· · · Institutional Repository

This research focuses on the behaviour of a granular scour protection around piles when the bed is subjected to bed degradation. Furthermore, engineering guidelines are provided to account for bed degradation in the design of the initial scour protection. To address the behaviour of the protection, a physical model is used and to obtain the engineering guidelines, a theoretical approach is used which includes the phenomena that are observed during the physical model tests. As stated by many authors [Breusers et al., 1991; Chiew, 2004; Melville et al., 2000], a granular scour protection around a pile can fail due to several mechanisms. In clear-water conditions these mechanisms include shear failure, winnowing and edge scour and in live-bed conditions two other mechanisms can be added, namely bed degradation and bed form induced failure. This research focuses on the bed degradation mechanism, which can endanger the granular scour protection around a pile. Just like Van der Hoeven [2002] observed during his research on falling aprons, the stones spread evenly over the front and side slopes of the mound, which form after undermining of the protection has led to the ‘falling’ of the edge stones of the protection. This coverage of stones on the slopes of the mound effectively rearmoured the slopes and prevented further erosion. At the back of the mound this rearmouring is not observed, because the high turbulence levels, introduced by both the mound and the pile, caused stones to become unstable in this area. Furthermore, at the front and side slopes, where the slope of the mound was totally covered by the stones of the protection, the slope angles were constant and comparable to the slope angles that Van der Hoeven found in his research, namely 1:2,0. However, a remarkable difference between the observations of Van der Hoeven and the observations of the current research concerns the layer thickness of the protection on the slope. Van der Hoeven observed that the layer thickness after launching of the apron was only 1*df,50 thick, while the current observations show a decreasing layer thickness from the top of the slope to the toe of the mound. The theoretical approach that is used to develop a design formula focuses on the side slopes, where the flow is parallel to the interface between stones and base material. A balance between the volume of the initial protection and the required volume of stones after bed degradation has led to the formulation of the internal slope angle, γ. This angle describes the assumed linearly decreasing thickness along the slope, starting with a certain required thickness at the top, Df,A, to a thickness of only 1*df,50 at the bottom. Because the hydraulic load decreases with increasing distance from the pile and the thickness of the initial protection is based on the maximum load near the pile, the assumption that Df,A is equal to the initial thickness of the protection, leads to a conservative design formula. This approach is however taken in the development of the simplified design formula, which results in a linear relation between the bed degradation and the extra extent of the initial protection that should be constructed to deal with that bed degradation.

Het ontwerp speelt in op de problematiek krimp. De ontwerplocatie ligt midden in een groot krimpgebied, namelijk Sesia Valley ten noorden van Milaan. De gekozen doelgroep speelt een grote rol in de oplossingsrichting wat gestuurd word door het thema in-between; Stad en land en publiek en privé.

In the construction of dams and in protection of civil engineering works in water courses often use is made of dumped stones. The dumping process can be divided in two stages. During the first stage the resulting mound of stones on the bottom of the water course is built up vertically. In the second stage the stones are under the angle of repose and increase of height of the mound can only be accomplished together with an increase in width of the basis. In the dumping process four sub-processes can be distinguished, i.e. the positioning of the dumping vessel, the start of the falling process, the falling process itself and jumping and rolling of the stones that hit the bottom. Three different partial processes can take place during the falling process. The first one is a (semi-) diffusion process, which results in a swinging motion of a singular stone. The mathematical description of this process is called the "Single Stone Model". On average the distance of the final position of the stones, measured from the projection on the bottom of the point where the fall is started is zero. The spreading ofthe stones is proportional to the square root ofthe water depth. A second possible partial process is caused by the Magnus effect. A constant horizontal force is executed on the vertically falling body. The third partial process can be caused by asymmetric flow separation. This also causes a horizontal force to act on the body. Drag causes the movement to be uniform. Last two mentioned processes result in a ring- shaped mound. The diameter of the ring is proportional to the water depth. The mathematical model of this process is called the "ring model". In practice all processes will take place. The mathematical descriptions can be combined in a combination model.

In a flume of BallastHam Dredging a falling apron model has been constructed and loaded by current. The tests have been done with different rock sizes, different layer thickness of the top storage of the apron and two different gradings. In summary it was found that for both the narrow graded rock layers as well as for wide graded riprap the final slope of the apron was always 1:2, while the riprap layer reached a thickness equal to dn50. It was also clear from the tests that the slope did not stop the transport of sand on the slope, but retarded it considerably. The process of settlement of the riprap was that the stones did not roll over each other, but that the whole structure moved down, like on an escalator. For the protection of the abutments of the Januma Bridge in Bangladesh also a falling apron has been applied. This apron has been designed according the standard design rules, applying also a (relatively costly) wide grading of rock. In the framework of the maintenance of the structure, yearly surveys are made of the falling apron. These surveys show that also in prototype the slopes reached a steepness of approximately 1:2. A volumetric analysis indicated that also in reality the thickness of the falling apron was in the order of only one layer of stones

Traditional guidelines on rock protection at circular piers predominantly focus on preventing shear failure (by choosing a sufficiently large rock size), winnowing failure (by designing an appropriate filter) and edge failure (by selecting a sufficient extent). In particular areas (e.g. in an eroding river channel or in an area with large-scale bed forms), the rock protection may also face a degradation (lowering) of the bed. As a result, the edges will be undermined and stones will roll down and cover the newly developed slope (a falling apron). This process is not well understood and theoretically-based design guidelines for falling aprons are not available, only empirical relationships for bank revetments. Indicative laboratory experiments were conducted in order to derive guidelines to account for bed degradation in the design of rock protection at circular piers under currents. This paper summarizes the experimental set-up, monitoring techniques, test program and results of the conducted experiments. All tested rock protection layouts consisted of a single stone grading with a sufficient stone size and layer thickness to prevent shear failure and winnowing failure. In total, 7 live-bed tests were conducted with varying current velocities, bed degradation levels and rock layouts. During the tests, the bed protection level near the pier was monitored with video cameras. Before and after each test, the bathymetry was recorded by stereo-photography. The analysis focussed on the successive failure stages, the redistribution of stones over the slope and the evolution in time. The results showed that, as the undermining progressed, the stones at the edges of the protection were redistributed through a combination of rolling, sliding and sinking. Finally, a protective mound was formed, with stones covering the slopes on all sides. The observed stone layer thickness on the slopes gradually reduced towards the outside. A design rule was derived to account for bed degradation by quantifying the stone volume required for a falling apron.