Large areas of the world are protected by flood defence systems. A common part of flood defence systems is a levee with grass cover, with the primary function to protect hinterland against floods. Waveovertopping might lead to erosion of the grass cover, followed by erosion of the inner slope and potentially levee breaching. The wave overtopping simulator (WOS) was developed for full-scale and insitu testing of the erosional resistance of levees against wave overtopping. The first experiment with this simulator was carried out in 2007 and more followed. However, the data gathered during these overtopping experiments has always been considered on individual experiment scale or in small subsets, exception being the Cumulative Overload method. The experiments proved to be difficult to reproduce, seemingly identical experiments gave different results. The objective of this thesis is to combine WOS data and multiple grass erosion models to make a prediction for the failure of the grass cover. Key features that determine the prediction of failure are the location and moment of failure. Therefore, these are included in the main research question, which is identified as: How can grass erosion models and WOS experiment data be combined in a new method to generate a prediction of the moment and location of the grass cover failure due to wave overtopping? A literature study on grass erosion and grass erosion provides the basis for modelling grass erosion, a review of combination techniques gives insight in combining results for a prediction. A selection of models is discussed and used for setting up a prediction method. These include three flow-based models which are semi-realistic simplified representations and time-independent over a single overtopping event: Cumulative Overload Method, Analytical Grass Erosion Model and Dean Stream Power. These models are, where required, adjusted to model grass cover failure and to comply with identical hydrodynamic input. A fourth model, the Wave Impact Approach, is an impact-based approach. Each model is calibrated on each WOS experiment, creating a set of calibrated models, which function as the set of predictors. The number of predictors equals the number of models times the number of experiments. During calibration, the moment of failure has been traced back using the control list of the WOS, and vice-versa, resulting in a moment during the experiment at which the grass cover failed. Detailed review of all factual reports on WOS experiments and the nature of the grass erosion models highlighted the need for a locationdependent resistance parameter to determine the location of failure. Therefore, calibration was based on a location-depended resistance parameter; critical flow velocity for the flow-based models and critical pressure for the impact based model. This set of predictors has been used to create a set of predictions for each of the five validation experiments, input being the geometry, loading and initial condition of the slope. The initial condition given by a registration of anomalies to the average grass cover. Failure is determined by vote and averaging. After validation, none of the final predictions proved to be fully correct. Meaning that none of the final predictions correctly indicated the location and moment of the first grass cover failure. Despite this, in four of the five validations at least one correct failure location was indicated when considering the specific anomalies. The main shortcoming is concluded to be the prediction of number of waves until grass cover failure. Based on this research and given a certain average grass cover quality, it is concluded that the resistance against grass erosion by wave overtopping is described by resistance against erosion of anomalies. For at least 22 of 28 sections in the dataset, grass cover failure occurred at an anomaly of the average grass cover. No grass cover failures occurred related to the average grass quality. Especially mole activity showed a significant 46% decrease in the averaged calibrated critical flow velocity with relative to that for the average grass cover. This leads to the conclusion that design and safety assessments should include conditions other than the average grass quality. For future research a method is recommended that divides mole activity into classes, based on certain characteristics. Each class distinguishable by unique combination of properties and assigned a resistance against erosion. For design, the probability of occurrence of each class must be determined and combined with the corresponding influence on the resistance against erosion. For assessment, an inventory or representative sample of animal activity must be available to asses if the occurrences of animal activity are within the design requirements.

Large parts of the Netherlands are vulnerable to flooding. Due to the great impact that a flood would have for the Dutch society, The flood safety standards are very strict, which means that the required failure probability of flood defences is very low. To keep the flood risk within acceptable limits of the hinterland, flood embankments are assessed on whether the probability of failure matches the stringent norms using the methods and rules which are stated in the legal assessment instrument (WBI). Within the WBI the probability of failure of the flood embankment equals the sum of the probability of occurrence of individual failure mechanism. This thesis specifically focuses on the failure mechanism: sliding of the landside slope cover due to wave overtopping of clay flood embankments. Due to wave overtopping the relative permeable top layer of the clay embankment becomes saturated. This leads to a shallow subsurface flow parallel to the landside slope cover which is unfavourable for the geotechnical stability of the landside slope cover. The stability assessment of this mechanism is currently assessed with the Edelman & Joustra formula. This formula method is not based on a probabilistic or semi-probabilistic analysis and therefore does not give an failure probability. Instead it uses safety factors. Levee managers therefore believe that the outcome is overly conservative (Waterschappen, et al., 2021). This thesis addresses the question: How can the stability assessment of the Edelman & Joustra formula can be optimized for clay flood embankments in the Netherlands within the legal assessment instrument? To answer this question, new semi-probabilistic safety factors were derived based on a full probabilistic assessment of the Edelman and Joustra formula. The outcomes show that the current approach is not always as conservative as is often assumed. Application of the newly derived safety factor allows levee managers to be more flexible in how to account for the contribution of this failure mechanism to the total levee failure probability. The newly developed method therefore prevents unnecessary rejection and over-dimensioning of flood embankments, so the available resources for flood embankment improvements can be used better.

