Compressive Membrane Action in Prestressed Concrete Deck Slabs

Abstract

One of the most important questions that structural engineers all over the world are dealing with is the safety of the existing structures. In the Netherlands, there are a large number of transversely prestressed bridge decks that have been built in the last century and now need to be investigated for their structural safety under the actual (increased) traffic loads, for the rest of their service life. This research is an attempt to investigate the bearing (punching shear) capacity of such bridge decks under concentrated loads (wheel loads). Using the actual design codes for the verification of the bearing capacity leads to values suggesting that the safety standards are not met. However, since the bridge decks are laterally restrained by the supporting beams it is expected that compressive membrane action (CMA) exists in such deck slabs, and that the transverse prestressing of the deck slab in combination with CMA will enhance the bearing capacity, making thinner deck slabs possible with no problems of serviceability and structural safety. This thesis begins with an introduction to the research topic, listing briefly the background and the objectives, and concluding with the research strategy. A literature review regarding the punching shear capacity of transversely prestressed concrete decks and compressive membrane action has also been carried out. First, the general mechanism of punching shear and compressive membrane action is explained along with the relevant analysis methods and code provisions and then important experimental investigations done on prestressed deck slabs are briefly described. It is concluded that there is a need to investigate the bearing capacity of transversely prestressed concrete deck slabs supported by and connected to concrete girders using a large scale model since most of the past research is either done on concrete decks with steel girders or on small scale models. In order to investigate the research problem experimentally, laboratory tests on a 1:2 scale bridge model of a real bridge in the Netherlands have been performed. The model bridge consisted of a thin, transversely prestressed concrete deck (with unbonded tendons), cast in-situ between the flanges of long prestressed concrete girders. Prestressed transverse beams were also provided close to either end of the bridge deck. The interface between the deck slab and the girder flanges was either straight or skewed and two types of loads were applied: single and double. Loads were applied at midspan and close to the deck slab-girder flange interface. All the tests showed failure in punching shear (either brittle punching or flexural punching) regardless of the type and position of the load. Failure always occurred in the span of the slab, whereas the interface remained undamaged. The effect of various parameters, like the transverse prestressing level (TPL), the type and position of the load(s), the inclination of the joint (interface), the size of the loading plate etc., on the bearing capacity were also studied. As part of the numerical investigation, a 3D solid, 1:2 scale model of the real bridge, similar to the experimental model, was developed in the finite element software DIANA and several nonlinear analyses were carried out. A comparison with the experimental results was made proving that satisfactory results were obtained that validated the finite element model. The normal forces arising from compressive membrane action were determined with the help of composed elements. A detailed parametric study was also carried out involving numerical modeling parameters, like the mesh size, displacement-load step size etc., and the material and geometrical parameters, similar to the experimental parametric study. In addition to that, the size effect was studied by carrying out a nonlinear analysis on a 3D solid model of the real bridge, showing that a size factor of 1.2 is appropriate to convert the results of the model bridge deck with 100 mm thickness to those for the real bridge deck with a thickness of 200 mm. A theoretical analysis of the model bridge deck was then carried out and it was demonstrated that the ultimate load carrying capacity as found from the experiments and the finite element analysis was much higher than predicted by governing codes and theoretical methods. The discrepancy was attributed to the lack of consideration of CMA in the theoretical approaches. In order to incorporate CMA in the analysis, the normal forces arising from compressive membrane action and determined via the finite element analysis were used in the fib Model Code 2010 punching shear provisions (based on the Critical Shear Crack Theory) to determine the ultimate bearing (punching shear) capacity. Calculations were performed at two Levels-of-Approximation (LoA); Elementary LoA (without CMA) and Advanced LoA (with CMA). Generally, it was observed that an increase in the TPL improved the behavior of the bridge deck with regard to both serviceability and ultimate limit state. An average safety factor of 3.25 was obtained when the projected model bridge design capacity and the real bridge design capacity were compared with the design wheel load. It can be concluded that the existing bridges still have sufficient residual bearing capacity considering the beneficial effect of CMA. Moreover it was shown that appropriate nonlinear finite element models can predict the load bearing capacity quite accurately. The research described in this thesis, resulting in methods for the analysis of bridge decks including compressive membrane action, has the potential to result in considerable cost savings, since the models are able to demonstrate that many existing bridge decks are safe enough, contrary to earlier expectations. A proposal has been prepared to introduce the effect of compressive membrane action into the calculation models for punching shear offered in the fib Model Code for Concrete Structures 2010. To this end two more Levels of Approximation are added to the first three given already in the code. The new level IV enables the use of the Critical Shear Crack Theory in combination with the calculation of the curvature of the area around the concentrated load with a nonlinear finite element analysis using shell elements. The level V enables the prediction of the punching shear capacity with a tailored NLFE-program using composed elements.