4: Simulation Celebration and A Happy New Year!

Hello. And good day to you from episode 4 of our new podcast series Project Breakaway. A metaphorical and literal time in the day when we here at Predator Cycling take some time away from working in the back of the shop to come and share with our listeners what we're doing, how we're doing it, and what it takes to do it. Our ideas, our innovative success stories, and even our missteps and failures. If you find yourself with an interest in bicycles, composite manufacturing, out of the box design, or even curiosities beyond, I encourage you to stick with us, settle in, and learn a little. I'm Courtney B, co-owner and project manager of Predator Cycling. I'm here with my partner Aram Goganian, the other co-owner, CEO, lead designer, and engineer, and pretty face of Predator Cycling. Aram, how's it going over there? Ooh, I got a pretty face. Yes, I yeah. On a good day. Uh, I'm I'm doing really well today. Just had the holidays. Yeah. We survived. We did. We survived the COVID, uh, holiday. The Zoom I think I counted seven, uh, uh, Zoom, uh, Duo, Facebook Messenger, what do they call Facebook? Messenger. Messenger. Yeah. Video Messenger. Yeah. So many calls. Yeah, it was pretty hectic. I mean, we didn't have to get dressed. I didn't have to wear makeup. Mm-mm. But man, It's a lot. It's just like just, you know. And time zones and time zones, different countries, the technology, nap schedules the nap schedules and the technically challenged family on the other side. Yeah. So, it's a lot of, uh, tech support tech support and a lot of, uh, slow talking at the screen. Do you hear me? Yes. Yes. Yes. Anyway, we And then doing it all over again. Yeah. So, we survived. And here we are. Back at work. Yep. So, we are finally going to touch on design simulation today after a 2-week pause. Ooh. Um, so Aram, what what is it? What is design simulation? How do we use it? Um, how is it's evolution in the bike industry and more specifically, how has it catapulted Predator here and how are we going about using design to improve our concepts? And finally visualize all these crazy theories in your head and to, you know, use the sh shape optimization for practical use. Yeah, so there's I mean, there's the way we use simulation here, we kind of use it in in two fundamental ways. One is in this like design simulation that you were talking about and that's what we, um, we refer to that as in the early process for us. So, when we're, you know, conceptualizing sketches and getting stuff together and, uh, when we first start on a project and a and a new product, um, introducing simulation at the very very beginning. Um, that's what we're talking about design simulation because that quickly, uh, guides us in the direction of the way the product should go. Um, So, you're not a you're not an artist in in a sense of paper, pencil. No. No. So, you don't you have let's say Aram comes and he has an idea. Well, that's silly. Aram always has an idea. He doesn't come. And he probably has a portfolio, uh, in his mind of a bazillion things. Uh-huh. But I only allow him to work on maybe four five things. Yeah. So, you have an idea for a bike. That's that's on a You're not that crazy guy drawing a bike on a paper napkin. What are you doing? I I I go straight into, uh, into, uh, design tools. So, I go You open your computer. I open my computer. I'm the guy that opens up my computer and I go into, uh, Fusion 360 or into Rhino and I start modeling. You're the guy in college who like, well, I'm talking college early 2000s who had the laptop taking notes. I'm here in 2020. I handwrite everything before I type it up. That's not you. No. Okay. No, I don't. You don't want to see me handwrite anything. So, how do you start? You have a design in your head. I have a design in my head. Sometimes I'll do something on the whiteboard. Usually, I'm I do it on the whiteboard if I'm trying to explain it to you and how we're having a conversation about it and I don't want to pull it up. I just want to let listeners know that the whiteboard to me is just a bunch of chicken scratch. It's pretty bad. Um, but yeah, I sometimes use a whiteboard but it very seldomly. Uh, I go straight into, uh, into Fusion or Rhino and I start sketching out concepts. Usually, I start with like, you know, like on a bike, like on our our bike frames, I have a a, uh, parametrically driven 3D sketch that What does that mean? I don't know that means. It's a sketch that's basically I can update. So, if I want to change geometries, I can change the angles of the So, if you want to make a bike your size for very long legged people or you want to make a bike my size for very uh, average. No average. I'm not short. Perfectly your perfect In every way. Perfect. In every way. Um, you can just make that and you can just make those changes. And those are kind of like the hard points. So, like I'll adjust the bottom bracket height, the the headset angle, like a head tube angles, headsets, what the races, what style I'm using, like those kind of hard points I'll update instantly in my drawing that I've already built and then, um, I will