Challenges for Low Volume Production

Brian:               Hello everyone. Thank you for attending this panel this afternoon. Hope you’ve been enjoying the show so far today. My name is Brian Levy. I’ll be moderating the session today. This panel is called Challenges for Low Volume Production. I’ll just give a quick intro of myself, and then I’ll turn it over to our panelists, and they can give a quick intro for themselves as well. As I said, I’m with Joe Gibbs Racing. If you’re not familiar, we are one of the teams competing in the NASCAR Cup Series. My primary job function is Aerodynamic Development, specifically Full Car CFD Simulations, and One Tunnel Development. But I’m also been responsible for overseeing our Additive Manufacturing Resources here at Joe Gibbs Racing. We’ve been involved with Additive for about 15 years now, predominantly on the Polymer side. And so balance out … my days balanced out between Aerodynamic Development and Additive Manufacturing. So at this point, I’ll introduce our two panelists today. First from Local Motors, we have Dylan Hoskins, and I will give him a few seconds too give him an introduction of himself.

Dylan:               Hello everyone. Thank you for attending. My name is Dylan Hoskins and I am the Research and Development Manager here at Local Motors. My background is all in Large Scale Additive Manufacturing, polymers and composites specifically. Most of my work revolves around hybrid optimization. So ad-living and subtractive on the same type of machine, and the challenges that come with it. I do a lot of material investigations as well.

Brian:               Great. Thanks Dylan. Also joining us today, we have Kenny Krajnik. He comes to us from Arrow McLaren SP. Hey Kenny.

Kenny:              Hi Brian. Hi Dylan. So as Brian said, I’m Kenny Krajnik. I work for Arrow McLaren SP. We are an IndyCar team based in Indianapolis, and I am a Design and RND Engineer with the team. We first started outsourcing our 3D printing needs in, I would have to say, 2016 and we got our first FDM printer in 2018. So we’re relatively new to Additive Manufacturing, but we’ve had a very, very quick learning curve and I’m very excited to be a part of the panel today.

Brian:               Thanks Kenny. Just a quick note, just so everyone … as a refresher. This panel is an interactive one. And so please feel free, if you have any questions, enter them in the chat on the side of your screen, and we’ll get a little discussion going with our panelists right now. But a little bit later we’ll try to get to some of those questions answered. So like I said, feel free to enter any questions you have on the side, and we’ll get started from here. So first off, we’ll start with a bit of an easy one for you guys. So this panel, of course, is title Low Volume Production. So for your two different companies, and two different industries, that you guys are with … Dylan, we’ll start with you. What do you guys consider to be low production?

Dylan:               So for Local Motors, and what Local Motors does is we 3D print the entire chassis for an autonomy shuttle, named Olli, at the moment. So right now production for us is around 50 vehicles a year. And a lot of what we are designed around is the fact that due to leveraging the advantages of Additive Manufacturing and Hybrid Manufacturing, that you can quickly go between iteration. So from one chassis to the next it could be slightly different. So low volume for a single chassis could be five to 10, but for our actual Olli it’s 50 per year right now.

Kenny:              So what we would consider low volume, at Arrow McLaren SP, dealing with racing and car racing in general, we’re typically talking about low volume stuff all the time, with every job that you do. And so for us, low volume I would say is two or three individual parts, and a high volume would be anything over 10, certainly.

Brian:               Yeah, I would echo what Kenny just said. I think over here, at Joe Gibbs, we are probably closer to where you guys are with over on the IndyCar side. Where low volume for us is generally less than 10 parts, maybe 10, but we do do some higher volume, quote-unquote, runs, where we’ll produce maybe 30 to 50 parts, but those might be something that we’re going to run on a car multiple weeks, or throughout the entire season. So we’ll do, if it’s a part that we feel like isn’t going to be changing much from the start of the season to the end of the season, we might print maybe 50 parts. So for us, 50 is a higher volume run. Whereas I know, and probably for most of our attendees, or for the rest of manufacturing, 50 is still incredibly low. So that’s … it’s interesting what the different uses you guys have for additive, and how you’re utilizing technology. So-

Kenny:              So, sorry Brian, let me just jump in and ask a question really quick. So-

Brian:               Sure.

