In this assignment we simulate a two meters wide dike stretch which is subjected to overflow. Three failure mechanisms could occur due to overflow: full erosion of the soil of the dike, macro instability of the levee if too much soil is eroded away, leading to insufficient counterweight on the inside of the levee, and surface slip failure due to reduction of strength of the dike by flow of water in the dike core. Prevention of failure can be achieved by either keeping the water away from the berm surface or holding the soil particles in place while the water flows alongside it. The desired repair is an emergency repair that can placed quickly during a storm, or between two storms to prevent a damaged levee from failing. This is a temporary repair that will later be removed for full reconstruction of the levee. This provides constraints to the installation and repair procedure. To decide which solution is best, 10 concept solutions are compared based on their score in the Multi Criteria Analysis with 8 weighted criteria. The solution with the highest score is the best solution according to the Multi Criteria Analysis. The selected repair option is to locally cover the damaged area with flexible overlapping sheets of Tyvek®. This is a light and thin material that can easily be transported and cut to size at the location of the repair. These sheets can be easily secured against flow loads using Gripple anchors, while pins are used to keep the Tyvek® in place during windy conditions. An installation plan and cost breakdown was made. Using 3 m wide Tyvek® rolls, the costs would be around 19.19 euros per meter height. The damaged areas of the levee are covered with Tyvek® which is a waterproof, damp open material. This prevents the penetration of water into the core through the damaged areas. The Tyvek® also prevents the exposed soil from being eroded. Moreover, Tyvek® is a very strong material that won’t be torn off. As there is no strength reduction due to water penetration or further loss of surface material the chance for macro instability is minimized. First, a design was created, with components dimensioned for a dike stretch in the Hedwigepolder. Subsequently, the design was validated with tests. Including two types of damage. The first was removal of a 2 by 2 m grass layer (0.1 m thick) at the top of the dike and the second was a dug-out step-section across the entire width of the dike with a height difference of 0.5 meters at the toe of the dike. Only the second damage went through the clay layer and into the sand core of dike. During and after the tests some important observations were made. During all test volumes the plastic sheets were kept in place due to the anchors and pins. Some pins however, came loose due to vibration of the sheets (we presumed the cause of vibration is flow of water between two sheets or turbulence of water). The trench of the lower damage was partly washed away by the water. The erosion of soil in the trench did not result in changing the stability of the top sheet. At the lower damage the plastic sheet was torn off after a tear of 1 m was made with a knife. Our solution is designed to protect any dike from macro-instability by keeping the water flow away from the berm. This is done by layering multiple sheets of plastic foil over the inner berm, fastened with ground anchors and pins. The solution is a durable and reliable design due to proven waterproof, damp open and UV resistant properties, together with Gripple's demonstrated anchoring capabilities. Another major advantage of our solution is that the repair of the damage is selective, i.e. the solution is applied only to the parts of the levee where damage has occurred. In case of only localized and small damage to the levee, the costs and workload are very minimal compared to other solutions that are applied to the entire surface area of the levee.

The demand to build with wood and other bio-based materials has increased rapidly the last decade, due to the ambition to build more with sustainable building materials. One of the major challenges in building with wood, is to account for the lower strength and stiffness perpendicular to the grain. Wood is an anisotropic material, meaning the material has different properties in different directions. Loading in the perpendicular to the grain direction is sometimes unavoidable and fracture often occurs in a brittle manner. Unfortunately, our understanding and knowledge of the perpendicular to the grain fracture is limited. This is especially the case for hardwoods, for which the material properties and the parameters important for the fracture process are often unknown. The goal of this thesis is to reduce this knowledge gap on the material properties and the relevant parameters in mode-I (pure tension) fracture process of (tropical) hardwood. The knowledge gained in this research provides the answer to the main research question of this thesis: Which parameters should be considered when describing the constitutive model of a (tropical) hardwood in mode-I tension? A single-end notch three-point bending (SEN-TPB) test demonstrated a strong influence of the orientation of the growth rings on the perpendicular to the grain fracture process of azobé. The value for the peak load and post-peak behaviour change significantly depending on the orientation in which the sample is placed. Furthermore, visual observations and results from a profilometer revealed that there is an inconsistency in the roughness of the cracked surface area between the different series. From the microscopic analysis is concluded that this inconsistency in general mechanical behaviour and the roughness of the cracked surface area is linked to the composition of ray and parenchyma cells. Moreover, the direction of the ray and parenchyma cells with respect to the fracture plane is different depending on the orientation of the sample. When the ray cells are orientated perpendicular to the crack plane, a higher strength is obtained. The thin-walled parenchyma cells significantly reduce the post-elastic stiffness if these cells are orientated perpendicular to the crack plane. The influence of the composition and orientation of both cells is reflected in the numerical value for the modulus of elasticity, the tensile strength and the fracture energy. This research showed that the composition of the cells in a wood specie and the orientation in which the growth rings are placed significantly influence the mode-I fracture process of a hardwood. This influence is reflected in the modulus of elasticity, the tensile strength, the fracture energy and the shape of the tension softening curve perpendicular to the grain. The results of this research provide new insight into the fracture modelling of hardwood. Both in numerical modelling and design considerations the variation in the material properties because of the orientation of the growth rings, should not be neglected.