Internal erosion is a frequent cause of dike failure, also known as piping or backward erosion. Uplift is considered a submechanism of internal erosion, together with heave and piping, where all three submechanisms must occur for the dike to reach a failure condition. The principle of the uplift phenomenon is straightforward: it occurs when high pore water pressures in the aquifer lift the cover layer, which is located at ground surface. If the pressure is great enough, the cover layer begins to float ('opdrijven'), and may also bulge ('opbollen') or crack ('opbarsten'). Currently, the assessment for uplift is based on a vertical equilibrium, which relates to a floating condition of the cover layer. The forming of an exit point for concentrated seepage is presumed to happen simultaneously with the increase in uplift pressure and is therefore not explicitly considered in the current assessment. This thesis aimed to investigate in what way the uplift assessment may be improved, both to describe the complex behavior better and to decrease the need for costly countermeasures. The evaluation of uplift was conducted in two parts: by making explicit what the causes of uncertainty are in the current assessment, as well as by proposing a new model to describe the complete uplift behavior better. It is recommended to evaluate the ‘floating’ and ‘bulging-cracking’ phases directly. In the ‘bulging-cracking’ phase, the forming of an exit point can either occur by a tension or shear failure.

Wave overtopping is responsible for many dike failures, dike breaches and severe flooding. The erosion resistance of dikes to overtopping waves is determined by the strength of the grass cover. Therefore, a thorough understanding of the erosion resistance of grass covers is desired to enable a good and effective dike management. In the existing approach to assess the erosion resistance of grass covers, the erosion resistance is empirically quantified in terms of local hydraulic loads. However, despite the cyclical nature of overtopping waves, no insights into the dependence of the erosion resistance on successive loads are available. To improve the understanding of the resistance of grass covers to wave overtopping induced loads, understanding of the development of the mechanical behaviour of grass sods is essential. In this study the mechanical behaviour of grass covers under tensile loads have been examined by means of direct pull tests executed in the field. Especially the influence of repetitive wave loading and the conditions during testing were of interest. The results of this study show that the deformation of the grass cover due to tensile load is time dependent. The failure mode of the grass cover is influenced by the degree of saturation of the grass cover. The maximum strength and mechanical behaviour of the grass cover is influenced by accumulation of damage. The findings presented in this study provide a better understanding of the processes that influence the mechanical behaviour of the grass cover. Moreover, insights in the development of the process of damage accumulation are revealed. Based on the findings of this study, a method is proposed to determine the resistance of grass covers against repetitive loads by means of direct pull tests. In this method the resistance of grass covers against repetitive loads is described by an estimation of the fatigue limit. Relating the fatigue limit of a grass cover to hydraulic loads, enables an assessment of the erosion resistance of a grass cover at a specific location.