Kenny:              Dylan, how many chassis did you say you produce in a year?

Dylan:               50 is our current benchmark.

Kenny:              50? Okay. And Brian, how many chassis do you guys have in-house?

Brian:               In-house in a typical season … Coincidentally, we’re pretty similar. We’re about 50. So we run four full-time cars, and each of our teams requires 15 to 18 cars per year, depending on scheduling, and unfortunately crashes, and damage, and things like that. But yeah, we’re generally in the 50 to 60 cars per year that we go through.

Kenny:              Okay. So we have six chassis total for two full-time cars. So I think that sets the tone for the scale of production that we’re doing amongst the three of us. You know, we’re-

Brian:               Yeah, for sure.

Kenny:              We’re doing very, compared to you guys, we’re doing lower volumes. And how many printers do each of you have? I know we’re talking about maybe some different processes, but Dylan, it sounds like you have one giant one?

Dylan:               Yes. We have a 15 by 40 foot printer and that’s our production workhorse.

Kenny:              And so we have a Stratasys F370 and 450mc. Those are the only two printers that we have.

Brian:               Yeah. And we, I guess we’re similar to you guys in that we have four Stratasys FDM printers and one of their object printers as well. Which actually, kind of leads into of my next question. So what kind of, what kind of parts, what kind of applications are you guys utilizing with the additive technologies? Kenny, why don’t you take that one.

Kenny:              Okay. So I would say the majority of the parts, the vast majority of the parts that we’re producing are for pit equipment, jigs, fixturing, test fit parts for a part that’s going to be actually machined out of aluminum or steel, or made out of carbon, we’ll print a test piece just to make sure it fits the way we think we want it to. And probably only about 5% of the parts that we produce, actually go on the race car as a functional part.

Brian:               Would you say, just as a follow up to that, would you say … Are you guys trying to get more production parts on the car? Or I guess a better way to ask the question is what, what limits you from getting more production parts on the car?

Kenny:              I would say the limiting factor for us is the rules of the IndyCar Series. There aren’t that many parts that we are allowed to make ourselves. And so, yeah that limits the possibilities we have. So right now it’s a lot of bracketry and electronics related things that are on the race car, certainly nothing structural.

Brian:               Yeah. I think we are in a similar boat in that we are limited in terms of the number of parts that we can put on now, number of parts that we can put on our cars. Now we produce a higher percentage of parts compared to over in the IndyCar side. But we’re still limited in terms of our rule book dictates what parts can be made of certain materials and utilize certain manufacturing processes, and so on and so forth. And so our rules in turn are one of our biggest limitations, as well as you guys. Dylan, what about you guys? What … It sounds like you might be on the other end of the spectrum in that the majority of your parts coming out of your machines are ending up on the production side.

Dylan:               Absolutely. So our entire vehicle is based around our 3D printed chassis. So really what limits us is weight, of course. With an electric vehicle, the more weight you add the shorter your range. The shorter your range more often you have to go to a charger. But a lot of it is just making sure that we are keeping the passengers safe. So crash worthiness is of the utmost importance to us. And however important it may be, there’s still a very long way to go before we really understand what it looks like, and be able to simulate crashes for the materials that we work with. We’ve done crash testing before, but naturally that gets very expensive whenever you’re having to produce an entire chassis just to crash it for a low volume manufacturer. So it’s entirely structural for what we use our main printer for. We do have some like 3D printed mounts and things like that, that are produced outside of our large scale printer. But for the most part, what we use our printer for is that main structure. And we just want to make sure people are safe.

Brian:               Yeah. That brings up … You touched on something that brings up a good point. So with the vast number of parts that you guys are producing that are ending up on production vehicles, what sort of steps are you guys taking to? … You could even start on the all of beginning in terms of qualifying the materials and the processes, and then validating the production parts, and so on and so forth. What kind of steps are you guys taking to ensure that safety is still a paramount importance?

