Going for Gold: How to Win an Olympic Medal with Autodesk Software

RICHARD DIDHAM: Cool. I guess we're all set. So to everyone here, welcome to Going for Gold: How to Win an Olympic Medal with Autodesk Software. I'm Richard Didham. I'm a Design Analysis Engineer for the US Olympic IR&D Program, and for the next hour or, so I'm going to be taking you through a couple of different topics within the US Sailing Olympic Program.

So first a little bit about me. I'm a Design Analysis Engineer for US Sailing IR&D, and I work jointly between the US Sailing Fast USA Technology Center inside the San Francisco Bay Area and Autodesk's Pier Nine Rapid Prototyping Facility. So the US sailing team is lucky enough to have residency program with Autodesk, and we've been using their facility for some of our projects.

So my primary role within US Sailing is to design and prototype advanced composite structures for Olympic equipment, and I have a bachelor's of science in mechanical engineering from UC Berkeley. And before I got involved in the tech side or the engineering side of Olympic sailing, I actually used to be an athlete in the Olympic Development Program.

So I have a lot of perspective on both sides-- the athlete side and the engineering side of Olympic sailing. So here's just a brief overview of what we're going to talk about. I'm going to start off giving a little introduction to the US Olympic Sailing Program, what IR&D does within the US Sailing, just a broad overview of the projects that we work on. Then, I'll go into detail about a specific design project that I worked on, which involved a lot of Autodesk software and a lot of collaboration with Autodesk, and then we'll finish off with some future development plans that we have.

And before I get into it, I'll say that this presentation is going to kind of give a very broad overview of a lot of different subjects and topics, and I'm going to go skipping through a lot of things pretty quickly. So if anyone has any questions with what I've said, what I'm talking about, please don't hesitate to raise your hand, yell something out, ask me, because I'm sure other people have the same questions that you do.

So anyway, I first want to start off to give a little perspective on what Olympic Sailing looks like, because most people aren't very familiar with it. So usually, when people hear the word high performance sailing and think about sailing in general, either they're not really sure what that means, or they'll think of things like I have in the bottom slides, like these big yachts with large teams of people racing in pro sailing and corporate sponsorship teams.

So you see on the bottom here, this is a catamaran from the last America's Cup, which is not related to Olympic Sailing. There's a little bit of overlap between the two, but broadly speaking, they are two separate things. To define this a little bit better, we have some key distinctions here. So with other high performance sailing not related to the Olympics, these things are classified with large teams of people racing on the boat at any given point in time.

So it's a large yacht. There'll be five to 10 people on it. And then the boats themselves are what we call Box Rule classes, which means basically, the design and build of the boat just needs to fit within a certain set of parameters, which we'll call its box rule. So it can't be longer than this much, it can't be wider than this much. There's certain rules on maybe the materials that you can use for the build. But other than that, you've got a really large range of different design approaches that you can use to build these boats, approach these problems.

So Olympic Sailing by comparison, it's very small teams of people, so all the individual athletes when they're racing will either be by themselves in a single-handed class, or there'll just be one other person sailing what's called a double-handed in class. You can see the picture down there, there's two people on that boat. And the specific boats that they race inside the Olympics are called one-design classes, where the idea behind a one-design class is everyone is supposed to have roughly speaking the same equipment to race on.

So it's supposed to be a lot less of a design competition and a lot more of a competition among athletes. So most of the time, there will be certain manufacturers, that those groups are only allowed to manufacture the Olympic equipment. And it's not allowed to come down to different nations building boats and different nations building certain equipment.

The few things that we can design and build ourselves, there's often more strict design rules around that equipment. So ultimately, a lot of the time when we're working on projects, it comes down to trying to exploit really, really small differences that are allowed within these rules to try to get really marginal gains above our competition.

