Riprap sizing in HEC-RAS version 6.1

[Music] welcome to the australian water school the home of demand driven industry design training for the global water sector hello and welcome to today's australian water school webinar covering riprap sizing i'm your host craig price um i'll also be a presenter on today's webinar and i'm joined by stanford gibson today welcome from us here in australia to wherever you're coming to us from around the world today i'm going to be presenting a couple of things about rip rap sizing but we are thrilled to have stanford gibson on board uh stanford you can turn on your camera and say hi we're just delighted to be able to have the developers themselves this new feature that is uh coming to us now in heck braz 6.1 which was just released a couple weeks ago and we'll have stanford to walk us through that yeah for just trying to say hello just let us know how long have you been working on this little tool here hi craig well you know i think that david may who's also going to be on he had this idea he'd been doing he had spreadsheets to do all this but he figured you know if we want to get young engineers following the process without spreadsheet mistakes that maybe we should just get it in raz he proposed that a couple of years ago and uh we kind of banged it out in about 18 months all right well that's uh yeah i've heard that it's coming for a while and i've been uh yeah eagerly anticipating seeing that come up and so i yeah we're very thrilled to be able to share that with you today but what i'll do initially is step through the background and the title has got australia in it i notice from the spread around the world that we've got attendees from all over and uh this is the australian water school though so this uh this is something that will take a bit of a focus on australia but it does apply everywhere else because i don't care where you're at uh my guess and challenge me if you've seen anything other than american references but almost everything that i've seen around the planet comes back to uh some of the us guidelines and so we'll come back to the same sources stanford have you seen any other riprap sizing methods from other countries that trace back to other places other than the us that's a good question in 6.2 we're actually going to be adding a second method the maricopa county method but uh arizona's a state not a country we're a federal agency so we're kind of invested in things here okay good um so again my point in that is that um that what you see here and as we trace back the ancestry of the australian rip rap sizing approach is uh anywhere you're coming to us from around the world i think you'll get to the same sources now just a quick summary of the poll results thank you for filling those out uh we've got commercial and consulting uh coming out on top uh hecarim has users obviously are tuning in to see what's new in hekaraz i have asked this question here doubling the velocity increases the required rock weight by a factor of what uh is it 2 4 8 16 32 or 64. thank you for answering those we will give you the answer at the end of this session is the correct spelling grip wrap or rip-rap i just put that in there for fun and i've seen both but that is correct that in most uh cases it's riprap a single word the next uh result here channels bridges culverts or spillways or other structures it kind of evenly spread but uh most uh focused on channels which is good because the riprap calculator is focused on channels so i'm going to start by tracing the australian methods and i'll dig through the pros and cons of using velocity or shear and then i'm going to set up the background for the core method which stan will step us through at the end and it'll take us through the calculator that automates that process what we've got here is hydraulic forces how they act on a rock how big is that rock um this is going to resist movement what size distributions we use and what side slopes should we size things differently for different applications like dams and spillways bridges covert outlets bank stabilization you know should those all be have their independent sizing methods so uh before we move ahead though i'll refer you back to a previous webinar we did last year it was number 100 called rockin it um we stepped through the basic considerations for riprap at that point um i'll provide you a link to that one here have a look at surfacewater.biz riprap following this webinar you'll see a link to the accompanying resources on that website it will also include a download link for the paper that this presentation is based on and it's called the advancing australian rip rap sizing approaches that paper started with some initial research for rio tinto those results were published in last year's international mind water association proceedings i'll provide a link to that as well it was expanded and then submitted to the hydraulics and water engineering conference which was postponed due to covid it was then cancelled in favor of adding uh the approved papers from that session to a hydraulic session at this year's hydrology and water resources symposium when it was first uh put together it was just tracing the osro's references um you see the links here um in in that accompanying website to these manuals uh but then when we submitted it to the hydrology uh conference i've added the australian rainfall and runoff references as well so austrades and ar and r both refly back on u.s methods as we've discussed australian and american rock sizing have a lot of similarities but before we get into any debate between australian and american rock sizing we've got to go back to german versus french rock if there is such a thing and so let's take a look at this on a map this is my favorite view of the world where did the current australian publications come from we've got to go back to europe to just before mozart's birth the story starts in germany back in the 1700s we'll go to france in the 1800s it'll take a layover in russia in the early 1900s then we go out to california for the development of the american west in the mid-1900s and then over to australia in the late 1900s so that's the journey we're going to take today these are the references that we're going to be tracing in the austro's manual we have different applications there are tables and charts which generally converts uh velocity to a rock size so you look down here you find your velocity and come across and you'll get a rock size sounds easy enough but this is a rabbit hole um in australian rainfall and runoff we don't just have uh tables and charts actually we don't have any tables and charts there's just one equation from bob keller's uh open chapter and apparently he had an accompanying rip-rap sizing document that was in earlier versions but was excluded from the most recent one so we've got this one equation a handful of references to state publications and one reference to ostroads now this if we're looking at applications we can break it up a number of ways i'm going to use these four which is just could be channels bridges culverts and spillways these charts obviously there's a lot of information here that we're not going to be discussing but i've put these charts in here just so you can see what's in the accompanying paper so for number one um channels if we look at post-around rainfall and runoff there's just a single unsighted equation and then we have this chart and that chart though is attributed to uh reuven who doesn't exist actually but the title of his paper is actually one by cohen simons and lee uh but it doesn't actually include this chart that chart is uh from some us bureau of reclamation data that traces back to a grad student named nicholas berry