Dylan:               So I’ll start off, I guess. What we do originally is we actually had a fairly large project where we went through multiple materials. We tried to understand what their print window was, their tensile strength, their modulus, their strain, strain to failure. So we went through a very long tensile testing, DIC testing to understand the Poison’s ratio. Three point bend, how adhesives work with it, how bolting works with it. Can you friction stairwell that if you have a crack, a lot of things to make sure that once you create a chassis, that it’s usable and functional, and not just something that if it’s damaged once or you hit a big bump in the road, then you’re out an entire vehicle.

So it was a very long thorough process just to qualify our material and qualify our printer, to where we were comfortable with starting to use simulations to have estimates of what we expected to happen. How much reinforcement we would have to have. What does adding in a honeycomb here do? Things like. So we went through a very thorough process of material characterization. And then from there we went into simulation, of course, using of course Additive Manufacturer, especially composites are extremely complex. They’re much closer to a fully anisotropic material which, as anybody knows, that’s a lot to try to actually qualify. So we simplified it into an orthotropic material and did assumptions based on that for a crash worthiness, stiffness evaluations and things like that, before finally actually doing physical testing of the chassis.

Brian:               And are you guys … Do you do any sort of quality control on the back end of validating parts on a certain time, or on a certain per part basis?

Dylan:               Yes, so actually our entire chassis is one, one print. So, but what we do is we print our wheel wells. So we don’t have to worry about support material or drop-in boxes or anything like that. And we’ll actually cut out each of those wheel wells, retain them, and make them either into tensile samples, three point bends, specimens. So basically we have a catalog of how every print performed during the process, and make sure that it’s available to us in case we ever need to go back and test things. But we do try to at least test some of the material after it’s printed. We might not go through a full material characterization kind of thing, every time, but we at least try pull a couple of dog bones off of each one and make sure it falls within the realm that we expect.

Brian:               That’s interesting. So you’re basically pulling samples out of the actual production part and using that as your test specimen for each chassis. That’s pretty neat.

Kenny:              That is really cool. That’s a good idea as well.

Brian:               Yeah. So we actually had a question come in from the audience. And just as a reminder, if anyone else has any questions, feel free to enter them at any point and we’ll try to get to them as best as we can. So for our panels here, how do you guys design or how do you leverage design for Additive Manufacturing for the unique capabilities of additive?

Kenny:              Yeah, I’ll go first on that one, Brian. I think having Additive Manufacturing capabilities in-house, for us really opened a lot of doors on the design side, that I wish had been available to us far sooner. We just … The budget wasn’t there for our team to get one any sooner than we did. But I think what it really allows us to do is get as creative as we want to. You can kind of throw out the rule book on that you used to follow on designing a part to be machine done, manual lathe and mill, or a CNC lathe and mill, for that matter. And you can get as creative as you want to, which makes it really helpful for me, for mounting electronics and sensors, and things like that. It just really opens up a lot of doors for that sort of thing. Getting something exactly where you need it to go, and not needing to worry about how exactly you’re going to get it there.

Brian:               Do you find that you can take certain … you can be a little more open with your designs, or a little more creative, or maybe take a little freedoms, because you know you might only be producing a handful of parts, versus working in traditional industries where you’re manufacturing hundreds of thousands, if not more parts?

Kenny:              Yeah, absolutely. Absolutely. Yeah, that’s one of the beauties of it. You can on each … let’s say you’re going to, you plan to do two or three parts as your final run of a single design. On the first one something might not work exactly like you want it to, but you can fix that in post-processing by either putting holes or bores wherever you need them to go, or removing some material here or there to make it fit the way you want it to. And then you can just update it on the next design iteration. And because it’s such low volume that we’re working with, it’s not a problem for us. So that’s very helpful to be able to do it that way.

Brian:               Dylan, what about you guys? I feel like the entire existence of your product would not exist without additive, but how do you guys leverage design for Additive Manufacturing to produce, to produce Olli?