So within the US Sailing Team, IR&D, our primary goal is to improve equipment where there is room for refinement. So the few things that we can improve, we want to make sure that that is as good as it can possibly be. And then, anything we can't change, we want to profile it to the best of our abilities. So sometimes when equipment is made in large scale from a designated manufacturer, the manufacturer will have nuances inside their manufacturing processes that allow for just a range of different equipment characteristics.

Sometimes when they build a mass, for example, it could be stiffer than others just because there's defects inside their manufacturing process, or there's variability in their manufacturing process. So what it comes down to most of the time is there's small teams dedicated to trying to refine our equipment, as well as possible. What we often want to do is try to model whatever we're trying to work on to improve.

So a lot of times it comes down to modeling different equipment. It could come down to trying to model a venue that athletes race in so that we can pair a particular type of equipment for that venue as well as possible. And then there's also a lot of performance analytics, making sure we understand that our athletes are using the equipment as well as possible to make sure they're getting the most out of it.

So there's often a saying within US Sailing R&D where the three most important factors to an athlete's success in sailing is boat speed, boat speed, and boat speed. So ultimately, there are other things that will influence an athlete's performance in a race, but if an athlete is superior in their both feet in that they just simply can go faster than all their competition, it can help them make up for mistakes in other errors of their sailing.

So we've identified five different areas that we think can influence an athlete's ability to perform in terms of both speed, and these are the five broad categories that we work in when we're working on projects. So the first of these is human factors, which is fitness of the body and fitness of the mind. Often, this will come down to biomechanical analysis, so we're trying to understand the mechanics behind how athletes race.

Certain athletes will use their equipment slightly differently than others. They'll have nuances and techniques that we're trying to figure out. Make sure that each athlete is, again, using their equipment to the best of their abilities. The other one is racing environment. So sailing, more so than probably any other sport that I know of, is incredibly crazily variable in terms of the environment of the race. So different venues will present completely different racing environments, and there's just a wide array of factors that will change from one day's racing to the next.

So you can have lots of breeze, you can have no breeze, you can have a very choppy sea state, you could have a really flat sea state. There's all these currents that vary from one venue to the other and also very with tides depending on time of day. And for an athlete, it can be often really difficult to make sure that you understand how to race in your specific set of conditions, the specific conditions you're presented in a day.

So one thing that we want to focus on, particularly for the Olympic games, because obviously that's the most important event that we're always trying to work towards, is making sure that we understand the Olympic venue and understand other important venues that fall into the Olympic calendar as well as possible. So a lot of times, this comes down to accurate measurements of the sailing conditions, either while people are racing, or we'll do little reconnaissance missions, tying to get a better understanding of the venue as a preemptive effort before athletes go and race so that they can have a better understanding of what's going to happen on the next day that they go sailing there.

So number three ties back into number two pretty well. So with sailing, there's often multiple ways to approach any given venue or any given race course. A lot of times, different strategies for approaching that venue will end up being the correct strategy from one race to the next. So going off what I was saying with the previous slide, you often need to try to understand what's the best way to approach a sailing venue from the athlete's standpoint, and how do they exploit all these nuances in how the currents are moving around the race course, and if there's more wind on one side of the race course than the other, and how the breeze is changing, if it's shifting from one side to the other.

And also, making sure that the athletes can understand how to use those advantages more so than their competition. Then, the last one is we consciously are working to try to make coaches' lives easier. So often when you're a coach, you are presented with the same set of conditions and the same environment as an athlete in that there's often a lot going on, and it's really hard to understand the nuances between one athlete's performance and another.

So we work a lot with trying to get quantitative data on all these different factors of how an athlete's performing as opposed to the traditional form of coaching, which is mostly qualitative. So the data acquisition platform, or DAP, that we've developed is a three stage quantitative measurement tool which allows us to combine data from athletes themselves, combine data from the equipment that they're using, so how fast the boat is performing, and combine that with comprehensive video analysis to make coaching more efficient so that the coach can get more condensed information and information more readily available to inform the athletes on how they can change their technique.

