Arne Jansseune

Influence of Finite Element Modeling Choices in the Assessment of Stress Concentrations at Fatigue Prone Locations in Orthotropic Bridge Decks

In orthotropic bridge decks, the rib to floorbeam connection is a major source of fatigue problems. Commonly, the trapezoidal ribs cross the floorbeam continuously necessitating clearance holes in its web, and frequently additional web cutouts are foreseen to relieve the ribs lower edges. This solution is favorable for rib cracking but will generate stress concentrations in the web itself. The shape of the additional cutout has a major influence on the sizes of the concentrations rendering differences of a factor 2 or 3 for corresponding overall geometries and loading schemes. Various authors have studied cutout shapes through full scale testing or by computation by finite element modeling. This paper presents such a study, but focuses on the influence of the finite element modeling itself. It is shown that the mesh density, the element type, the choice between shell and volume elements. This is an important finding and should not be overlooked when comparing finite element based results to coded values or measured results. In order to do so, the results of the numerical work are compared to strain results from full scale tests.

Elastic Properties and Failure Behavior of Tiled Laminate Composites

This paper focuses on the elastic properties and the failure behavior of tiled laminate composites. Such laminates, in which the plies are not parallel to the outer surfaces are found in InfraCore® based GFRP panels. This technology is developed for the construction of a robust FRP panel that is applicable for highly loaded structures, e.g. for bridges or lock gates. In general, the drawback in traditional FRP sandwich structures has always been debonding of skin and core. Such a debonding problem may occur after impact, followed by fatigue loading. Through the use of the InfraCore® technology, debonding is no longer possible, as multiple overlapping Z-shaped and two-flanged web structures are alternated with polyurethane foam cores acting as non-structural permanent formwork. Consequently, the fibers in the upper and lower skins as well as in the vertical webs run in all directions, especially in the connection between them, rendering a resin-dominated crack propagation impossible. As a result of the integration of core and skin reinforcement, a skin material is created in which the reinforcement is not parallel to the outer surfaces, but at a small angle. Such stacking is called a tiled laminate (TL), as opposed to plane-parallel (PP) and is not fully described by the classic laminate theory. In the paper, finite element analysis is used to assess the effect of the ply angle and the interlaminar properties on the assessment of stiffness and failure behavior of a tiled laminate.

Properties of Fiber Reinforced High Performance Cement Based Overlays for Orthotropic Bridge Decks

An increasing number of steel orthotropic bridge deck require repair due to fatigue cracking. For the rib to deckplate joint, a possible solution is to reinforce the fatigue prone deck plate by replacing the asphalt wearing surface with a reinforced high strength concrete overlay. However, this process has been proven very difficult to execute and consequently has some issues regarding durability. This paper reports on high performance fiber reinforced cement based overlay solutions with gritted epoxy interlayer for a typical orthotropic deck with inadequate plate dimensions. Material characterization of three mix designs includes determination of compressive and tensile strength, and post crack behavior by CMOD registration from 3PBT. Furthermore, the steel-concrete composite action in the positive and negative bending moment zone is investigated experimentally. The test results give insight into the concrete behavior, the steel-epoxy-concrete interaction, and are used as input for a 3D finite element model of the reinforced deck under fatigue loading. The results indicate that fiber reinforced cement based overlay solutions may be a viable option for orthotropic decks with inadequate deck plate thicknesses

Review of FE-Based Fatigue Evaluation Methods for the Rib to Floorbeam Welds in Orthotropic Bridge Decks

Complex welded structures such as bridges are very often designed with the help of FE analysis. However, one should remain cautious when using such an analysis, since the results are mesh sensitive, with especially the mesh density and the element type influencing the results. In addition, these results are in most cases retrieved in hot spot areas with high stress gradients, where the maximum stress even cannot be correctly determined with linear elastic finite element analysis. For that reason, a stress evaluation method is required to obtain relevant stress levels that can be directly related to fatigue detailing. The most complete set of stress evaluation recommendations is given in the Recommendations for Fatigue Design of Welded Joints and Components from the International Institute of Welding. Nevertheless, several authors have recently commented on the difficulties regarding the application of these methods for the rib to floorbeam welds in orthotropic bridge decks. This paper provides findings for this type of connections based on both shell and solid model analysis and relates these findings to work from other authors.

Plastic collapse of steel water towers

The aim of this contribution is to verify an existing design rule for steel liquid-filled conical shells by performing numerical simulations. Questions were raised regarding the reliability of the design rule as early yielding comes into play if an axisymmetric imperfection is located near the lower rim where the stresses are elevated. In our study, full nonlinear numerical simulations are performed with a realistic weld depression as imperfection shape. This weld depression is placed at the most detrimental location near the lower rim. The results of the simulations show that for cones with a moderate relative slenderness the design rule indeed overestimates the buckling strength, but that the error is not pronounced and that the design rule may be considered as still acceptable.