CSD spillage is defined as “any soil that is dislodged above the lowest cutter tip trajectory of a single swing, but is not sucked into the suction pipe”. In addition to higher energy consumption and material wear for delivering the targeted depth, spillage can lead to a variety of environmental issues. As of yet, no analytical model exists in literature that can estimate spillage rates for a given set of cutting parameters. This thesis presents the Sand-Rock Cutting Spillage Model (SRCSM), an engineering model that is particle size-agnostic and makes use of cutting parameters that are all available to the dredge operator.
Prediction accuracy of 5 percentage point is achieved with a two-disc potential flow model complemented with empirical closing relationships. A triad of forces governs flow in the cutter head for typical cutting conditions: a centrifugal, suction and gravitational force are considered. For the centrifugal pump effect, and centrifugal pump effect only, the flow inside the cutter is considered steady, non-gravitational, inviscid and non-axial. This allows for the derivation of a pressure-discharge affinity law from the Navier-Stokes. The axial pump effect is governed by the mixture velocity at the suction mouth. It is hypothesized that the pressure difference over the discs drives an inflow at the disc closest to the nose. Centrifugal advection and rapid redeposition spillage are considered the two most significant spillage types out of the six classified. Centrifugal advection can be determined by identifying the onset of radial outflow at the disc near the cutter ring. The magnitude of rapid redeposition flow and its concentration depend on mixing effects that are proportional to the ratio particle settling velocity and the mixture velocity squared. The model is calibrated with three coefficients. User input parameters are the cutter geometry, cut-type factor f_d,type (-1 for under-cut), bank slope angle ξ, cutter inclination angle λ, bank height h, step size l_step, rotational velocity ω, settling velocity v_ts, swing velocity v_s, mixture velocity v_m and material densities.
Spillage=f(D_ring,D_nose,D_pipe,b,f_d,type,ξ,γ,h,l_step,ω,v_ts,v_s,v_m,ρ_q,ρ_b,ρ_w )
For calibration, an inverse flow number θ^-1 is used that is proportional to the ratio of centrifugal flow over mixture flow. Spillage rates from SRCSM are in high agreement with reference data for sand (Miltenburg, 1983) and rock (Den Burger, 2003) in an under-cut swing. A sensitivity analysis suggests that most cutter head dynamics are adequately incorporated. The model is less reliable for (non-typical) inverse flow numbers of  θ^-1= 6 [-] and higher due to a mixture velocity that drops below zero. In addition, the model is calibrated for a relatively high cutter inclination angle of 45 [deg] and bank angle of 45 [deg]. Caution should be observed with the results. It is also suggested that mixing effects related to the swing velocity are incorporated more explicitly in the model. For typical sand cutting conditions, the highest spillage reduction (-4.6%) is achieved by a 1 [%] smaller step size. For rock, the highest spillage reduction (-0.63%) is achieved for a 1 [%] decrease in swing velocity. Spillage appears to follow the theorem of Ellington (1934): it don’t mean a thing if it ain’t got that swing.

Dikes constructed of a sand core, clay cover and natural vegetation as (landside) slope protection are common flood defence structures in the Netherlands. The design parameters, such as the crest height of dikes, are determined through risk assessment to reduce the remaining probability of a failure mechanism, such as overflow. Little is known about the exact process leading to failure of the typical Dutch dikes in case the failure conditions actually occur. The most notable floods in the Netherlands have all been caused by dike breaching, where local damage in the dike lead to breach formation. This research describes the process of breach formation due to overflow and determines what design parameters affect the resistance of a dike against breaching. The proposed model (BRAM, a Breach Resistance Analysis Model) combines existing methods to determine the time-to-failure of the grass and clay cover with a newly developed process-oriented headcut erosion model. The model assesses the total time-to-failure of a dike through sequential failure of its components. The following process is simulated. Overflow over a dike causes a turbulent flow along the landside slope, resulting in local damage to the grass cover on either a prescribed weak spot or slope transition. The clay cover is subsequently eroded and the granular core material becomes exposed to the flow. A key assumption in the applied method lies in the fact that every newly exposed layer is less resistant to erosion than the layer above, resulting in slope steepening. Therefore, a near-vertical cliff, referred to as a headcut, is formed. When the overflowing water no longer follows the profile of the steep slope, a cascading flow, resembling a waterfall, forms a scour pit. The turbulent flow in this scour pit also erodes the core material of the dike, allowing horizontal erosion to undercut the slope. The undercutting of the slope causes the headcut to tumble over, moving the headcut and jet impact point to move towards the waterside slope, essentially restarting the scour process. This iterative process continues until the invert height of the dike is lowered to zero and the overflow rate is increased further. Model validation shows good agreement with test results on both large and small scale, as long as the assumptions regarding the structure of the dike hold (a dike profile consisting of a granular core under a clay layer and grass cover). When applied to three case studies, the BRAM-model predicts a similar total time-to-failure as the most advanced current model, posed by d’Eliso (2007). The predicted time-to-failure for the grass cover shows a significant difference. The BRAM-model predicts the grass cover to fail significantly faster than the model by d’Eliso (2007). This difference is offset by a longer predicted time-to-failure of the clay cover. Finally, the proposed model predicts a 50% slower headcut migration rate, as compared to the results by D’Eliso. This is a result of the more advanced description of the headcut erosion process. As the time-to-failure of the headcut migration phase is relatively short compared to the total breach formation process (between 10% and 25%), the total predicted time-to-failure is quite similar between both models. The main advantage of the BRAM-model is its suitability for design testing. The main weakness remains the unpredictable location of headcut initiation, which depends too strongly on spatial variation of the resistance of the dike to be predicted accurately, based on available data. Design parameters to increase the time-to-breaching have been identified. The most relevant were the steepness of the dike slope and thickness of the clay cover. Sensitivity analysis showed that the steepness and thickness of the clay cover, but also the porosity of the core material can be adapted to increase the time-to-failure of a dike. A breach-resistant dike may be designed for a limited overflow duration. Case studies showed three test dikes to sustain an overflow rate of 100 l/m/s for various hours by the grass and clay cover. As soon as the core material becomes exposed, the breach formation process accelerates significantly. The exact duration depends on the profile and construction material of the dike.