I will go into, you know, into T-splines or into sub-D's or whatever modeling system I'm going to use, how I'm going to design it. And T-spline is what? Uh, T-spline is a is a proprietary software, uh, that's inside of Fusion. It's, uh, It's proprietary by Autodesk. Autodesk owns it. Um, they bought T-splines, uh, five or six six seven years ago, maybe maybe more. Um, and it's a sub-D modeling so it's, uh, And sub-D is what? Um, Some dimensional It's yeah. So, you basically it looks like, um, it it's basically using plane surfaces to affect complex curves. So, you can basically it's like clay molding in modeling. So, you can take a a block and actually just grab and push and pull it and move it in directions that you want and create surfaces and angles and finish. But how do you make a frame of a bike? Uh, I mean, you don't just start pulling off of a mold making a tube, do you? You have to Yeah. So, I have an an I have I mean, Like a base model? I have a base model basically I have a a system that I basically start with and I have all of my surfaces and, um, like high tol- tolerant like my everything that needs to have a specific tolerance like a bearing race or a a, um, uh, a a thread fitting or some sort of clearance issue, I already have surfaces built for those. Just the industry standards. I, well, it started with industry standards and then it became And now it's the Predator standards. Okay. Um, and so I basically just evolved that and have a massive library of models. I mean, I have I mean, I think I have two or three hundred model sub-models now. And you model everything down to like the screw. Mm-hmm. Which is I think Joe Schmo looking at your model online is like, "Oh, it doesn't look right. You pulled this from somewhere." And I 'm like, "No, you actually sat down and designed that." Yeah. And you actually talked to a real designer which I I know I'm not making this up because I sit and I listen to all of his conference calls and actual designers give him many many compliments. Uh, yeah. So, we we I More so than me because I'm Joe Schmo going, "I don't know what you just did." Uh, yeah. So, we model everything. I and I I actually model in my threads as well in most cases. So, um, which pays it it plays a really big factor when you get into the simulation side. So, we have massive libraries of models, assemblies, um, all kinds of stuff. And you just don't model I mean, we're talking about bikes right now specifically but you model everything. Yeah. You've modeled the floor plan of our shop. Uh-huh. You modeled our son's push bike and threw it in simulation in case you want to know how fast I actually threw it in roll down a hill. I I did that. I also ran simulation on, uh, CFD simulation for our air vents in our shop. Oh. Just for fun. Okay. I didn't know that. Uh, yeah. I done it just cuz, uh, Sounds like a great time to me. Uh, I was testing out one of our new, uh, softwares and I didn't know what to do it with so. Okay. Um, so we're talking about simulation. So, you make a bike in Fusion or Rhino. Yep. And then what? Um, well, so once we have like a rough shape of what we're trying to do, we then I I typically will actually I'll I'll go into, uh, ANSYS Discovery, um, which is a a real-time simulation engine. It's a it's a really cool program. You can design in it as well. Um, I'm not as, uh, I'm I'm not as quick as designing in it as I am in Rhino and Fusion. So, I usually start in Rhino and Fusion, model out what I'm trying to do, bring it into ANSYS, uh, Discovery, and then in there I can actually run simulations. I can run CFD, I can run, um, And CT CFD uh, CFD it was computational fluid dynamics which we discussed which explained Uh, so it's basically it's basically a virtual wind tunnel. So, I can look at Is it just a wind tunnel or does it tell other other things? Well, so the CFD portion of it is essentially for how we use it in the bike industry is essentially a wind tunnel. Um, we can also in Discovery run, uh, thermal dynamics, we can run, uh, str- static loads, we can run topology optimization, we can run other types of simulation which is weight weight normal people speak weight Uh, weight forces potential source in a crash Yeah, buckling loads. You're talking thermal so what leaving your bike on in the sun for Yeah. Like we wouldn't use I mean, thermal is not really relevant to bike industry. I mean, you there's not very many relevant we don't look at thermals inside a bike. Now, we do do thermals for our molds, all of our thermodynamic for our molds are all calculated in disc- I use a lot of in Discovery. And we have learned that's very important. Yes. Uh, thermal dynamics is very important when you making molds. Uh, yes. Uh, yeah but so we can actually take like a for instance like a head tube, like the RF20, one of the things that we spent a lot of time on was the head tube design of it. And really looking at how those bearing races are and the the the profile of that head tube and really making it lean and transition into other tubes. Now, in typical simulation structures, you have to you model this all out, you build your assemblies, you then throw it