 Okay, so our next talk is the first of the student talks in the morning session going to be given by and the plan was turbulence resolving or layering a two-phase model for coastal sediment transport applications. It's a walking truck. My name is Vincent and today I'm going to introduce the Terminus as well for you in the early two-phase model for coastal marine transport applications. And this is a trail that's much more than the previous trail, so this one we found with my writer council and we are collaborating with Junior Tutor and Dr. Tibun Kibir-Bek and this work is founded by NSF and ONA and all the latest images are done in the Indian SPC Community Center and XT. So why do we tell so much about coastal sediment transport? It's because the coastal ground is a very important human habitat of highly industrial diversity and creating their economic importance. According to the survey in 1995, over 38% of all coastal glaciers are in the Indian rivers of the coast or estuaries. However, due to global climate change, the sea level has been rising rapidly and this actual rising sea level rise has made the coastal ground more vulnerable to a massive hazard such as storm surges. As we all see here, the sea level and the earth level show an example as a low sea base up ahead in Sandy. We can see that a lack of sediment are really caused by a hurricane and this leaves the houses along the beach at risk. So what is really helpful is essential for this erosion and recovery. However, good evidence shows that mechanisms created in the major storm conditions from now will characterize in the state-of-the-art sediment transfer motors. The state-of-the-art sediment transfer motor is usually based on a single base of coasts which the sediment transfer is divided into a dead load and a sustained load. So the dead load is not a result of condom rising in the imperial formations and a sustained load also will have a peak of foundation from the bottom which is also a pretty empirical. So we prefer the two-place model to study sediment transfer. We spread the two-place model equations with closer interphase modern transfer practices, total experiment in wagons, so we can get a full transfer profile from the dilute transfer all the way to the immobile depth. So the conditional dead load sustained load is not a result. Recently, we have developed a total average two-place model which is called and this model is assembled to the CSLIN as model repository. Now here we have hosted CSLIN as a clinic and it will be interesting if our team can find the intro video from this website. So this model is very, very useful to determine from down here on the lines as we can see from the concentration, comparison with our model is doing a very good job and this model has been used by many researchers on different applications. Here we show one example of the scour, however this is a time limit, I cannot go into detail of this example. Hopefully you've come to me after talking. So now that we have a total average model working, so what is near being this total reserve model? In the total average model that the closure of total spending in action is still highly imperative. In our third machine flow, the total average model works well for the mediums end or to the cost ends. However according to the experiment of the domain answer in 2001, the one that in terms of the transfer near sickness, the fine spend is, the steady load or fine spend is much higher than the mediums end and this feature is cannot be captured by the existing models including the turbulence average, turbulence model. So this is a little advantage over fine spend. All of the processes is that the various savings in actions are critical for fine spend and in typical wave convenience in coastal environments, the flow is transitioning to the end and is present during the flow processing. So that's why we are more afraid to develop this turbulence resorted model. That's what we call the Uralian flow phase flow equations. In the yesterday's student talk by Anders Stamskak, they usually are very insisting of course that the motion of these particles are resolved. However, we noticed by a larger scale model, we actually average over the particles and get a continuum picture. So in this free phase flow, both the fluid phase and particle phase are treated as a continuum and they are solved separately by mass conservation equation and the motion equation equations. The fluid momentum and sediment momentum are copied by a jet force and the fluid stress are more of a result of the turbulence equation and the particle stress is due to the collisions and collisions. More of a result of the planetary and the friction stress model. So by turbulence resorption, we actually solve these two phase equations in a 3D domain, which is large enough to capture the largest eddies and we also use a very fine-grained revolution to capture a lot of the emotions. So a lot of eddies and structures are directly resolved and the effects of the small eddies and structures on large eddy motions are more of the separate motions. It's updated in seven and goes into two major components. The first component is due to the unresolved separate eddy motions, which we can model with a shear stress-like model and the separate eddy viscosity is used here. And the coefficient as is behind me in using a dynamic procedure, we use a very thin structure for both fluid phase and sediment phase. The second component is the solution to the jet. It is reported that there are metal structures of sediment particles, such as stringers and gases. And these metal structures may not be well resolved by mesh size. However, they can have a dramatic effect on the overall sediment dynamics. So the effect of unresolved metal structure can be more than with a sub-rejective reaction. And rotation k here is dependent on the mesh size and sedimentization. And it's also become in the unit dynamic procedure. So this model is further developed by using electric experiments, which in the shear flow is the Schener flow. This experiment provides us a very valuable data set of the leading two component velocities, string-like velocity and vertical velocity, as well as the sedimentation. The flow is a unit direction flow with a bottom shear velocity of 5 centimeters per second. And the water depth is