Coral reefs are degrading across the entire Great Barrier Reef. Rehabilitation of the Great Barrier Reef is crucial for Australia, both socially and economically because it provides $6 billion in revenue and 63,000 jobs. The Queensland state government issued a challenge within the "Small Business Innovation Research" program to quickly restore ecological functions provided by the reef. Van Oord and CSIRO participated in this challenge by testing a concept designed to scale-up rehabilitation efforts. This concept involves rehabilitation of the reef by increasing the number of larvae deployed onto the reef. To achieve this, coral-spawn slicks should be collected from areas with healthy coral populations that show a redundancy of embryos. During transport, these slicks should then be stimulated to grow into settlement competent larvae. Subsequently, they can be deployed at a degraded coral reef. The collection phase should be executed with pumps to reach considerable volumes of coral embryos for significant ecological scale. Unfortunately, the mortality rate of coral embryos due to pumping is so far unknown, yet crucial to determine. Therefore, the objective of this research is to develop a method for designing a pumping system that can pump coral embryos and larvae into a vessel-based container or hopper with maximal survival rates.
To achieve this objective, the strength of coral embryos and larvae as well as the stressors in a pumping system are investigated. In order to design and optimise the pumping process for survival, it is useful to have a framework that estimates the balance between strength of coral embryos and larvae and stressors. In literature, comparable strength and stress balances are known in, for example, thermal stress onto corals. The cause of failure (or mortality) is a combination between the strength, the exposed stress (or load) for a certain period of time and the capability to regain strength. Failure happens when the strength cannot withstand the stress (or load). These failure effects are also described through threshold limit values (TLV) which indicate a limit that the stresses may not exceed. Examples of such limits are time weighted average, short-term exposure limit and ceiling limit. Within the design of a pumping system, these values are expected to be acute (e.g. in the pump) and chronic stress (e.g. in the pipeline). In order to distinguish between pumping systems, the stressors in the system can be calculated by using tools such as a mathematical model.

The coral life cycle commences with gametes that contain eggs which are fertilised in the slick at the ocean surface, before developing into embryos. Because the buoyancy of coral embryos is positive until 36h following spawning, dispersion is limited. Hence, they are easiest collected within this period of time. The pumping related strength of coral embryos and larvae is therefore investigated in this research. The assumption is made that the strength of the embryos and larvae differ in the first period following spawning. In order to gain insight in the strength limits of coral embryos and larvae within this period, strength tests are executed. During the mass-spawning event in November 2018, fertilised eggs were collected and used in a Couette-rheometer test at the Heron Island Research Station. These tests consist of a cylinder rotating at different speeds in a container filled with water and embryos. Due to the rotation of the cylinder, the coral embryos and larvae experience stress. The living coral embryos are counted before and after each experiment by use of a microscope experiment in order to define the survival rate. After 5-7 hours the embryos have developed a considerable strength (larger than can be applied by the Couette-Rheometer). Therefore, it is recommended to start the collection process after 5-7 hours following spawning. Nevertheless, the tests do not define the exact strength of the embryos. Hence, it is advisable to investigate the exact strength in further research.