Dylan:               So to start with, whenever you think of a normal like, let’s say the chassis, something underneath a truck or a car, you have to realize how many parts are going into that. And just the fact that we can simplify all of that into one part that we make. We pick it up and we set it down, and we can start bolting to it, bonding to it, whatever we’re doing. That’s our main way of leveraging it is, we can really reduce the number of parts in that structure of the vehicle. Another way is due to the fact that we are, we have slightly different rules. We really don’t have support materials in large scale printing. So we have to come up with unique ways to make structures, but the fact that, how long would it take Fortune 500 Automaker to switch from one body style to the next? For us, we could do it in a couple of months, with no issues, if not even shorter. And a lot of that would just be running the simulations.

So the way that we’re leveraging it is we can always stay ahead of what we need to do. So if we notice, or our chassis needs to be stiffer because this is going to be a bumpy road, or less stiff because it’s going on cobblestone roads in Europe, or something like that. In theory, we can adjust that very quickly. We can add support structures to it in a matter of days or weeks rather than years. So the way that we leverage it is being able to make minute changes to our print, and to our milling, to make sure that what we produce is always the best that we can produce. Not the best, not always just planning for the next model year, or the next major update to a vehicle.

Brian:               Mm-hmm (affirmative) Just as a follow up to that, and you kind of touched on it, but … So you talked about some of the ways you can leverage additive, but at the same time, and this is a question for both of you guys, but Dylan, you could take it first. What are some of the challenges that you guys face just dealing with additive technologies?

Dylan:               So, because large scale Additive Manufacturing and polymers has only been around for six or seven years really, started over at Arnel with, with the bam. So we’re still a relatively new technology. I mean, the technology that make it up have been around for … Extruders have been around forever, CNCS have been around for decades, but it’s combining those and having an actual fundamental understanding of it. How does a structure fail? Well, if you give me a piece of aluminum, I can go pull up all kinds of books and understand what that looks like. But really what limits us is the science that currently exists, and the software that currently exists to support our efforts. For us to slice something like a Strata Slicer could in a four by four square, and it would tell you exactly how much material you were going to use print time down to seconds, everything.

We’re still a long way away from that, from the software side. Right now slicing for us is very manual. And I think that kind of subtracts from that ability to really rapidly iterate. If I have to spend three days making a change for two or three print paths versus an FDM or SLS printer, where it’s, I want a different infill, triangles or hexagons, that’s really what’s holding, holding back a lot of Large Scale Additive Manufacturing. It’s just that fundamental science that needs to get there and catch up to what we have already. And the software.

Kenny:              Yeah. That’s amazing to me that you do all that work by hand, so to speak. It would be interesting to see your process on doing that, because it kind of still blows my mind. Because yeah, I have the luxury of just clicking the dropdown box and selecting exactly what I want.

Brian:               Yeah. You certainly take for granted.

Kenny:              Yeah.

Brian:               Even with additive, you still, you sometimes take for granted the capabilities and the resources that you do have versus really a niche process like Local Motors is using. So we can-

Dylan:               It’s a very iterative process. You know, you go back.

Brian:               Sounds like it.

Dylan:               Test it, hope it printed right. If not, select the surface in a slightly different section or offset it. Redo it, run it again.

Kenny:              Yeah. Very labor intensive.

Brian:               Kenny, what about on your end? What are the biggest challenges that you face?

Kenny:              So I think for us with Additive Manufacturing, I would say the biggest challenges that we have are post-processing the parts as they come out of the printer. So a lot of times they’re going to need inserts. They’re going to need some details cleaned up. A lot of times the layer lines removed by sanding or acetone vapor bath. So that’s kind of the biggest challenges for us, because we are such low, low, low volume production. Some of the challenges that Dylan has, it’s just not even in our realm of what we’re dealing with. So for sure, the biggest challenges we have are some of the post-processing things that need to go into it, and how that plays into the timeframe that we’re usually working on. Which is a very, very short time span usually for the time part or design is requested, to the time it needs to be put into action. You don’t always have a lot of time to do that post-processing, that you need to do.