about 21 centimeters per second. There are factors, or light factors, which are supposed to be density, 1.19, and the mesh size is about 3 millimeters. The measure of certain velocities is about 5.59 centimeters per second. And the long dimension of bottom shear stress is about 25. So we carry out a 3D simulation with the virtual dimension is very close to the experiment set-up. And we determine the stream-wise and spawn-wise direction dimensions and the bridge-reversing by following the velocity, fluctuation, correlation, and anspecial. So here is how our 3D simulation results like. Here we show the control of the sedimentation, which is larger than 0.08. And the vectors here show the velocity, vectors, and this plane out. So before we get too excited about the result, I want to emphasize that we actually carry out a very careful vibration in this model. By comparing the ensemble average process with this of the stream-wise velocity profile, sedimentation profile, iron mass of stream-wise velocity, and the wet velocity fluctuations, and also the layer stress distribution. From the layer stress distribution, we come to the conclusion that the flow condition in our model is similar to the experiment. If we look at the velocity profile and sedimentation profile, we are doing a very good job in sedimentation. It's a model that is about 0.08. We even tested a linear light distribution of velocity. In this region, the particle fluctuations are very important, and we are doing a very good job in this region. However, when the sedimentation is lower than 0.08, we underestimate the velocity profile and underestimate the sedimentation as well. If we look at the current resistance in the stream-wise and the water direction, our model predicts a very similar magnitude of the RMS of stream-wise velocity fluctuations and velocity fluctuations. However, we are overproducing the work of component and underproducing the stream-wise component. It is well established that the sediment existing of sediment can dampen the performance, and this statement can be quantified by the reduction of the concomitant constant. The concomitant constant can be obtained from the velocity profile and our removal result is about 0.2. This is well close to the point dimension data of 0.2 to the 3, and this value is significantly lower than the critical value which is about 0.4. The sediments induced by the effect can be usually quantified by the density sediments, which is quantified by the gradient richest number. Here we show a comparison of our removal result and measure the equivalent, which are a very similar magnitude. However, we found another mechanism that can dampen the flow torvents, which is the drag induced torvents sediment effect. And this sediment effect is much more significant than the density in this kind of flow, condition, and sediments. One advantage of the torvents result in more is that it can give us an instantaneous picture of torvents sediment in action. One interesting feature is the prevention concentration, which is defined as belonging that the heavy fibers will proliferate in red in the means of low resistance and high stream rate. And this we can identify by the Q value, which is the excess of the obviously negative to the stream rate. Here in this figure we show the coherent structure identified by the axis of this Q equals to 250, along with the plane path of the sediments condition, and the mean condition is about 22. We can clearly see some of the sediment clusters. So this just allows us to move the sample in a jaded direction. We may see the problem, which has the blank out of the sediment concentration. We found that the question Q is finally correlated with the low sediment concentration. This is exactly the prevention conversation. And the similar problem has been reported by the previous study of fine cement. Another interesting problem is that it's immediately caused by the problem notions. According to Orless 2016, they divided the total notion into four regimes according to the sign of the humanized velocity fluctuation and virtual velocity fluctuations. And the measured data showed that the dead in the means is highly correlated with the ejection event and the stream event. Here the picture shows a 2D control of sediment concentration in terms of virtual elevation and time. And the black story told here shows the dead matter. We can see that wherever the ejection events going on there's an increase of the dead matter. And we can also see that the dead erosion is highly correlated with the stream event. We can have a similar analysis and found a very similar feature as the measured data. And however, we can notice some difference. For example, we under predict the sediment concentration in the lower condition part and we'll also predict the more frequent dead in the means in the measured data. So no work has been done in this part. And here's the conclusion. We have to have the turbulence that's open to a 30-year-old motor. And this motor is likely to be in the next stage of the experiment. We found that the dead in this condition is more significant than the circumstances of the patient in the next experiment. And this motor is able to capture the sediment in the perversal condition. However, more investigations have to be done. For example, we under estimate the stream mass growth piece and sediment suspension. And the inward and outward interaction events are under predicted, although it's not reported here. And more chronologically, the dynamics of that in a minute. Thank you. We have time for, I guess, about one question over here. To the children's part. So it was not done to me about what was mentioned, eh? Second part only. You know, according to data, don't the human transport the preliminary condition of, uh, the preliminary condition at the bottom, do you have, like, no full condition at the bottom, for example? Yes. We have, uh, low flux and no, no motion at the bottom boundary. So there's a many of the mobile deaths in our simulation domain. And, according to that, we under estimate the same association. I think we're not doing a very good job in the subtly part of some of the drag interaction part. So maybe I would use the most sophisticated subtly drag motor to test whether there's turbulence in the interactions. Okay, great. Thank you.