To design a pumping system that can pump coral embryos and larvae with low mortality, the stressors in the system should be minimal. The main technical criteria of the pumping system include low shear stresses, low-pressure fluctuations and low flow accelerations. Practical criteria such as availability, scalability and handling should also be considered in the design. A mathematical model was designed to calculate the previously described factors by calculating the magnitude of the stressors in a pumping system. Based on the model, it can be concluded that the pipeline configuration of a coral slick collection system should contain minimum pipe length, maximum diameter, minimal surface to encounter (e.g. bends and connections), low rpm and flow velocities and submerged in- and outflows openings.

Laboratory tests in the Netherlands and field tests in Australia have been executed in order to validate the relation between the strength of coral embryos and larvae and stressors in a pumping system. During laboratory tests, pumping system aspects that contribute most to damage have been investigated, and potential practical issues have been identified. Damage rates were estimated for different pumps and pipeline configurations using different proxies such as, hydrogell balls, peas, berries and fish eggs. From these tests, the Hidrostal and Diaphragm pump resulted in low damage rates and were selected to be applied during the field tests. Additionally, a skimmer (intake) to collect the floating particles was designed and tested, both in the laboratory as well as the field.

The field test has been executed in the Southern Great Barrier Reef around the Heron and Wistari reef. Currently, this part of the reef boasts the highest level of coral cover throughout the entire reef, and is almost at its historical maximum. Therefore, it was chosen as research location because of its expected large supply of coral spawn. The main goal of the field study was to investigate the possibility to pump coral embryos following the mass spawning in November 2018. The experimental setup consisted of a tug vessel with two pumping systems, each with a different pump (i.e. the Hidrostal and Diaphragm pump) and the same configuration. A total of total twelve tanks was used for cultivation of which six plastic and six steel. The total number of living coral embryos that have been pumped was approximately 29 million from which 19% developed into competent larvae within five days. The competent larvae were pumped through the same system to investigate their survival rate, which was around 88%. This indicates that deployment of larvae onto degraded reefs should be possible by pumping without much loss and that coral larvae are less fragile compared to coral embryos.

The objective of this research was to develop a method for designing a pumping system that can pump coral eggs, embryos or larvae, from the sea surface onto a vessel housing an aquaculture facility. By using the previously described criteria, it is concluded that coral embryos and larvae are able to survive pumping related stressors. Further research should aim to find the exact strength of coral embryos after which a more precise model can be created to estimate mortality. Furthermore, limitations for pressure differences should be determined. For now, the assumption is made that for designing a pumping system, atmospheric pressure should be the minimum local pressure which is conservative. Additionally, the controlled deployment of the competent larvae onto degraded reefs should be further investigated because this project was not focused on that phase. To conclude, the project needs up-scaling towards hopper sizes, with a broader range of coral species and various weather conditions.

Given the previous described challenge, this research proofs that it is possible to pump coral embryos, after which they can grow into competent larvae and be deployed onto degraded reefs. Significant quantities of new recruits can be collected at healthy reefs and by using hopper size vessels, the collection of coral embryos from slicks become promising.

In the Netherlands flood protection is immensely important for the safety of the nation. The shocking outcome of the 1953 flooding proves this point. In modern days, the development of socioeconomic and climate change factors casts doubt on the effectiveness of conventional approaches to flood risk management. Consequently, this project explored new approaches to flood risk management.
An analysis of climate change effects led to estimation of future loading conditions. Subsequently, a detailed hydrodynamic analysis was conducted. It highlighted the significant levels of uncertainty that climate change introduces into loading conditions. Also, it confirmed the team’s perception, that the Westkapelle case region requires additional safety measures to guarantee an acceptable level of safety in the future. But how to guarantee the acceptable level of safety in the most efficient way? The team adopted the concept of robustness to find an answer. In a keynote publication Mens (2015) describes robustness in the following way: "Robust flood risk systems have some degree of resistance and some degree of resilience: the system can withstand some floods (no response), and for other (larger) floods impacts are limited and the system can recover quickly from the flood impact (response and recovery)." The team set out to include robustness as an integral part of the design process to handle uncertainties. The project shall be seen as an explorative study how this can be done, revolving around Westkapelle as a case study that proves the methodology’s feasibility. Robustness and uncertainty were included on multiple levels throughout the design process. Firstly, the range of uncertainties was quantified. Secondly, the effect, that single parameters have on the magnitude of uncertainties, was assessed. Thirdly, the system’s capacity was analysed to find the required overtopping reduction for guaranteeing sufficient safety. Fourthly, constructive measures were assessed on their robustness potential and satisfaction of stakeholder needs via a Multi Criteria Analysis (MCA). The MCA was then employed to select the type of constructive and non constructive measures to achieve the required levels of overtopping and safety. With the information on uncertainties, the measures were combined to form a robust design, consisting of living breakwater, dike heightening, surface protection and two policy measures. Probabilistic analysis was also done to see the sensitivity of the failure probability to sea level rise in different loading and design scenarios. A thorough comparison between the conventional design, that has been applied to the project location, and the robust design followed. The robust design came out on top. Robustness was found to be an effective tool in countering uncertainties. Where conventional design methodologies are lacking flexibility and precision, the robust design methodology makes use of the system and its resilience to find an optimal solution. Its applicability may not be limited to flood risk management only but stretch out to other civil engineering disciplines.

