Intro HEC-RAS Sediment Demo (Part 2 of 3 - Sediment Transport Data)

welcome to the introductory series of videos on how to create a sediment transport file in HTC R as this is the second video on how to generate sediment transport boundary conditions and the said file my name is Stanford Gibson I am the sediment transport specialist at HEC in charge of the sediment capabilities in rasz and in this video I'll walk you through how to create a sediment file for a 1d HC r as model if you haven't done the first part and created a quasi unsteady flow file you can do this standalone if you already know how to do that we provide those files or you can just build on to the quasi instead a flow file you created in part one so a sediment transport model and rasz needs two basic types of data you need bed data and boundary data the bed data the sediment that goes into the bed those are your initial condition what is the control volume and gradation of the sediment that are in your bed at time zero and then the flux data is the boundary conditions at at every kind of upstream external node and so you need to specify not only the sediment mass that comes in at those boundaries or concentration but also the gradation of that material which can be a challenging part the files we've provided under sediment demo2 we have solution files in case you want to go in and see what it looks like when it's all done but the data files have a project file a geometry file and then just the Kwazii and steady flow file that we developed in the last video and then we have sediment data and so we're just going to go and press the open or the file open and say sediment demo two data files and select the project and so you see we have a project file a geometry file and a quasi and steady flow file so next we have to enter sediment data to enter sediment data you go to the menu edit sediment or you press the sediment data button now the sediment data editor can look a little intimidating when you first open it up because it has all these blanks that you may not know how to fill in but we're gonna fill these in pretty quickly and zero in just on the actual data we need and so there are three tabs here initial conditions and transport parameters boundary conditions and then this optional B stem tab that's really only if you're going to do bank filling analysis which will get its own separate video really what we're gonna focus on is the initial conditions which is really the bed gradation in the bed control volume and the equations and then the boundary conditions which is really about the flux at the upstream boundaries or any lateral fluxes that you want to add so let's start with the initial conditions and transport parameters you have to define bed sediment for every cross-section and if you come over here you can see the profile plot or the individual cross section and so each cross-section needs a sediment reservoir or a sediment thickness below the bed and so the way that we're gonna do that is you can either put in a maximum depth which will be how deep the sediment goes below the the thalweg or a minimum elevation a minimum elevation is much less common that's if you say no the bedrock elevation and wanted to find that kind of distinctly in the real world but generally we just put in a max depth um particularly if it's an alluvial channel that can erode without much constraint vertically so we're an SI so let's just start out with a 5 meter maximum depth and that means that will allow all of these cross sections to erode 5 meters we can update that to make it a bigger sediment reservoir if we need it and then the other question is what are the extents that we're going to allow to erode and so this you can go in and define more carefully later but this button right here use Bank extents allows you to define your movable bed limits as your banks and so now if you come in here you'll see this cross section has a sediment control volume or sediment reservoir associated with it full of a roadable sediment and so the next thing we need to do is define the gradation associated with that sediment now this again looks pretty intimidating because every cross-section needs a better gradation there's no way around that we need Begg rotation and from for every cross-section and you almost never have that gradation at every cross-section but we have a couple of options that allow you to interpolate or to associate a big relation with multiple cross-sections so if you come in here to the sediment data you're going to open the tutorial data and this is the flow data we used in the last one if you come down here to the next tab bed sediment we've just defined a single bed gradation Begg rotations are usually defined in the geotechnical cumulative distribution function and so this is a size and percent finer curve you're going to do first is you're going to enter all of the bed rotations you have into this generalized bed gradation database here in R as where it says define and edit beg rotations you'll put all of your bed gradations in here in this case we only have one so we press that button and then whenever you see this new sheet of paper and razzed that's that's where you create something new so we're going to create a new Begg relation and we're going to call it sample one and then I'll just come over here and copy all of these and you're going to copy starting at the top to make sure they go in the right place and then paste them here and you can see we're using percent finer you can't put in the grain class percentage as a discrete percentage of each grain class and you can see will let you convert between them as well but we're going to use the percent finer nomenclature because it's a little bit more common and a lot of times these are the data you get back from the geotechnical lab except sometimes they're not in this log base to scale where there are every grain class is twice the size of the one below it and so in that case you have to set it to fit or you can go in and define your own grain classes right so I'm going to say okay and so now I have a bed gradation and so the next thing we're going to do is we're gonna associate that Begg relation with the cross sections if you click over here in the bed gradation field now you get a drop-down menu and in that drop-down menu it will be all of the bed gradations that you have defined so you can come here and say I'm going to choose that sample there and you can choose another sample downstream and then you can press the interpolate button to interpolate between them or you can choose the sample hover over the the lower right hand corner to get the crosshairs click and drag and associate the same sample with multiple cross-sections now I'm going to come in here and save the data I haven't saved this yet so it's gonna prompt me for a name and I'll call it sediment data and we're on our way so you can see that even though this sediment data editor comes up with a lot of blanks and there might be some cognitive resistance to that we can fill them up pretty quickly and then go back and work on them as we develop the model really there's only a couple other things you have to do here and these are some of