who has an unpublished university of colorado masters thesis that just took advantage of six methods here and took an average of them and we'll dive into the history of that one shortly when we look at bridges we've got this kind of circular reference here back to ostroads which refers to this table which is fairly well known in australia but it can be converted into a chart for riprap sizing but again the thing to keep in mind and this is why we've got stanford on today is that this traces back to a california method that's been superseded by the army corps of engineers method likewise now this is a big chart for culvert riprap we've got a bunch of methods that are cited here that are intended for natural channels but for the design charts they all fall back on a uh on these same documents here we've got a tremendous compilation of resources from catchments and creeks but again they all refer back to some of these documents here and trace back up through the california document that has been superseded so that's again why we have stanford on for the fourth application that we talk about in that paper is spillways and other structures and for that we've got some references to some unsighted tables in australian rainfall and runoff and then some the same tables here which again have been superseded by what you're going to see from stanford today so let's go back in time we're going to have a quick little history lesson here um christmas 1717 have a look at this inundation on the german coast um it is uh if you if you zoom in on uh this plate here you'll see this is a depiction of a breach here that's a through break breakthrough it's a breach of a dam and it says it gets inundated from the inside so we've had this one here with 14 000 fatalities so you can save the little baby moses imagery here big ships coming and plucking people off of buildings and off of rooftops and out of the water a lot of cattle couldn't be saved and there were huge economic damages the rulers wanted to make sure it didn't happen again so they commissioned albert brahms same name as the well-known musician but this was well before his time the lord of the territory appointed brahms to write this massive uh guide on vasabao which is uh hydraulic engineering and uh this this word have a look at this one this is kunst which is the art and as we'll see today um hydraulic engineering is still a bit of an art um here's some of the artwork in his publication and what i wanted to get to here was this uh term called the cubic quadratic now when you look at uh what he's got here in his report he lists uh w which is the sharing the weight um the head which is the velocity and then he says that varies by the cubic quadratic so three times two as an exponent gives you six and what we get is w being proportional to velocity to the sixth now if w is proportional to velocity to the sixth then we take a stone in this case a round stone take its diameter figure out that equation basically that means the diameter of the stone is proportional to velocity squared so bronze came up with this relationship in 1753 and then over the years a lot of engineers and scientists put this to the test so here's a flume by gilbert at uc berkeley dubrot is the first one who did it in france that's our french connection uh and then shelford studied the river tiber in italy and um wrote a paper about that and cited aries law erie hadn't actually published this as a law yet but in his response to shelford's paper said it is just a well-known relationship for the last hundred years that i'm going to claim was uh initiated by brahms himself so um this was all pretty theoretical at the time but it took the expansion of the american west to put ares law into practice and to give us some publicity so we had a couple of things going on in the american west number one was canal construction and that all comes back to a paper by um fortier and scoby who surveyed ten engineers and asked them when do your canal start falling apart they responded um with critical failure thresholds that became the background for all of the shear-based methods once we have some things and see what shields did when he put some science behind it now um if you get a chance to read up on these the accompanying links um will give you a bit of a background on shields it's a fascinating story if you want to dive deeper into how he hitched a ride on a freighter to present his landmark paper in nazi germany um he never realized until the end of his life that it actually made a difference to anyone so read up on that if you can um but uh again these these slides are there for you in the background on that accompanying website um for highways then uh again this is quite a chronology you'll have these slides um but uh basically i'll sum it up by just saying it's a history of implementation of a solution failure and then reassessment and again the big point here is that this 1960s manual that the austral's guides are based on has been superseded by what stanford's going to present to us today the u.s army corps of engineers method so for dams why does the u.s war department get a russian engineer to deliver a paper um to them in the 1930s well hoover dam and other large dams were all the rage at the time most of them needed coffer dams to be built um so they could be built in the dry but then does a cofferdam need a coffee damp so they could be built in the dry well at some point no you dump the rock in a flowing river and that's what this russian engineer named is bash did at a very large scale and like others before him he relied on aries law where it varied by the square and he went in and took that relationship a little bit further and found this coefficient so if we find a riprap size here the d it varies with the velocity squared but that velocity squared is multiplied by a coefficient that can vary with all sorts of other variables and that's the um that was his contribution to it and at the same time barry this grad student went and did the same thing he summarized these coefficients um averaged them all out and made this curve now this is when the usbr the bureau of reclamation adopted barry's approach they put in some disclaimers they said it needed more tests and change well 70 years later there's no change this is still in the ostrod's manual um based on his averaging and uh one thing to keep in mind is the bottom velocity here um that's the uh the bed velocity uh that's what he's saying you should be using so you average all these uh these sources um we see here that um that equation that he came up with matches ostroads they cite another one that doesn't match um and which would give you unstable particles uh but again this is uh they're saying to use the bed velocity but when they're actually making these recommendations here they say to use the average so that's always a big debate is which one do you use and then what's the gradation and this is very uh this is this is very vague here saying most of the ripper app should be that size um and again don't trust everything you read um this statement right here refers to your poll questions and they are wrong by a factor of four so watch out for that um this is the a value here that's used and that's in austral's at the moment and um if i trace this back here that's the austros one i'm taking each of these sources uh back through time and if you line them up and convert the units uh back and forth you'll see that that relationship is the same so this is uh the same charts that came back from 1948 they haven't been changed um and that's what we're relying on so this a value right here is crucial and each one of these uh traces back each one of these um uh applications that i'm talking about