffeners. Figure 1a illustrates the alternative mechanical model with a single spring in tension (assuming there is only one row of anchors) and a distributed spring system on the compression side of the connection. In the figure, the black solid line indicates a rigid bar, as it is usually assumed within the component method. The neutral axis location will clearly depend on the eccentricity of the axial force. Therefore, in the case of a non proportional loading path, the neutral axis location moves during the loading process; particularly, for a monotonic increase of the bending moment and a constant axial force, there is a continuous shift of the neutral axis towards the plate edge on the compression side. The change in the location of the neutral axis clearly corresponds to a change in the rotational stiffness of the connection. Consequently, a typical moment-rotation response, for a given axial force, will exhibit a reduction in tangent stiffness, as far as the bending moment increases, until the plastic moment resistance is reached. A schematized moment-rotation response curve, corresponding to a constant axial force and an increasing bending moment, is shown in Fig. 1b. To simplify the model, the behavior of the connection is assumed to be approximated by the following four distinct phases of response:  Phase 1: There eccentricity of the axial force is so small that equilibrium is possible with no tension forces in the anchors. The corresponding rotational stiffness will be indicated by S j,1 . This phase ends when a moment symbolized as M j,Dd is reached. 2.2. The “component method” for steel-concrete connections with stiffened plates