And then the last one, which is my favorite, is sailing equipment. So going back to the earlier slides, we design and prototype specific equipment that we're allowed to change, and for the stuff that we can't change, we're trying to measure and profile the equipment in order make sure that the specific athlete is using the equipment for the right conditions and [INAUDIBLE] application. So we're lucky and fortunate enough to have a few partners with us to try to help our athletes perform at the best of their abilities.

So as I mentioned earlier, we have a residency program with Autodesk Pier 9 and the Boston Build Space to work on prototyping. Saildrone is a startup into the Bay Area that allows us to work on environmental or venue reconnaissance in terms of weather and tactics, and then University of Miami is another partner which helps us do some of the more mechanical engineering simulation side of things. And then SAP is helping us work on our DAP program in order to help expedite coaching and data processing.

So this is just a little slide to show where all these different partnerships fit in. For the remainder of this presentation, I'm going to be focusing more on Autodesk and the sailing equipment side of things, just because the next part of our presentation is going to be specifically focused on the Moth Hydrofoil.

So moving onto that, so the Moth Sailboat is a high performance Box-Rule class which is not in the Olympics but has several beneficial key attributes. So the relaxed-box rule encourages development and allows us to build and prototype equipment or test out different ideas without having to get caught up inside all these intricate little details. That if we're building stuff and we have very, very strict design rules, changing small things is a lot harder to understand if that's improving or detrimental to performance.

So another good benefit is many of the US Sailing Team athletes actually race Moths as a supplement to their sailing, just because it's a high performance sailing discipline that is fun and is a bit of a more relaxed atmosphere than the Olympic Sailing disciplines. And then, final key attribute is all these projects can be more readily shared with the public, just because it's not the secret equipment projects that we're looking to hide from other countries. So it's fortunate that I can talk to you guys about this today.

So just give a little preview or a little insight on what sailboat racing and Moth sailling looks like, this is just a little highlight video from Moth World's 2017. I just threw this in here to give a little perspective on what sailboat racing looks like on the small scale, so you'll have large fleets of boats with all racing together on a start line. Inside these races, there's about 50 people racing at any given point in time. And like I said earlier, so this is a single-handed class.

This guy just broke his mast, but he's not an American, so we don't really care about him.

[LAUGHTER]

But yeah, so one interesting thing about the moth is it's a hydro foiling dingy, which means it actually has a little wing underneath the boat that lifts it out of the water, and it allows it to have significantly higher performance than most other comparable sized dingies. So if you see all these boats off the starting line, they're actually flying slightly above the water.

And the only thing that is in the water creating hydrodynamic drag is the hydro foil itself and the structure that attaches that hydrofoil to the boat. So this is similar to some of the other disciplines inside the Olympics. If you ever look up a boat called a Nacra 17, it uses a similar concept, where it has a hydrofoil lifted outside of the water, except it's a double-handed boat instead of a single-handed boat.

Anyway, what we wanted to work on for this project was to design and prototype a hydrofoil foil on the goal what we wanted to work on specifically was to test out hydrodynamic design optimization, so we're looking to build something similar to a hydrofoil for some of the other Olympic classes. And then most of the equipment inside Olympic Sailing are all carbon fiber reinforced composites, so this was a good project in that we were able to play around with and test out our manufacturing capabilities for fiber reinforced composites.

Here's just a brief overview of what this project looked like from a high level scale. So we started out designing hydrofoil profiles inside Computational Fluid Dynamics software, or CFD. Once we had finished the analysis design workflow for that, we moved on to Fusion 360 CAD in order to create a solid body model of what we wanted to make. From there, we kind of split off into two different roads. So we used a CNC router to machine moulds for the bodies that we had designed inside CAD.

And then alongside that, we also defined a layup scheme for the composit layout that we're going to do. I'll talk more about that later once we get to that part. And then finally, we fabricated hydrofoil, and brought the design to life. So to begin with, in terms of how hydrofoil design analysis workflow, there's a couple of different ways you could go about this. But the way that we decided to approach it was to try to parameterize all of the potential shapes that we could use for this application, and then throw that into CFD and have CFD tell us which was the best one.