traces back to this a value how should you vary the rip-rap um size based on this velocity what should you apply to it they all come up with this uh equation which is seems very simple but there are a lot of things uh in but uh some considerations that can lead to discrepancies i'm gonna pick four things to highlight here um there are more like density and ice mud and debris and things like that that can affect your size but um this is what i'm talking about right here the velocity distribution is one that is huge uh this is one that makes a big difference this is the original source material for ostroads you'll see here that when you have your average velocity which in this case they're saying in the example is 15 you actually need to factor it a factor of two to three to four on that ratio um they say that you either take four thirds of it or two thirds of it now this text that one in the original material never came through into the ostrow's manual so that factoring is something that we've got to watch out for because when you go from two-thirds to four-thirds you get 64 times the weight there's no provision for using the mean velocity in this original equation so um watch out that adjustment has not been included in all strokes and that's uh you know this isn't intended to account for bends um there's quite an extreme effect if you look at this five times um this is simon's lee where i started my career they published this uh note on what happens at a bend and how the hydraulics get affected um the weight would go up 125 times at the severe bend versus a straight channel so it's a huge impact here's the ostros on top of it um it's it's your class four ton rock versus facing class can make a big difference don't double count this if you're pulling localized velocities from a 2d model watch out because this has already been factored you don't want to factor that up again with culvert aprons i'm going to just flip through these pretty quickly this is from the original chart here and the reason i go through this quickly is because it's actually well documented in catchments and creeks and um you'll see these charts very handy charts um that uh that you can just pull things straight out of um so that one i think has been tried and tested grant withridge who put those together um sent me over the background to that one and where the spreadsheets they came from and so those can be used uh without much debate but there's a couple of references in here again if we're looking at velocity-based sizes got to watch out here's one if you ever see a chart or a table of velocity-based riprap sizing have a look at it plot it out and see what it actually does if b is one watch out this chart is not doesn't have any citations with it says it's compiled from various sources um but it shows a linear relationship there's one point here that falls below it but this linear relationship you know i've got over 200 sources that i've pulled out they all show that exponent being either two or two and a half or three somewhere between two and three it's never one so that's one to watch out for also it says seven meters per second um is the recommendation for the threshold between turf and rock um that's not going to hold if you even if you took a linear relationship up here you'd have one ton rock but if you take the actual exponents up here and take that curve it meets up here somewhere around seven tons and uh turf is not going to hold up to seven tons so when you see these charts again it might be helpful to think of this in terms of um uh where the danger zone is you don't want to be in this zone you're likely to fail out there again big uh small smaller rock with bigger velocities um is likely to move uh larger rock with smaller velocities is likely not to move so we've got to stay out of that danger zone um i'll flip through these ones uh a bit quickly here and just show you that some of these things if you plot them out on a chart if you see a velocity and a diameter like in this case well a 300 mil diameter and a 5 meter second velocity if you plot it out in the chart and it falls into this zone watch out it's likely to fail so um and again there you may want to do a conversion between bed velocity and uh and average channel velocity and that can affect where you are in this zone but stay out of the red zone for the shape um we assume that a shape is based on uh when we convert weight to the diameter that it's spherical but we assume also that it is angular and if you did have round rock you'd have to ramp up the safety factors for the angle of repose this is something that i've got to watch out for the angle of repose 70 degrees for randomly placed rock that is not going to happen if you look at the source material again this is the background for what's in austrades it says they used it cemented into a plaster base that's cheating okay you can't do that to get your angle of repose so that is not 70 degrees if you look at all these uh charts that you'll see with the angle of repose and uh how that increases with really angular rock you're never going to get to 70 degrees even the gods of the field here simon's encenter wrote the sediment transport bible basically if you dig through this stuff watch out some of it um you will find some mistakes theta is not the angle of repose if it was in these equations here you would have infinitely sized rock because you get down to zero here so 70 is not the angle of repose and if theta was the angle of repose that would go to zero so watch out you need a difference between the angle of repose and the angle that you place your rock otherwise just pushing it with your finger would knock it over so 70 degrees is not possible another thing that i wanted to chat about here before we turn over to stanford is uh the gradation two-thirds of the stone should be heavier that's the original source material two-thirds of all rocks this is the australian translation that went into uh the austro's manual that those are different okay if we have a single rock here um here we have a rock formation two thirds of these rocks if two thirds of these rocks are heavier um well the w33 is then somewhere between chris and kid because these two thirds of the particles are larger so um watch out for that because i've talked to the authors here anytime you see these references you've got to watch out because the implication is that it's by total weight so the number of particles if you see any references to that watch out and that can give you an order of magnitude difference so if you have these nine stones in a dump truck what would the d33 be well if we talk about the original uh source material if two-thirds of stone should be heavier than d33 right there these two these three rocks right here add up to more uh weight um than the these six stones here whereas if you counted the rocks individually and you had these six rocks here you know they're going to be 10 times the weight of these stones here so watch out if we take a sample say of 24 rocks and split them in half same thing two-thirds of the stone being heavier you get a d50 of 20 kilograms if we counted the individual particles you'd have a w50 of 2 kilograms so that's an order of magnitude difference the cube model can help us work that out this is something from a 1970s document and when you look at an equal number of particles in each sample if you weighed them out you just get quite different results the idea behind this is if you are going to count rock to get your gradation counted at even intervals and then you'll see that you're actually skipping a bunch of smaller rock that you're not counting unless you actually touched it at your even interval and that will get you a more accurate representation