Brian:               Yeah. I would agree with that. We’re in a very similar situation in that. We often refer to it as rapid prototyping, right? But you know, sometimes it’s not rapid enough. And we often talk about the race is going to start on Sunday at one o’clock whether or not we’re there. And so for us, we know that we need to get parts done and out the door by Thursday afternoon, or Friday morning when the trucks load up. And yeah again, whether or not our part is done, the race is going to start on Sunday afternoon, whether or not we’re ready.

But yeah, I would also agree with what you also said, in terms of the challenges that we face are post-processing for sure. Getting parts cleaned up, getting inserts installed, maybe some secondary processes, getting those completed on time. But I think one of the other challenges that we face is just in terms of finding materials that have the properties we need.

Kenny:              Yeah.

Brian:               Whether it be strength properties, or thermal properties. We’ve tried to put some parts in close proximity to maybe exhaust parts, or close to the brakes on our car, where the temperatures are really, really high. And so there’s … Luckily there are some good FDM materials out there that can allow us to get some parts close to those other, those heat sources on the car. But it’s definitely something we have to be very mindful of when we design some of our, some of our additive parts that end up on the cars.

Kenny:              Absolutely.

Brian:               Let’s see. So we’ve got a couple audience questions, in the few minutes we have left here. So this one I think is probably geared more, mostly towards Dylan. So one of our attendees wants to know, how economical is it in terms of cost, time and material properties to make a whole chassis compared to conventional methods?

Dylan:               So I think a lot of this also has to do with, again, what I said is that iteration. So for us to change an iteration, how much would that cost GM or Volkswagen? Hundreds, thousands, millions? I have no idea. I … Numbers I can’t fathom easily, but I think that’s where we see our biggest cost savings. But we can produce a chassis full-time less than 30 hours. Before it is start, on the printer, to bonding, 30 hours. And that’s printing the upper, lower and milling both of them to their final shape. And then after that it just becomes how fast are people working? Are we working overtime? Are we working…? So as far as the Additive Manufacturing part goes though, it’s around 30 hours for us to produce one chassis and we’re only getting faster.

And as far as the properties. The properties, I think still have a long way to go. But economically I think we spend less than $10,000 a chassis in material. And that’s very economic, in my opinion. By the time you would account, how long does it take welders to weld a chassis, bolts, rivets, nuts, man hours, one person can run a printer. How many people does it take to build a chassis?

Brian:               Yeah, that’s … that all makes a lot of sense. So I think we probably have time for just probably one more question. So Kenny, this one is probably more towards you guys. How do you decide when to produce a part using additive versus conventional manufacturing?

Kenny:              Okay. So I think there’s quite a few things that we first take into account, so like major considerations. The first one’s going to be the strength that’s required. So that’s obviously a huge consideration that we put into it. How strong does the part need to be and how complex does the part need to be? Because that’s something that really plays into the hand of Additive Manufacturing. The other thing that we’ve just talked about a few minutes ago is the timeline of the project. So how quickly do we need this? Is it going to take two or three days to machine it, or could we get it done in 10 hours on the printer? And then there’s also the physical size that we need to take into consideration.

Brian:               Yeah, that makes a lot of sense. I would say from our standpoint, we’re pretty similar in that we consider all those things as well. When … We try to produce as much using additive as we can. The advantages are clear. But yeah, I mean, we just have to consider all those things at the end of the day and decide if it makes sense, or if it just makes sense to produce something using conventional technologies. Well, unfortunately we are about out of time. So I would like to thank our guest, Dylan from Local Motors and Kenny from Arrow McLaren SP. I’d also like to thank our audience for attending, and please be sure to step over to the Aerospace or Automotive Chat Room, to connect with some of the other attendees and continue the conversation. And check out some of the other panels that are going on throughout the day, today and tomorrow. And hope you guys enjoy the rest of the show. Thanks everyone.

Kenny:              Thank you.

Dylan:               Thank you all.