into a modeling engine, uh, a simulation engine and and then you you set it all up and then you run it and then it usually takes anywhere from a couple minutes to a couple hours maybe even days and how complicated the part you're running and how sophisticated your hardware is. Um, and then you take those results, go back into design, update, change, and repeat. Right. And this can take days. I mean, this can take weeks, this process. Um, years actually. Yes, it could. It could take a totally it could take years. Um, so a simulation early in this design eliminates, uh, any possible flaws that of performance that you might find after manufacturing. Yeah, and you can also it can also speed up that process. Like we can actually now design and simulate within minutes. I mean, we're within seconds we're getting feedback. So, when people look at your new frame your RF20 and they say, "Why is that so skinny? Why is that line like that?" You can actually and say, "Well, I threw it in a test and let me tell you why." So, and and it also helps you like kind of understand what direction you're going to model in early on. So, like I early we very early on there's on the down tube of the RF20 there's this like ridge that comes off of the bottom race of the head tube and goes down the down tube. And that ridge came from CFD simulations. Right. It came from the understanding on how the the airflow is going over that tube in that fork area and at the same time trying to build mass so we could get some torsional stiffness on the front head tube. So, that area was just the all of that shape there comes from uh, from both structural and, uh, CFD simulations. And those areas have gone through change how many changes? So many changes. Oh, yeah. It's I mean, it's it's I know there's photos online available of the RF20 and it's had so many iterations because we use this bike as a teaching device in our design software. Shows that we've done at Autodesk University that you've done in other webinars ours so We just did one with KeyShot and Right. So, you know, there's a lot of, uh, information on this bike that is, uh, both, uh, current and old. Yep. So, we're really working on the the frame still. We have the final design now. Yeah, the final design is there. It's it's actually not on the site. That's just bad marketing on our end. Uh, but we we do. We're working on the bike and we are in in coming up here in the new year releasing the final and full version. It is a a labor of love and a work in process. And, uh, we, you know, And it's really I mean, And the thing is we invite anyone to contact us if you have any questions about this frame. For sure. Reach out to him. Aram will talk to you for hours about it. You just Yes. I mean, I might actually tell him to hang up on you and get some work done but Yes. Yes. Yeah, no and Any questions, please reach out on any social media. Yeah, for sure. Cuz I'm always on it. I'll respond. Uh, yeah. And the other thing too is the RF20 is is really a it is a product itself but it is a it is one of our new manufacturing platforms and showcasing what we've been working on for all these years. Specifically in the design and simulation realm. Design, yes. So, Because that's where we're at now in Predator and going forward with most of our products. Yeah. So, speaking of more products, Yes. Let's talk about the water bottle cage that we're working on. Yeah. So, kind of a little so when we talked about this the simulated design for the RF20, the RF20 started as a model and then we introduced simulation to it very early on. The water bottle cage is is really interesting because it started in simulation. It's a product 100% designed specifically in simulation for simulation by simulation as a cool art piece Yeah. And a functional generative design. So, discuss generative design. Okay, so in generative design, um, and when I'm speaking about generative design, we're in most cases talking about using it inside of Autodesk tools. Um, and we basically within most Autodesk tools you can use generative and generative allows you to define, um, um, loads, um, mounting locations and restraints on a product on a on a dev- an an object and then let AI and generative design Artificial intelligence. Yes, artificial intelligence, um, allow you to create a part. It's cooler when you say it like that by the way. All right. Artificial intelligence. AI. So, and and that's what generative does and it's incredibly interesting, it's incredibly powerful. Um, it kind of takes your bias away from that you have as a designer and engineer. Um, and it's one of the reasons I like it so much. Um, but you can really use the software to help you leverage your materials, your property, your manufacturing process, um, everything to design a part. So, in layman's terms, you put in your hard points, your mounting points, your weight requirements, etc. And you throw it into the software and it spits out a bazillion different iterations for you. Yes. And from those, You can basically filter it out based on different parameters that you have and work on it. And the way I way we it's a relatively complicated workflow