So this started off by using shape parameterization formula, which is just a parametric math formula that will give us different shapes depending on numeric values that we input to it. So the five different numeric values that we input relate to bluntness of the leading edge, location of the max thickness, value of the max thickness, sharpness of the trailing edge, and the chamber-line.

So just an example of what this looks like. The two hydrofoils in the bottom right corner that we have there have all the same input parameters, except the location for the max thickness is slightly further forward for the top one than the bottom one. So if you look at it, the max thickness is probably somewhere around here for the top one and then for the back, somewhere around there for the bottom one. So what it allows us to do is it streamlines the design process, and all we have to do is type in a few numbers and then a new shape will automatically be generated for us.

So this feeds in very well the CFD, because often you're trying to analyze of a wide variety of shapes in order to figure out what's the best one possible. And if you have a streamlined process for generating shapes, it makes it much easier, because most of the work has to go into just the analysis side. So the performance metric that we had for this project was pretty straightforward. All we want is the maximum lift verses the drag ratio.

So most of the time, the CFD plots will be giving us lift versus drag curves. And just to illustrate this, what we want is this axis right here is lift, this axis is drag. So for specific operating conditions, maybe we want a coefficient of lift to be-- we think that the foil needs to perform well between coefficient lift of 0.2 and 0.4. So given this is the operating conditions, we just want these lines to be as low as possible, because that means for the given lift that we need out of them, they'll provide as low drag as we could get.

I guess I jumped forward there. So how do we design these shapes, or how do we choose which is the best one? So one potential possibility is to create and analyze every single possible foil shape in the world and then look at all these graphs, and simply pick which is the right one. So that's one option. I don't know if you guys think that that's a good option. If anyone sees something wrong with it, you should just let me know.

Or someone just yell something out. What do you think is bad about that idea? Yeah, it's very inefficient. It's very, very inefficient and costly. So I'm not doing this on some supercomputer with a million processors running at the same time. I'm just doing this on a laptop. So we want to try to be more efficient, or we want to try to attack this in a better manner. So what we decided to do was to set up an algorithm where we analyze a really wide range of shapes to start with, where we don't have that many different analysis going on but every shape is radically different from all the others.

And there, we'll pick the top few foils, the top few shapes from those batches, and then we'll run another set of analysis, figure out which is the best one, generate a bunch of shapes that are similar to those, and then run more analysis. And you keep going down, and eventually what will happen is you'll start to converge, where the differences in performance from mew shapes versus previous shapes don't really give much performance benefit. And at that point, you'll stop the algorithm, and you'll have your awesome foil shape.

So just to illustrate this a little bit, we were using Xfoil CFD to do all the analysis, and this is illustrated in the little video to the right. So basically, once you set up the program, we'll hit run, and it will know which foil shapes to start with and then how many times to go before want it to cut the optimization algorithm and tell us what we think is best. Just a little bit about Xfoil, it's not the fanciest CFD tool out there, but it lends itself really well to this specific application, just because it's pretty fast to run.

It's also really easy to automate the analysis side using a programming language. So for this project, we use Matlab, but you could use Python or other similar languages, similar programming environments to automate it. And like I said earlier, it really lends itself to large batches of runs. So if you want really, really detailed calculations on a specific foil shape, it might not be the best one to use, but if you're trying to run through a bunch of different shapes and just try to briefly figure out which is the best one, this Xfoil program works really well.

Now, once we've chosen the best foil shape, this example, like the foil that we had, all these numbers are the parameters that we chose to be best. So this is the bluntness of the leading edge, and then all the other ones relate to my previous slide. Once we have our optimum foil shape, now we're looking to bring that into CAD and create a solid model of the body.