the biggest decisions you're gonna make you have to decide on your transport function which is the equation that you're gonna use to kind of turn your hydro dynamics into sediment transport potential or capacity and you have to choose your sorting and armoring method which is the bed mixing method that determines whether or not you're going to develop an active layer that will reduce your erosion will have to deal with these in a different venue or a different video because these are relatively complex ideas but for now we're going to select the Larsen Koplin transport function and the Copelan or Exner 7 mixing method why am i choosing those well because I know that they're going to work on the system a priori and let me just say again in this video that this system does not exist in the world it's completely contrived and should not be used for research or planning purposes now all of my initial conditions and my equations have been selected what we need to do now is define our sediment boundary conditions and that's how is our sediment going to change in time we basically need a sediment flux and a gradation of that flux for every time step but we almost never have time series of sediment data the way we have time series of flow data so there are multiple ways to do this if we go to boundary conditions you'll notice that we require a boundary condition at the upstream boundary of our model and if we go and open the geometry file you'll notice that it's a this is a single reach and so we only have one upstream nodes so R as only populates one required boundary condition so you'll notice there are three possible upstream boundary conditions you can define a sediment load series which is just a time series of sediment data that you can put in manually or bring in from a DSS file this again is pretty rare but sometimes you know the USGS will develop a sediment time series for your gauge based on sediment data or rating curve or sometimes you've got a sediment wash-off model like SWAT or HMS and you can develop a DSS time series and bring that in the other option is equilibrium load this is a pretty popular option and is frankly way more popular than it should be it's very easy because you don't have to input data and it just computes the equilibrium flux at the upstream end but if you're doing a sediment model it is unlikely that your system is in equilibrium and this can introduce a lot of data the most common more credible method is to define a sediment rating curve at the upstream boundary condition that will look at the incoming flows and compute a sediment load for that upstream boundary condition now I'm going to do an entire video just on developing a sediment rating curve there's some statistical nuance to this but let's just say that we've got one now and we're just going to talk about how to put it into your model so we're gonna select rating curve what you see is we get a rating curve with actually a lot of rows in it it seems like there's more rows than there should be well here we're gonna put in our flows and here we're gonna put in our sediment loads but then that will give us our mass flux and you'll notice down here you can do this either in load or concentration but that'll give us our sediment flux usually it's a power law generally this is some sort of power function you know the the relationship between load and flow is usually load as flow to some power usually around to or between 1.5 and 2.5 but then you're going to have to break that flux down into the different wrinkled which is where this gets particularly challenging now you can define any number of sets here and you would want to define more than two sets if you have a flow load relationship that isn't linear in log space let's say that it's supply limited so at the higher flows you get less load or maybe vice versa or if the gradations don't change linearly and so we're going to open the sediment flux boundary condition and here we've actually defined for flow load rotations and so I've actually included the power function here I used to generate my flow load curve and so this is the flow lower load relationship in these top two rows but then our sediment flux has to be broken down into the individual grain classes so for each of these flow load pairs the loads are then broken down into the percentage or fraction of each grain class now you can pick these in as percentages or you can put them in as decimals it doesn't matter to R as whereas is gonna add them up and then divide in order to get an actual true ratio they could actually just be masses but you can see that here we have basically forty percent very fine sand for the smallest flow load pair and then we have sixty five percent very fine sand for the largest flow load pair and so a valid question might be where will you come up with these data well sometimes there are load gradation measurements associated with gauges or studies and so you can look at how sediment changes with flow does it get coarser does it get finer um which would you expect would you expect sediment in general to get coarser or finer with flow it's actually a pretty interesting and complicated question it turns out that it could do both in some cases as flow increases the the sediment load gets finer in other cases as flow increases the sediment gets coarser I've included a link on a paper Weaver and on this topic that actually we'll talk walk you through the different mechanisms that could cause it to but in this case we've got something that works for the system and so we'll just copy this and then paste it and you'll notice that I actually went in and I selected all the cells that I wanted to paste into before I paste because rasz doesn't always work like excel where it will automatically populate everything out you actually have to sometimes select the cells before you paste into them and then you can ply your rating curve and because we just used the power function here it's linear in log space which of course means that the higher flows have disproportionately higher load which is the way that rivers work right we say okay we go to file save and now you'll notice we have a sediment data file here and our sediment data are complete and in the next video we will run this and then look at some output and then think a little bit about what those output mean and we'll just have a couple of comments about some of the things that go wrong and some of the things that you can do to improve your sediment transport modeling at this point I'd like to recognize that these videos have been supported by the flood and coastal storm damage reduction research and development work unit of the Corps of Engineers and a lot of the sediment transport features in HEC Raz were based on the sediment transport features in a previous program called HEC six which was developed by Tony Thomas and much of the technology that we built this program on is based on his decades of work and insight all right we'll see you in the next video

Channel: Stanford Gibson

Views: 16,301

Keywords: HEC-RAS, Sediment transport, sediment data, mobile bed, sediment modeling

Id: 9YiL3Men9as

Length: 14min 54sec (894 seconds)

Published: Thu Jun 06 2019