there are some errors in the manual so um sometimes you know the wording can have a big impact but the numbers as well have a look at these if you are going to use off-road sizing uh try and correct these watch out as well i'm recommending in the future manuals that we don't use a minimum zero percent larger than um you know that's kind of like saying zero percent of macca's quarter pounders weigh more than a quarter pound well what does that mean it doesn't actually tell you anything uh you want a d85 or a d90 that will tell you something so taking velocity based to shear based all the shear based methods i'll just show you down here up to about 100 mils and so anything beyond that is extrapolated most shear based methods flip these axes around so let's flip it around and put d50 on the side here and then when we turn this into attractive force on the y-axis and the diameter on the x-axis all the charts i've seen stop at 100 mils and so that max has then been extrapolated a lot of us agencies use this table here from fishenich and when you look at these it's just a linear relationship that's been extrapolated way up so we're way down in this zone right here and we've extrapolated out to here is that true does that need some testing i'm recommending we look at shear based but when we use it for big rip-rap you probably want to compare it to something else bob keller put this spreadsheet together that has not been updated in some time but it is a sheer based spreadsheet i've got links to this one if you wanted to have a look at this one and it shows rip rap sizing increasing with depth using the shields coefficient i've got another link to a paper we did then where we compared velocity versus shear and for small channels and culverts for rio tinto and we showed that the sheer results include a safety factor of two if they do we still have a size reduction if we went from velocity to shear based so that's a potential cost savings watch out as well though if we take average versus localized peak hydraulics you're going to see a size increase if you apply like 2d peak hydraulics instead of average velocities which is what some of the methods are based on so if i take a look at this one right here just as an illustration if i had the same velocity in both of these channels and that was equal i've got a bigger dv uh flood hazard in this one on the left and i've got a bigger shear stress in the one on the right that doesn't necessarily make a lot of sense intuitively but when you look at the way the shear stress varies with the slope then it starts to make sense because in profile view the steeper your channel the higher that shear stress becomes and your dv though is going to be lower like these ones right here having a constant velocity of one that's going to give you the same rip-rap size if you use velocity-based sizing but you're going to have a very different dvs and their ability to move uh big uh you know buildings and people and uh and vehicles uh is gonna vary so again if we look at something that's got a constant velocity here four meters second or four meters a second this larger channel actually has a lower shear stress and if you did a velocity-based sizing you would get the same riprap size for both of these channels does that make sense um it's something that we can debate a little bit about when stanford shows us here the results that came from the army corps of engineers experiments um and uh we will look at uh how how things vary with depth keep in mind these do have some limitations uh but the thing i want to talk about um the the stanford we'll we'll get into the guts of here is this uh coefficient or the exponent on the depth so if we have the depth on the army core of engineers method uh in the end when you add all these up it ends up being inversely proportional it's a negative okay so does that make sense um the velocity is raised to this two and a half power not the two power which we see in the austros but the two and a half power and again this method has superseded the methods that ostrows relies on so it's different um and so should we be using that um keep in mind also that these exponents out here if they're outside the uh the the the uh sorry if the coefficients are outside the exponent you're gonna get very different coefficients um using one versus the other um i will uh just this one um i'll refer you back to the presentation that we did um that number 100 uh we discussed this a little bit about how whether depth should uh whether the rock size should be uh increasing with decreasing depth you'll soon get to zero and that wouldn't make sense but um this method does have some significant limitations and we want to make sure we stay in the green zone so i'll refer you back to the previous presentation for those limitations and again we don't size rip-rap like that for practicality-wise but if you looked at um rip-rap varying inversely with depth you'd actually have it looking like this and obviously you never designed rip-rap like that so how how does that how do we account for that um also uh if you take rip-rap here and have the weight of the rip-rap on top of it one of the things that's not accounted for in the equation is that these particles end up more stable because there's weight pushing on them and squeezing them down and keeping them in place if we look at horizontal and vertical variation we have rook rap sizings that take 1d results and factor them to account for both horizontal and vertical variation 2d models are always going to use depth average velocities but the methods won't tell us how to separate the two so we got to watch out for that and stanford we'll talk about that shortly so here's some typical 2d results where should you be extracting your velocity results from and how do you plug that into the equation what is your v seems like a simple equation but what is that v now there are some advancements a lot of people are doing this the recent 2d modeling guidance by the federal highways administration shows this figure on how it could be done but you got to be careful if you look right here 14 feet uh for the d50 that if you took a d90 for that and converted that up that would require a pretty massive truck so watch out um in in this illustration um this is using hec raz um and this is why i'm excited to have this tool in heker as to do the rip-rap sizing um but uh one of the things that you could do and that i'm recommending uh doing is checking it by shear which i'm doing here as you go in and take the shear you can multiply the shear up by a given factor and get you a riprap size you know your millimeters of uh particle size can turn into a recommended um you know rock size based on the shear stress likewise we can use the uh like a raster calculator to take any coefficient any a value that you like and put it put it in here in hecara's or any other gis software take your layer in this case v times v times 35 and that will give you your ostroads sizing but proceed with caution because um if again if you're taking this from 2d results some of these methods were based on 1d so down to my recommendations uh take at least three methods why do we trust the usbr when they say do three methods well the the usbr method is in os roads as well um and it is worth checking a couple of different things um the these three recommended methods that i'm putting in here uh this is covered in that paper that i presented um i do recommend checking the velocity-based uh riprap sizing and the shear based and then this is the one that stanford will walk us through so this is the core of engineers equation but what i want to also look for some clarifications on here is that we need to clarify when we use these different