So who here is familiar with Solid Modeling Fusion 360? You understand pretty well how that works. OK, so a couple of people. So depending on how familiar you are with this, sometimes you'll have more or less perspective to understand how this works. But how we start off is we import CSV files of the bodies that Xfoil told us would be best into the CAD environment. So this is just basically what we're starting off with.

It's a bunch of data points or a bunch of points from data files that we'll put into CAD, and then we'll use that and something called a loft command to turn this into a solid body. So this is kind of just a wireframe, if you will, but then this is a solid body that we can use inside the later stages of this project. So we also had to design something called a bulb, which for this specific application was necessary. Just it's extra material around the place that the foil attaches to the boat in order to give it more structural stability.

If we just had the foil shape right here, it wouldn't be as structurally sound, the attachment points. But this doesn't serve any actual hydrodynamic purposes. The perfect foil shape wouldn't have this, but for our design constraints we needed it. So once we have our solid model of the hydrofoil, we need to prep this for CAD design. So we need to prep this for CAM.

CAM, for those of you who aren't familiar, stands for Computer Aided Machining. And whenever we're working with composites, you're always going to have to make a mold in order to create your actual part. So what we had to do was make a two piece mold for this body, where we split it into two halves. So this is one half of the mold, and this is the other half of the mold. And if you close them together, the void between those two is the body that we need to make.

So once we define our solid body, we need to send this to a CAM software, which will allow us to program this big machine that will create these designs for us and allow us to make the molds that we then use to make the hydrofoil. So the CAM workspace within Fusion 360 is really nice, and it's definitely superior to some of the other CAM software that I've worked with in the past. So the best thing about it is it's fully integrated into the Fusion 360 software.

So in the past, when you've finished your design, you've finished the mold and you're ready to go start machining it, what you've had to do is download that solid body, and then open it in a completely different software environment. So you had to go load some competitor software which allows you to define the body. But the problem with that is it's really hard to jump back and forth between solid modeling and CAM programming.

And what would often happen is if you start working on your CAM program, telling the machine how you want to it to make the mold, if you realize at one point that something needed to be changed with the body, you had to just throw away that file, go back to your old solid modeling software, create the changes that needed to be made with the old solid model, download it, reload it into your CAM environment, and then start from scratch with the Cam programming.

So the nice thing about Fusion 360 is CAD and Cam are integrated within one package. All you have to do if you realize that there's something that needs to be changed with your body is just go to the tab up to the top and switch from model tab to CAM tab. If you start working on your CAM programming, realize something needs to be changed, go up there, do whatever needs to be changed, and then you can go back to CAM, refresh it, and then often you don't even need to put any extra work into the CAM programming.

So this is just a little illustration about what CAM programing looks like. It allows us to specify the machine process and tell the machines that are going to cut this mold or going to make this mold exactly how we want it to be made. So this is a little simulation that you run when you're doing the CAM workflow in order to make sure that the body is being cut in the exact way that you intend it to be.

And it's often important whenever you're working in CAM just to always check and make sure that you don't have any surprises in terms of how the body is going to be made. So this is just a little example of the different tool paths that are required in order to make this mold. So we started off with a facing operation, then went down into a roughing pass to machine the majority of that body, and then finished with a finishing tool path, where it just went back and forth over the surface a number of times in order to make it as smooth as possible.

All right, so now that we finished the CAM programming, we can move to the CNC router. And for those of you who aren't familiar with CNC machining, it stands for Computer Numeric Control. Basically, if I were to sum it up, it's like a giant robot with a variety of cutting tools on the end of it, which will cut anything that you want and make anything you want as long as you tell it what to do in the right way. So what we used for this project was the DMS 5 axis router inside Autodesk Pier 9.

It's kind of a midsize CNC machine. There are certainly bigger ones out there, but it's got a pretty decent size to it, and it was perfect for the scale of what we were doing in terms of this project. Just to explain this a little bit better, the CNC machining, we start with our CAM programming. So this model right here, we defined exactly how we want the machine to cut it. And then we'll load a program into the router called G Code, which details exactly how the router is supposed to machine it.