A growing risk of flooding in coastal areas and the corresponding development of legislation drive continuous development in the field of hydraulic engineering. Failure due to
wave overtopping is not fully understood from a physical point of view. Existing knowledge gaps, processes that are not yet explained by physics or not even discovered, are
explained by empirical relations. The fact that knowledge on the relevant processes for overtopping load and soil resistance is fragmented and limited, causes limited application of it in practice.
To improve the understanding of failure caused by overtopping, a model that computes the soil stresses (soil stress is related to failure) in a dike cover, during overtopping load, is developed. It is possible that this model approach leads to additional knowledge on wave overtopping, e.g. new failure mechanisms or shift in normative mechanisms.

It is shown that overtopping has a large variety of appearances and so does failure due to overtopping. Literature study shows the large number of processes and characteristics, all with variable magnitude, that is relevant in the wave load and the soil strength during wave overtopping. Main wave load processes are shear stress (gradients), turbulence and impact. The considered soil strength process is the ratio between the stress (relative to the strength), for which the Young’s Modulus, Poisson’s ratio, soil weight, hydraulic conductivity and the subsoil stiffness are shown to be the main soil characteristics.

The developed model, a numerical 2D model with the model domain oriented parallel to the flow direction, computes the stress distribution and development in a loaded soil. For this case it is focused on dike land side slopes loaded by overtopping waves in particular, that is on dealing with clay, saturated conditions, fast varying load and a sloped surface. Stress equations in the model are derived, based on equilibrium of forces (horizontal and vertical) and motionless soil. The model is able to indicate initiation of failure for failure mechanisms that are associated with soil stresses, e.g. lifting of soil and head cut erosion. Soil in the model is loaded by normal and shear stress. The latter is, deviant from conventional methods, modelled as a shear stress gradient.

Test runs for verification and validation show that the model gives good results for cases with constant loads. Comparison of load cases with non-continuous overtopping wave loads shows both similarities and differences in the resulting soil stress. The most outstanding difference is the magnitude of the stress at depth. The differences indicate
starting points for further research.
Model employment demonstrates the future possibilities. A single wave load, computed by the model itself, is modelled. This run gives, at first view, plausible results.

The current model gives a reasonable representation of soil stress development. Further development is recommended to make the model useful for dike assessment and design. To do so, the model should be improved by probabilistic parameter definition, addition of soil stress damping and considering the impulsive nature and turbulent oscillations of wave load. Furthermore, it can be extended by the comparison between soil stress and strength, addition of spatial variability of the soil, enabling computations of non-planar parts of the slope and enabling the connection with other hydrodynamical models.

This experimental research is focused on fail mechanisms in non-cohesive soil beneath the jet with a Jet Erosion Test(JET). A vertical plain jet impinges fluid to a sand bed, which causes zero shear stress beneath the jet averaged over time. Many erosion theories would say that no pick-up happen at that point.

E_c=u/(√(H⁄d) √(ΔgD50 ))

Large range of erosion parameter was done in this research with video recordings to the activities beneath the jet. Tests varies with those parameters outflow velocities, stand off distances(SOD) and grain sizes. Comparisons are made to Rajaratnam, who has executed similar JET with varies erosion parameter too. The cavity depths and hill heights correspond well as function of the erosion parameter. Cavity widths have deviations for values above 4.0, which might cause by different grain properties or by the limitation of flume width. Secondary flow limits the width growth over time.