zones so for structures this will be an okay one this is the osro's tables converted to a chart channels this is the osro's chart right here so if you're going to design a channel you can use these coefficients stay out of the red zone but if you go too far into the green zone as well it gets a little expensive okay up in here you might be overdoing it and over engineering something so this is the sweet spot in here uh checking between these sizes so what we'll talk about today is how to apply these methods some clarification is definitely needed some of the recommendations in ar and r say that you know we don't like to use riprap it's not favored but it's going to remain one of the primary scour counter measures so if we've done our job right selected the right one to your 2d or 3d tool and if we've sized it appropriately and we've done our design parameters and our construction methods correctly then we hopefully won't deal with uh need to deal with the remediation so back to our quiz question then uh i noticed that uh on the poll results a lot of people were up in here zero we're at 64. but again take a look at this if you took a three meter a second velocity with a 300 mil rock size and you went up to a six meter a second velocity and you've got a 1300 mil or 1200 ml rock size you're going to get a weight that varies by factor of 64. so that is surprising to most people that's why i spent so long on the velocity versus sheer methods and things like that because it is hugely sensitive to the velocity and where we pick it so there are some additional resources for you these are links that you'll have available to you catchments increase an awesome compilation of rock sizing resources everything that we've talked about in the webinar is here at this website there is a toolbox we talked about in our previous webinar from the federal highways department and we also have stanford gibson's youtube channel stanford's going to walk us through some very basic things but i don't want to repeat what he's got in here so i do recommend that you uh subscribe to that and have a look at that one so with that um that was just a whirlwind of material um i know i've thrown that out there pretty quickly hopefully in the background you've got some questions coming in keep your questions coming in on the chat line and we'll hope to get to you live momentarily after uh stanford walks us through this calculator so with that over to you stanford so over to you all right thanks crait that was equal parts illuminating and disturbing um there's some uh some uh you you don't want to just use use the equations or the the charts you need to think this through you need to be doing your own work um and uh you know that's always the downside about putting a calculator in raz is that you know it could be easy to type in the numbers and so when craig asked me to do this you know i have youtube videos and we have a whole user's documentation for the riprap calculator um but what he asked me to do is actually walk through the core of engineers approach two years ago i did not know the core of engineer's approach i've been working very closely with our sister lab urdik was the the the lab formerly known as the waterways experimentation center david may down there who's on here as well as chris herring and david biedenharn um really the core riprap brain trust and then i just want to note that zack morris coded all of the riprap calculator and so everything i show you today in the wrap riprap calculator was developed by one of our software engineers zack morris all right so let's actually talk about the method and approach behind the scour calculator rear prep and scour calculator and then i'll just take you on a a really quick tour of the tool itself um and so the cores method comes from what we call em 1601 and uh this is called hydraulic design of flood control channels and so you know since this came out we've changed our language twice we don't call them flood control channels anymore because we don't believe we control floods um so we changed it to flood damage reduction and we don't even really believe that anymore so now it's flood risk management so this document has survived two paradigm changes uh but chapter three of this document which you can go online and get describes the core of engineers rip-rap approach and the whole the whole process came out of this series of elaborate experiments at waterways experimentation center now urtic the engine where you know they built this mesoscale experiment this is a giant flume it's called the riprap facility um it's not quite prototype scale so it's a mesoscale experiment but it's not really a lab study either and what they did is this was the work of steve steve maynard who's like the grandfather of um of riprap in the core and you know we call the equation we put in raz the maynard equation is he essentially you know modeled different bends with different radius of curvature um and a straight channel and the the idea behind his approach was that we're going to actually measure a reference velocity in the upstream channel in the straight upstream channel because once you actually get into bends the channel deforms and the the the one-dimensional computation of velocity is actually deformed by the irregular channel shape there so what we're going to do what he did is he measured the stability of the rip-rap at these different radius and curvature bends but always tied it back to what we call a reference velocity in the upstream reach and so when this is the distinctive one of the distinctive things about the core of engineers approach is that most other approaches look at the velocity right where the right at the design bank but the core of engineers approach always looks at an upstream reference for velocity either at a crossing or um or a run where their channel is well behaved symmetrical and one-dimensional the other implication of this is that it's 1d you might ask hey how why can't we tie our rip-rap calculator to the 2-d results well it's because our equation is not tied to 2d results it's an empirical equation that is connected to the upstream reference cross section and the average channel velocity in your in your channel we just kind of need new theory in order to um in order to do better than that all right and so when you're in raz um you're going to choose a design cross section and an upstream reference cross section and so you're essentially your critical condition you're essential you're going to get your most scour on the downstream outside of the bend because that's just the way rivers deform the downstream outside of the bend is going to be generally your design cross section and then you're going to choose an upstream reference cross section which is kind of the upstream the closest upstream cross section where the channel is kind of one-dimensional and symmetrical now we've moved all of our user documentation online if you're not using our online users documentation you're kind of missing out because um you know those old pdfs were kind of hard to find things in our user documentation is over a thousand pages long and so you don't want to be hunting and pecking for things you want to start using it searchably and so if you haven't actually moved to our online documentation um you should and here is i'm going to drop in the chat the actual link to the riprap documentation and one of the things we do is we kind of walk you step by step through not only the equation and putting in the numbers not just the number punching but actually the process um in the overall core of engineer method and there's just a couple of things that we need to talk about one is you know what flow do you use well in em 1416 it states that