but we start there and then we get, um, a model that we like. And then from there we do some some some Massaging. Massaging to make it work a little better and get rid of some surfaces thing issues and to look more like a Predator design. Yeah. You work on it. And also just to understand like some of the load forces and stuff. And then from there we actually go into, um, ANSYS Discovery. Um, and in ANSYS Discovery we can fine tune well, one of the restrictions you have in in Fusion is that you can't put in custom materials. So, we need custom materials. Um, so we go into ANSYS and actually design and build out all our custom materials and then put that into the parts and uh, finalize our analysis and we use their topology optimization which is very similar to generative because you can directly input it based on the results. So, you can the way generative kind of like builds these these structures, you can in ANSYS because you're watching it live happen, you can actually as it's live running, you can go in there and say, "Okay, I'm going to allow more material here. I'm going to change this." And you can make change design changes as it's simulating. Right. Um, and then that's how we finalize that design. Um, Now, a water bottle cage was not on our radar. We weren't like, "We're going to make the coolest water bottle cage ever." No. We were teaching a design, uh, class, a generative design class on for Autodesk University. So, we decided to make a water bottle cage. Yeah. Good as an example. And it is um, not going to be made made out of, uh, composite carbon. We're specifically designing this part to be 3D printed. Yep. Why? Because it's cool. But you can explain better. Why? Because it's cool. Uh, so a couple things. Uh, we like I think we've talked about in in I know we've talked about in previous episodes, uh, we use 3D print here a lot for, uh, tooling and setting up. And we've always thought of using it for, um, direct printing of products. It's always been something we've talked about. Uh, we've never done it. Um, one of the issues we never did it is we couldn't optimize it to the level that we needed to. Um, and, um, with recent new software and ANSYS and new hardware that we have now here, um, we can we can, um, and so we decided to give it a shot and the water bottle cage was the first thing up. And we actually got it down to price point so that it made sense, uh, and could be manufactured at scale. And the other thing that you think about is is, you know, we mostly do everything in composites but we do it in composites because that's the best way to make it. Um, not because we just want to make everything in composites. Um, And a carbon water bottle cage is just not like an optimal product. It's not a cool product. It's very cool. But, you know. And people make really good water bottle cage out of composites. But when you start looking at the weight of the product, the durability of the product, Functionality and the price point. And the price point. When you start balancing all of these things, you can achieve most of all of that outside of composites. Um, and with, um, with with direct 3D printing of products, you can actually make a really cool cage. And it's the generative design quality of the water bottle cage is just, uh, complimentary to the RF20 because the RF20 does have generative design base parts in it that we, you know, massaged and made our own. So, it kind of complements, um, Yeah, for sure. Them. Um, it's also, yeah, it's a very it's an interesting and clever way to make a part and I think this for us is going to be a gateway into many new parts. It's scalable. Yes. It's competitive against composites in terms of drink durability, mhm, and it's competitive because of price. Yeah. Um And it's competitive against and you have to also realize that I mean I think there's only one or two water bottle cages that are composite that are made in the states. I I think there's only one. Um there there may be another one, but m- 99% is made overseas um parts and And there's a there's a multitude of reasons that we made this. First of all, it everything kind of sort of aligned. We partnered with amazing software and hardware companies that made the design um easily accessible to us. Yep. Um we got into more and 3D printing, higher quality 3D printing, everything aligned, and COVID happened. Yeah. I mean it created a a bunch of issues in supply chain Oh yeah. parts for bicycles and the entire industry. Um there was a lot of things um the material all started being poor quality. Yep. The injected plastics of previous water bottle cages had faded colors, they didn't match. Um we had customers that weren't pleased with the quality. Um it has affected our brand, so we decided let's change it. Let's make our own. Yep. Let's do it here in the United States. Well, and yeah, and exactly and and that whole the supply chain issue started coming up and it was just like and we started looking at some of the prices too that were going up and it's like wait how much are you charging for this ki- like this doesn't make sense. Material prices in general are just ridiculous right now. It