Weak deflected jet showed only surface erosion by micro turbulence at the soil bed till erosion parameter 1.0. One smooth bed was formed between two hill tops. Static bed did not exceed the concerning sand during and after start.
Increasing further the erosion parameter (1.0-5.0), soil deformation was also seen with ejection beneath the jet. Bearing capacity by Prandtl's theory was insufficient and failed the soil bed. Consequence, the inner cavity exceeds the internal friction angle of sand with dynamic erosion.
Erosion parameter larger than 5.0 had a wider cavity. Grains in soil translate and rotate in the start phase by insufficient bearing capacity. Vortices sweep around the jet with chaotic particle transport in the inner cavity.

Theories of Van Rijn, Meyer Peter Müller and Van Rhee underpredict for JET. The pick-ups rates are in all cases lower, which are based on the shear stress on bed. The impinging jet has different flow field and failure mechanism, which can be better related to the jet momentum.

This research focuses on the development and feasibility of an online grain size estimator for a trailing suction hopper dredger (TSHD), based on the mass balance within a hopper, and using hopper settling models. The idea of the grain size estimator is to use existing hopper settling models iteratively for a certain time window with ‘online’ sensor input and different soil inputs. The resulting loading graphs are compared to the measured loading graphs, after which the most probable soil type is indicated.

This research is a continuation of the research of Strooker (2017) where not only the mass balance within a hopper, but also different occurring processes were taken into account. The grain size estimator developed for this research, is solely based on the mass balance within a hopper, and is focused on selecting a suitable
hopper settling model, simulating a great amount of trips of various projects and interpreting the results.

The research shows that the Modified Camp model performed well in validating the hopper settling model and was therefore chosen to be implemented into the grain size estimator. After determining the limitations
of the estimator, two projects and two vessels were used to simulate 1500+ trips in total with the available sensor data. The results are presented visually in different ways and compared to the measured soil model
of the considered project. It is concluded that at one project (Liverpool), there are multiple areas where the results of the grain size estimator clearly matched the soil model. The second project (Kaohsiung) showed less conclusive results. It seems that the results of both the first part of each trip as well as the whole trip,
produces results where the areas of soil are best distinguishable. The time window for where the soil type is determined also affects the accuracy and reliability. Longer time windows mean less accuracy, but a more reliable result. The accuracy was found to be only changing slightly between time windows of 30-600 seconds, but the reliability of the model is difficult to quantify.

When dredging close to underwater sand slopes, steep slopes might form. In dense sand with low permeability this might lead to the so-called breaching process. The creation of a steep underwater slope marks the beginning of a breaching process.
Pore volumes of densely packed sand tend to increase during shear deformations. This effect is called dilation. As the grain slides over each other, the pore volume increases resulting in increased underpressures. Water, eventually, has to flow in to compensate for this underpressure. The flow rate depends on sand properties. The underpressure keeps the sand body, temporarily, stable. When enough water has flowed in and the sand has dilated enough, sand particles release at the front. This is at the start of the breaching process. This leads to a density current consisting of sand mixed with the surrounding water, which runs down the slope and might cause erosion. The steep front of the slope moves with a certain velocity, which is called the headwall velocity.

A breach can be stable or unstable. A breach is stable if the breaching height decreases in time and unstable if the breaching height increases in time. This work aims to improve the prediction of the stability of a breach. For this purpose I carried out a series of large scale breaching experiments. During these experiments the initial breaching height, slope angles, and sand types and varied.

Literature suggest that the stability of a breach can be predicted using the headwall velocity and the angle at the toe of the breach. The experiments show that this is indeed the case, but that breaches are more stable than literature suggests.

Therefore, to predict the stability of a breach we must know the angle at the toe, and the headwall velocity.
The experiments show that the angle at the toe converges towards the angle that is predicted using equations found in literature.

A formula to predict the wall velocity can be found in literature. Comparison with the experiments show that this formula can predict the wall velocity when no slides are present, but at large breach heights these slides often occur.

The experiments show a clear correlation between the breach height, and the frequency of sliding wedges.
Empirical equations following from the test data predict the percentage of the sliding wedges at different breaching heights are proposed.

A steady state numerical model, to calculate the pore pressure during a breaching process, was programmed in MATLAB. Using these pore pressures, a stability analysis was carried out. This analysis confirms that breaching height is an important factor for the prediction of sliding wedges.