you should use the flow that produces the maximum velocity in the channel and so here we have a very well behaved channel where you have velocity that increases about to the bank flow a little bit above the bank flow and then it drops um the channel velocity drops as you get flow over into the over banks um and so that would be ideal um you could go in you can plot this in the rating curve calculator and raz you can plot flow versus velocity and find that maximum velocity and that's your flow but that's not the way things always go like there's often actually a you know there could be an inflection point but then it increases afterwards or sometimes velocity just increases monotonically with flow and so the question is what's your maximum velocity in that case you just have to kind of take responsibility for a risk-based analysis you need to take responsibility for hey um we're not going to model to the maximum velocity we're going to choose some sort of recurrent recurrence interval maybe bank full but we're definitely going to do a risk-based analysis to look at a cost benefit of the chances of this failing the other thing that's distinctive about the core of engineer approach that i think as craig and i were talking he suggested that maybe this is why some people find it a little bit confusing is we actually come up with two velocities and two um d30s the the most obvious one is the one in the bed um you get the d30 in the bed and to do that you have the hydraulic depth of the whole cross section and the average velocity of the whole channel or of the channel portion of the cross section that's pretty straightforward but that's not where you often put rip-rap you know sometimes you're doing a pipe crossing or something like that you need to put rip-rap at the bottom of the channel but often when we're putting down riprap it's protecting the toe at the outside of a bank or a levy or something like that and so we really care about the d30 of the slope now craig talked about how sometimes how we're kind of under accounting for the weight of the of the rip rap above it but in general the weight required on the slope it has a downward gravitational component and so we're going to compute that to be higher and also it's on the outside of the bend so it's going to feel larger velocities and on the outside of the bend and so there's some empiricism to take to take that into account and so to do that we compute the velocity and the depth so 20 up the bank and this is stuff that kind of used to be annoying um this is one of the reasons we just put in the rip we just created the rip rep calculator so we can do these computations for you but you should know what's going on and so if you want to compute the depth 20 of the bank well you're gonna have to go google cotangent which is what i did you're gonna have to dredge up your trigonometry and eventually you can figure out you know just with some trigonometry what that um depth of the side slope is the velocity of the side slope you can't do there's no trigonometry in the world you can do to find that vss and so um this is one of the findings of manor's experiments is he went in and he found for a given radius over width ratio what's the ratio of the side slope velocity to the average velocity and as the ratio of the the radius of curvature um to the width decreases you're going to get you know you get a a sharper bend and you get higher velocities outside of the bend relative to your average velocity and so it kind of upscales your bend velocity and so now we have two sets of depths and velocities and now we're going to put them in the main equation the main equation computes the d30 and it has all of these coefficients which we'll talk about a little bit it does have the depth which does have that negative power which is interesting um but then it has the velocity raised to the 2.5 and the question is which depth and which velocity well we're going to compute 2 for you we're going to compute the d30 of the bed that uses this average velocity and depth and then we're going to compute the d30 of the side slope which uses the at the side slope velocity and the side slope depth and so um we include this helpful if busy uh diagram in the user's manual that shows you exactly what each of these are and in the riprap calculator we pull three parameters from the hydraulics and raz you put in a couple of parameters yourself and then all these intermediate parameters like the side slope velocity the uh the side slope depth the the that ratio these coefficients come from these categorical user choices up here and we populate those for you and um fill out this equation for you and compute the d30 of the bed in the side slope now if you kind of want to see how we actually did that we've we did some validation verification computations and we've included these in the user's manual you can go and this is an example calculation that was in em 1601 but we've had some others where we go and we actually show you the computations so that you can actually replicate them in in raz the idea here is that you don't want to just push buttons and get a number you you want to you want to trust the software you're using and so it's nice to be able to reproduce the numbers and see how we got them but it doesn't actually matter if you size the rock correctly if you don't protect the toe toe protection is the most important part of the core of engineers design process which is why we combine the riprap calculator with the scour calculator and here's the basic idea you go and you put your rip wrap in let's say you use a safety factor of eight right like you're just gonna you're gonna oversize this rip rap because you're just gonna protect this bank that you've got to protect this bank at all costs um you shall not pass water um but what's gonna happen with this riprap design well if it's on the outside of the bend you're gonna get toe scour and your rip rap is going to slough and it's going to expose the bank and you're going to get damages um so it does it doesn't matter if you get your riprap sizing correct if you don't protect the toe and there are two ways to protect the toe one is to key the toe and so you can go in and you can you know kind of pre-excavate the toe and go in and put it and key it putting your rip-rap to depth so that when it scours it doesn't get below the toe the other way and kind of the more common way is to develop a launchable toe you put your extra material at the toe then when it scours it fills the hole dynamically um it kind of it's launchable it launches but what you'll notice for both of those designs is you need to know how deep to make the how deep to make the toe protection you need to have a scour depth and so that's why we've added scour depth calculator to razz it's just you know we have the riprap tab here and the scour depth tab here and so you can do these analyses side by side because we think you should we don't think you should be sizing riprap for a channel bank without thinking about scour and toe protection and this you know um this again david may um uses this in the core process uses this suite of equations to compute ben scower or a general scour if if you're looking at the channel and so that's um what we added here okay so let me just uh let me just note uh this is a team effort you know i'm giving this presentation but you know zack morris developed it um david made david biener and chris herring um were all they all designed the algorithms and then this was funded by the regional sediment management program of the corps of engineers let me go now to a demonstration and so here's raz and the first thing you're going to do is you're going to run a steady flow because you need steady flow