was just like yeah and but like I'm looking at some of these prices and I'm like this doesn't make sense. I'm not going to pay that much for this. So And and we're talking about our industry, our boutique bicycle industry. It might make sense for other countries with for sure. But for us, we can scale this we can scale this 3D print type of components going forward. Mhm. In addition to our carbon fiber composites. For sure. So, that's pretty cool. Yeah, so we're super excited. So you start in the next couple weeks you'll start seeing um our uh some designs coming out. Yeah, we're running test parts right now. Yeah, we've been printing out test parts and getting production batches ready. We're learning our way around this 3D printer. Yes. Yeah, we got a new printer, so um and then the other thing that we should I mean our goal is to have them for sale next month. Yes. Um end of the month we should On the website. Should be And probably also available on Amazon. Yeah. Um so yeah, we're excited. We're super excited for this. Um and it is the first part that we have, like I said before, is that is entirely designed in simulation and AI Oh. I silenced my phone, but I did not un-vibrate it. If anyone's curious, my son's doing well at daycare. Um let's see. So we're talking about did you want to mention our partnerships with companies that Uh yeah. So yeah, I mean we've been working with uh I mean it's no if you follow us on anything it it's no surprise. Uh we've been working with uh Nvidia now for Few months. Uh yeah, six six seven months now we've been working with them. We were one of the first ones to test out the RTX A6000, which is a monster of a GPU. Um and we've been using that and that led to a partnership with Lenovo. So we've been testing out um their P620 uh which is a uh monster uh workstation. And then uh also we moved all of our laptops over. We're using the uh P53 53. P53. I'm looking at it right now because as you hear this podcast because of my Lenovo P53. Uh yeah, so we're using their um hardware. Running on my Adobe Audition. If Adobe would like to also give me some software, that'd be awesome. Uh yeah, but it's been I mean one of the reasons we've been able to scale out our our design simulation sta- work is because we have hardware um that can leverage it. Uh Absolutely. I mean none of this was This literally I think I said earlier we were catapulted Yes. I don't know why I enunciated that so much. We were catapulted uh because of this uh amazing partnership. Yeah. These partnerships. Yeah, no it's been it's been um Because we I mean I attempted to try this um about a year ago, a year and a half ago. I tried to use a lot of the same technologies that we're using now um to figure out how to make this and make it cost effective. And every time I did it, I s- I mean I spent more time than I should have on it and I could never make it scale. Like it just didn't work. I I couldn't make it work. Um and the limitation that we had was strictly well, it was some software limitation we had at that point, but um it was the biggest limitation we had was our hardware. Our computers just couldn't figure it out fast enough to give us the results that we needed. We literally set like a simulation to run and then we would just go home and come back the next day. And I would still and it still didn't it still hadn't finished Right. figuring I mean Or it had frozen and crashed. Yeah, that too. But now I'm just like hey babe come over here. Watch this. And then it instantaneously Pretty much yeah. Which I love Yeah. Because I have little patience. Yeah. No, it's it's uh it's it's pretty ridiculous. Um we're able to compute so much work now um that we we're able to do it and that's entirely why we're actually pushing into this 3D printing space more is because we now have the hardware and software that we can leverage it. Right. So what um can we look forward to on future podcasts? Going into the new year, we plan on inviting some of these guests. Yeah. Some of these partnerships, um software companies, hardware companies, some of their reps and um friends that we have with them. Um as well as maybe some former customers. Yeah. Current customers. Yeah. And if any future customers on like be on my podcast. Buy a bike. Um and then I don't know. Maybe maybe we'll invite your mom on. Yeah, why not? Former owner of Predator Cycling. Yeah. Um anyway, we're going to figure out that whole calling in thing uh next year. We're good at a lot of things, but audio engineering is maybe further down on the list. Yeah, no. Maybe like fourth or fifth. That's pretty high I would say. Yeah, it's pretty high. That's pretty good. Okay, well we thank you for choosing to take some time with us and we look forward to future breakaways. I'll try my hardest to put any mentions in our news section on predatorcycling.com. Also, look for us on Instagram and LinkedIn, Facebook, Twitter, and in person here in Tennessee. Aron, let's go get back to work. Please share, like, and subscribe. We're available on all major streaming platforms. Anyway, thanks for listening and have a good one and find some time to break away.