Case study: Nonlinear behaviour skew-plate viaduct Burgerveen

A necessary strengthening of a fly-over at Burgerveen has served as a case study for nonlinear structural analysis of a skew plate-viaduct. The main objective of the study was to assess the degree of safety, concerning the critical end-field of the viaduct, before and after strengthening. The geometric model of the viaduct was justified by linear elastic analyses. Analyses were performed in order to quantify the influence of element formulation, the influence of horizontal curvature and the influence of reinforcement under linear elastic conditions. In particular, the significance of an accurate representation of the horizontal curvature, was demonstrated. In schematization of the existing viaduct, lateral spring elements were used along the lines of support, in order to allow a realistic distribution of reaction forces. The critical end-field and the close part of the adjacent field was represented by nonlinear material models, while the three remaining fields were treated as linearly elastic. In the nonlinear area, a fixed smeared crack model was used for the concrete, and an elastoplastic formulation for the reinforcement. Three distinguished formulations of the crack model were applied, a Tension Stiffening model, a Reduced Tension Stiffening model and a Brittle model. With all models, a significantly higher load-carrying capacity was found, than obtained by simple design calculations under Ultimate Limit State conditions. The results of the nonlinear analyses show that local failure occurred in the critical field of the viaduct before yielding in the reinforcement appeared. This was due to the small reinforcement ratio in this area. Strengthening of the viaduct was carried out by casting deep reinforced concrete beams along both of the free boundaries. After curing, the boundary beams were posttensioned. Also in the analysis of the strengthened viaduct, the critical field was modeled nonlinearly, while the remaining three fields were taken as linearly elastic. In the nonlinear area, the concrete was represented by a Reduced Tension Stiffening model, whereas the reinforcement was modeled by an elastoplastic model. The analyses were performed by a phased modeling technique, allowing a realistic simulation of the strengthening. Hence, the deadweight of the viaduct and the boundary beams was applied in advance of the posttensioning. The results show that the load carrying capacity was significantly increased, and that the presence of the boundary beams povided a ductile failure mode. Due to the inclination of the posttensioning cables in the anchoring area, relatively large vertical reaction forces appeared. These caused local negative bending moments about the end-supports, which necessitated additional top reinforcement.