results to do this um hopefully in the next version you can do it with unsteady flow but right now it's it's it's steady flow only and then you're going to go to the hydraulic design calculator the hydraulic design calculator has a lot of cool stuff in it including the grid bridge scour but we're going to go to type rip-rap and it's going to bring up this cool new rip-rap calculator the zac designed with oh you'll notice it has a little bit more of a modern look and feel than the rest of raz and that's because we are designing raz 7-0 to have a more modern look and feel after kind of 30 years of the same raz um and so uh what we have these hydraulic parameters um b average hydraulic depth and width that come straight out of the cross section we're actually going to go in and define what our ups our reference cross section is let's use 3 500 as that cross section right there and what our design cross section is let's use this 3 100 as the design cross section and then i'm just going to put in two variables i'm going to put in my radius of curvature let's say it's 300 feet and my side slope angle let's say it's 25 and you'll see that we compute our d30 in inches um you know you'll get millimeters if you're an si um and uh we've got a d30 of the bed and a d30 in the bank now you see that these are plotting here as well and so if you go to these are the gradations that are em1601 and so you in most cases it's not sufficient to just define a d30 you also need to define a gradation and so you can go into this tool and define all the gradations that are available to you at local quarries but in order to get the thickness the equation we use for the thickness of your layer is either one and a half the d50 or the d100 and so in order to get that you need to compute both the d50 and the d100 but we've only computed the d30 for you so now you go in and you say okay for the bed i'm going to use this gradation and then for the side slope i'm going to use this gradation and that'll populate your d50 and your d100 and it'll compute your thickness all right so we have our rip wrap size and gradation and thickness now we're going to go over to scour depth again we're going to add the radius of curvature 300 and the d50 of the bed not of the material or not of the riprap and what you'll see is that we immediately get um some some scour numbers now this your general scour show up as you know your channel dropping here but your toe scour will show up as actual scour on the outside of the bank if you want so i'm actually going to turn off the general scour here and we add these helpful little buttons that will tell you if um your parameters are outside of the range of that tool um and so here it says hey you know um lacy was developed for silt and it seems to me that your your material isn't silt maybe you shouldn't use this one and so then what you can do is you can come in here and we can say okay my toe is at you know 168. so you put in your toe station 168 and now we can visualize the scour where it is and we give you this kind of ensemble approach you know these equations are all pretty simple they're all pretty old and so what you do now is you gather a set of kind of regional specialists who have a sense of what the scour is actually like on systems like this you look at the suite of scours that um the equations provide you and you decide how much launchable material or how deep you have to excavate in order to key your toe um to you know feel comfortable that you're going to um that your toe protection isn't going to get undermined and with that i will wrap it up and turn it back over to cray and aws great thank you stanford for walking us through that and again um the point of all this you know the these methods are not uh just you know magic bullets you don't just uh take one and accept the answers uh there's gonna be a little bit of investigation uh required um when you're looking at these and so again because the australian riprap sizing methods fall back on some uh publications that have been superseded by this method uh my recommendation is that at least give it a shot now one of the questions then if we go back to just plain rip rap sizing um aside from the scour volumes uh stanford is uh you know the the maynard equation has some uh limitations you know i think two percent for the channel so if you were designing anything else are there plans or do you have recommendations for some of these other applications like you know spillways and bridge bridge scour and things like that or bridge counter measures yeah i mean that's a good question the uh the core of engineers method actually has a equation for slopes over two percent we're kind of evaluating whether that is something that should go in the tool the uh the riprap smes are taking a look at that it's not widely used in our agency um there are some the omaha district uses it there are some districts that use it but it is not widely used and so we're just we're just kind of taking a step back and evaluating whether that should be in the tool um and then you know we are you know we are surveying other methods also if flow is turbulent then the major equation is considered kind of out of bounds that you you shouldn't be using it and we have a different approach for turbulent flow thanks yeah and that that question did come up maybe back to that question you had there or the the the comments you made there stanford about turbulent flow mixed flow regime uh what do you think about that as far as turning that on um and and should that be uh applied when you're doing rip rap sizing i'd be very cautious with mixed flow regime um i think that i mean you saw the you saw the experiments these are based on right i mean these the equation's not magic it's impera it's empirical and so i think that when you when you think about whenever you apply an equation you have to ask yourself where did it come from and is it remotely like my situation um and this is true all across sediment transport because where everything we do is empirical um and so you just think about think about maynard setup there uh completely subcritical plus to get super critical you're probably going to exceed that that two percent slope um you know limitation so i think that you just need to be very careful with mixed flow yeah and as far as um a couple people have asked the question about um steady versus unsteady now you know in australia hardly anybody does uh 1d flood models um we're not limited to you know fema trying to replicate old fema uh model runs and things like that and so there isn't a lot of uh 1d flood modeling going on um so i guess the question is if first of all if you had a 2d model um you know what what guidance can we provide as far as trying to do the same thing that you've done here with getting extracting an average uh cross-sectional velocity out of that 2d model and then if you did have 1d as an unsteady flow what are your recommendations as far as setting that design flow and then converting that into a steady flow you know understanding that a lot of engineers are going to try and do as little work as possible as far as additional models that you would need to set up once you have one that you already trust first let me defend 1d modeling the the it is it is totally appropriate to do 1d modeling um in any country the uh you know and one of the issues is our country's just really big like we got a lot of land mass and we don't have you know uniform national high quality lidar and so it's still uh you know we're still in a situation where and even so like you have to model the channels and sometimes burning a channel if you're going to burn a channel from cross sections you've essentially got a two-dimensionalized one-dimensional