So you see here, this is a picture midway through the machining process, where the CNC machine is down and cutting away material from the mold, or from the stock material to make the mold. And then to the left here, we have our final finished mold, the finished part. This is just a little video of the machine process. So you can't see everything that well, because there's this protective cover around the actual tool itself to make sure that dust and other particles don't fly everywhere.

But if you look closely inside the middle, you'll see this little cylindrical thing. That's just a cutting tool. It's spinning really quickly in order to try to remove material. Post machining. The mold's pretty much done. There's just some minor post-processing that's needed. For composite molds, you need to coat the mold surface inside an epoxy or some other type of mold sealant to make sure that when we do our composite manufacturing, that we can remove the final part from the mold properly.

And then we also need to apply some other materials like specific wax that's used for mold releasing. But anyway, the main takeaway is that once it comes out of the machine, the tool is pretty much ready to go. You're 95% or 99% of the way there. So now we've made our tool, but we're not exactly sure how to make the final hydrofoil. We're not sure how to make the final parts yet, and this is Autodesk Tru-Composites comes in.

So another question. Who here is familiar with Composite Lay-ups? OK, a decent number of you. Nice. So this will be a bit of a review. But for people that aren't familiar, carbon fiber reinforced plastics-- often they'll just be simply referred to as carbon fiber. But whenever you're making a part or whenever you have some structure, some part that's made out of carbon fiber, it's never just carbon fiber. So carbon fiber by itself is this sort of fabric like material, where it's incredibly stiff and strong in tension, but it doesn't have any rigidity to it.

So if you have a piece of carbon fiber cloth, you can twist it and deform it and twist it around and do all these things to it. And you need to surround the carbon fiber inside what's called a matrix material in order to actually get the optimum mechanical properties from it. So often-- this is just a little example of this. So over here, we have two different composites.

One's called unidirectional, because all the carbon fiber is pointed along the same direction. This is called a woven carbon fiber fabric, because you can see the rods [INAUDIBLE] the carbon fiber are woven and intertwined with each other. But all the volume that surrounds the carbon fiber is the matrix material. And it's something that's rigid but not as strong and stiff as the carbon fiber itself, and it's literally just to try to lock the orientation of the fibers in place so that they can't do stuff like the normal fabric could.

So they can't bend and twist. And every time you put load into your carbon fiber part, it's only loading the fibers themselves in tension. So here's an example of what's called a carbon fiber Lay-up. Basically, you'll have sheets of carbon fiber cloth, so kind of like two dimensional panels that are stacked on top of each other in order to get the varying thicknesses or structural properties that you need. Often, what we'll have is either just a uni-directional panel where-- if you see all these lines right here indicate the direction the fibers are going. So all of these are lined up in the same direction.

Usually you want to use a composite that's oriented like this. If you know there's only one axis that needs to take all the strength and all the other axes don't have any structural requirements. So if you look at this panel right here, because all the composite fibers are aligned on this axis, there's a lot of strength in this direction, but there's not much strength if you were to try to pull it across the fibers, just because the matrix material isn't as strong as the carbon fiber itself.

The main point of the matrix being there is just to give stability to the carbon fibers, but it's not to add actual strength to the system. So when we have more complicated loading paths, so it's not just along one axis, what you'll use is a varying layout orientation, where you'll have different sheets of carbon fiber oriented in different ways. As you can see, all these degrees right here indicate the direction that the carbon fiber sheet is rotated relative to a baseline.

And what it allows you to do is you'll get more uniform characteristics in terms of mechanical properties. It's still not a perfectly uniform material, so it won't give you a simple loading characteristics of steel, for example, where with steel, you can pull it in the same direction or any different orientation, and it'll give you pretty much the same mechanical properties. This is closer to something like that but still not quite the same.