model anyways so let's let let's have a little bit of space for our friend that one-dimensional model rivers have high aspect ratios okay that said um you know of course we're all too be modeling now that like that's what we're all doing and so the quest the idea is oh we've added this this cool new tool to razz except you can only use it with 1d models how cool is that well you know it's just it's the state of the art you know if someone wants to go out and measure two-dimensional velocities and come up with a new theory of rip-rap then we'll add it to ras um and we've actually had discussions about maybe it's time the core does this maybe it's time for maynard part two david may to uh to actually do something like this but for now it's all tied to that one-dimensional velocity in the upstream cross section and so if you have a two-dimensional model you can go and you can cut a transect and you can get a a velocity profile from that now craig when we were talking about this before pointed out helpfully that we don't actually provide any useful like one-dimensionalizing metrics on that we just give you a profile velocities and you have to like go in excel and do a like depth weighted average and actually it's it's it's weirder than that because we're just giving you the velocity magnitude not the velocity perpendicular to the cross section and so we actually just had a little meeting today about how yeah we should probably provide some 1d metrics um in order if we're going to ask people to you know one-dimensionalize their 2d model to use in the rip-rap calculator we should make that a little easier it's not actually mathematically trivial um and so we're actually going to have another meeting on how to do that but we're going in that direction but the way that i would do it is well actually the way i would do it is i would write our code against the hdf5 file that's how i did it for a paper i wrote once but uh the way i would do it if i wasn't doing that is i would write i would create create a transect um right click and get my velocity distribution put it in excel and do a do a depth weighted average okay now thanks for that that's that's probably the main question that uh comes up and gets upvoted um apologies to everybody who is uh frantically uh that over the last few minutes uh typed in some questions so we probably won't have a chance to answer all of them but what we will do is have a look at these and potentially answer these individually and what we generally do is provide a pdf file afterwards that have q a uh questions and answers um that you can look up and then maybe you can find the answer to your own question and again you know this is something that uh we're throwing out there as a discussion topic you know have a look what should we be doing uh going forward um you know should we be like stanford said um coming up with a bunch of 3d modeling to validate new tests uh you know and and you know 1d 2d 3d um and and make those comparisons between uh all of those modeling types so we can know which velocities to extract and which uh you know how we ought to be doing our rip rap sizing so um stanford any that maybe i'll just uh give you a chance to just uh pick pick one uh if you've seen one on there um to to highlight uh before we close off for the day yeah i mean there's one it's not the highest upvote but i think it's an important one is uh it says you know if i have a long river with many meanders i think this this is the the uh the heart of the question um how do i apply this tool and that's a good question because with like all the theory that you use that uses local velocities you can just do a spreadsheet and do it for every cross section at once but in the core method we can't actually produce those like large scale global results because it's not tied to the velocity of the individual cross section it's tied to an upstream reference cross section and so we've actually one of the things that we do is we allow you to keep the previous we have some sensitivity stuff in in the model that uh you can use to look at the sensitivity of different cross sections but we really we're kind of unrepentant about the fact that we make you look at each bend individually we don't want you just printing out a list of numbers and going to your quarry and getting the rock we want you to actually look at each bend think about that bend as an individual bend is it an actual bend or is it a compound bend we have a lot of material from david on compound bends that you really need to think carefully like what is an actual band and or is this a bend in a bend we want you to think about each bend individually um and so the fact that you have to you can only look at one bend at a time in the core method i think that's a feature not a bug good excellent um so thanks so much for coming on stanford um this is this is quite a treat for us to have the developers themselves um on on the line um in including you know those who who wrote the code and know what's in the black box do subscribe to stanford's channel if you're into this um you will see a a lengthier walk-through of what we've just covered um do fill out the survey at the end let us know what you want to see more of we've got some exciting uh webinars coming up and courses for the next year and we are we're planning 2022 at the moment so if you look here at uh some of the courses and webinars coming up you know this is what we've planned but what do you want to see more of so do fill that poll results out have a look at the accompanying materials this will be available for you on youtube if you want to slow any of this down and and have a deeper dive into any of the parts that we covered maybe a bit a bit quickly today um i would encourage you to have a look at those resources so with that um any closing room our stanford before we uh let everybody go get back to their day i kind of double dog dare anyone to listen to either cray or me on two times speed at uh on youtube that's true it'd be a little uh yeah it was already a lot yes sometimes i recommend people do that uh but in our case we tried to cram as much information in for you as we could in this one hour session and again uh invite you to ask us further questions um you know we nerd out on this stuff and uh you can see it's a rabbit hole i thought okay sure you know somebody asked me let's just trace the ancestry of the australian rock sizing equations where did these things come from i thought it would be like a two-hour exercise and i'd be done but you know i've i've got a spreadsheet with hundreds of documents now just trying to get to the bottom of where these things came from and it's fascinating story to see where they came from and it's also very important to know what were the limitations uh of the methods that were used to to develop the equations that we just sometimes apply blindly so watch out for that so with that thanks so much stanford for coming on thanks to our panelists in the background frantically answering questions um we will see you next time around at our next session do subscribe and give us your feedback thanks we'll see you next time thanks for watching subscribe by clicking the link below and click on the notification bell to stay up to date with new releases for the latest in significant innovative and critical advances in water science technology and management subscribe now to build your skills enhance your technical knowledge and learn from leading experts in water visit the australianwaterschool.com.au and discover our online training courses both live and on demand [Music]

Channel: Australian Water School

Views: 5,913

Keywords:

Id: cJ8994fHJg8

Length: 60min 25sec (3625 seconds)

Published: Thu Oct 14 2021