So the big question for a project like this is how many sheets of carbon fiber do we need to fill this complex geometry? So it'd be one thing if you were making a carbon fiber plate, where it's just this square panel that needs to be a certain thickness and needs to be a certain width. That's pretty straightforward. You don't need complicated engineering software in order to design for that, because it's literally just figure out how thick every individual sheet of carbon fiber is, lay one on top of each other until you get to that dimension, and then you're good to go.

So if you look at this mold, there's all these complicated contours everywhere. It's a lot deeper here than it is in the sides. And it's not straightforward to figure out exactly what these sheets of carbon fiber should look like when we put them in the mold in order to get our final parts' shape and size. So the solution to that problem is using software such as Autodesk Tru Composites. So what it allows us to do is to start with, we load our solid model mold into the Tru Composite software, and then we can determine the necessary lay-ups that are needed for our specific part in order to get get optimum mechanical properties and the thickness that's required for the strength that we need to get from the composite.

So if you look at this, this is just a lay-up scheme similar to the previous slide that I had, where we've got all these different composites oriented, or all these different sheets of composite oriented in different angles. Let's see. So for this specific part, we can say most of our load is going to be going along this axis, so we want more of the fibers pointed side to side there than are pointed longitudinally or on a different axis.

But because there is a complex load path going through this foil, we do need some fibers to be off that axis. And all that design process can be determined using the Autodesk Tru Composite software. So moving on, we have a few different manufacturing analysis tools at our disposal with the Tru Composite software. So one of these things is-- I'll just run through these. So Fiber Angle Deviation. Basically, we want to make sure that when we the lay the sheets of composite down, we'll suffer in our mechanical properties if the fibers deviate in angle too much.

So we want them as much as possible to just be straight so that they can be loaded along their axis. And if the software defines a lay-up scheme for us that has lots of weird geometry inside of it, the Tru Composite software will automatically tell us when we've deviated fiber angle too much, and we could potentially compromise the strength of the part. So Layup Templating is sort of self-explanatory.

So basically, it gives you templates of what these individual layups are going to look like. So what the specific sheets of composite are going to look like as we put them down into the composite tool or the mold. Fabric Wrinkling, it's kind of similar to Fiber Angle Deviation. Again, we want to make sure that these fibers are laid down flat and straight, and if you have a carbon fiber fabric where instead of laying down flat it compresses a little bit and wrinkles, again, it compromises your mechanical properties and you'll wind up with a part that's not as structurally sound as it could be.

So to the best of your abilities, you want to make sure that when you make lay-ups for your composite structure, you minimize fiber fabric wrinkling, and the software will tell you if that's potentially going to be an issue. And 2D Flat Patterns. Again, wherever we're doing these lay-ups, it's always going to be very thin, 2D sheets of fabric that we build up to our composite structure.

So here's just a little example of what all these individual flat patterns are going to look like that we're going to put down into the mold in order to get the composite part. Here's a little example of composite nesting. So when you're making these composite parts, you want to try to minimize the wastage of carbon fiber fabric as much as possible. And often, you'll start out with a large flat sheet of carbon fiber.

It could be six feet by eight feet, and it's just all interwoven fabric that you need to cut out in order to get the specific templates that you defined in the previous slide. So you need to cut all these templates out in order to lay them down into the mold. And this Nesting Software allows you to find the most efficient way that you will panel all these things together to minimize wasted material.

So all of this blank space outside of these patterns or these panels is pretty much just going to be wasted material. So you want to have these nested or set together as efficiently as possible. And the nice thing, we didn't use it for this specific project, but the Nested procedures, the Nested programs, you can export those as a DXF file to get a machine to cut these panels out for you, or you could do it by hand.

What we ended up doing was doing it by hand, just because in the prototyping environment that we're working in, it was simpler to do this for the low volume manufacturing. But in larger scale manufacturing, you often get a machine to try to automate this process of cutting out all these different panels, and the DXF file would be loaded into that machine. You'll have a cutter that goes out and cuts all these sheets out for you.

So now, last step, we're ready to fabricate hydrofoil. So what we used inside this process-- there's a bunch of different ways that you can manufacture composite parts. But for this project, we used Resin Infusion Manufacturing, also known as vacuum assisted resin transfer molding. The colloquial term for a matrix material when it's in its liquid state is known as the resin, because basically it's this thick polymer that hasn't been mixed with a hardner yet to turn it into a solid.

It allows you to work with the matrix material and spread it out amongst your carbon fiber parts. Then, once the time is right, it will harden and give rigidity to your composite structure. So the first step in this process was to cut out the dry carbon fiber fabric, which doesn't have any matrix material inside of it. So over here, you can see this is just a roll of carbon fiber that you're cutting out one of the templates from the previous slides.

You'll then take the carbon fiber sheets, and you'll lay it in the molds that we made on the CNC machine. And then after that, you'll seal the mold with a vacuum bag that allows you to effectively transfer the resin into the carbon fiber. So how this works is we'll use the vacuum bag and pull a vacuum on the carbon fiber.

So over here around this perimeter there's a bag, and then going in and out of it is these tubes that allow you to suck air out from inside the bag. Then what you can do with that is mix up some of your resin while it's still inside this liquid state. And then, you'll use a pressure differential from one side of this vacuum bag to the other to suck resin through the carbon fiber part and have resin replace all the voids of air in order create your matrix structure.

So once we've created the pressure differential and we suck resin into the parts, you'll have resin filling all the voids and uniformly distributing itself amongst the carbon fiber. And that will create a part with uniform mechanical properties and as few voids as possible which would disrupt the strength of the part. So anyway, the resin will harden depending on some chemistry that we use depending on some chemistry that we use for the manufacturing process.

So you'll determine beforehand how long you need the part to cure for, but there will be different hardners that you'll mix in with the liquid state resin that will determine how long it takes for it to harden. One it's hardened and you no longer have any liquid polymers inside your parts, you'll remove the vacuum bag, demold your composite part, and then you're left with your final product. So here is just a picture of the final hydrofoil after it's been cleaned up and polished and looking fancy.

All right, and then the last part of this presentation will just be pretty quick, but I'll go over some future development plans that we have with Autodesk. As much fun as building a moth hydrofoil was, ultimately it's not going to win an Olympic medal, because it's not an Olympic class. So what we want to do is take what we learned from that project and then apply it to an Olympic discipline called the Finn, which is pictured over here.

And specifically what we want to work on is something called the Finn rudder, which shares a lot of characteristics with the moth hydrofoil. So again, it's open to development with a few constraints in the design. And specifically, the design can be optimized for specific applications. So you might want to use a different rudder if it's completely flat water than if it's a very rough sea state like pictured here. Just different design aspects could influence why one would be more competitive than the other.

So just to summarize the performance goals for this we have. Ultimately, we want to optimize for the 2020 Olympic venue. And then on top of that, we want to optimize for particular athletes. So different athletes will sail the boats differently, and those nuances in terms of technique will influence the design. And then ultimately, we also think that broadly speaking, we can improve the current designs in a few key areas that will just give general better performance amongst a variety of conditions.

So design constraints for this project involve the physical shape obviously needs to conform to the class rules. And then the performance strength, as you want to have as low drag as possible for a wide range of operating conditions, and then be structurally sound, because we can't have athletes breaking equipment while they're racing. So where does Autodesk come into this?

Optimizing for the Olympic venue and optimizing for particular athletes, it's imperative that our DAP program is giving us reliable data so we can analyze and understand exactly how certain athletes will benefit from different design considerations. And then, we're also using laser scanning and CFD simulations in order to try to profile the designs that are currently out there.

And then Autodesk is going to be big help with the rest of this project in terms of all the design we're planning to do inside Autodesk Fusion 360, just like [INAUDIBLE] hydrofoil, and then we'll be using Autodesk CFD and Nastran for the flow and structural simulations. So anyway, that's all I have for you guys today. Hope you enjoyed it, and at this point I'm open to any questions.