How Can We Design Sustainability into Additive Manufacturing?

KIERAN MAK: Hello, everyone out there. Thank you for tuning in. And while we wish you could have joined us in person at AU 2022 in New Orleans, we really appreciate you taking the time to listen to our class virtually today.

My name is Kieran Mak. And along with my co-speaker, Ryan Abel, we'll be trying to address the question that you see in front of you. During this talk, we're going to look at best practices for achieving more sustainable additive-manufacturing outcomes through design.

Now, before I jump in, I do want to acknowledge that we may make some forward-looking statements in this class. And while they reflect our current plans-- and, as far as we know-- they're not a promise or a guarantee of any specific results or product being delivered on a specific timeline or at all.

Now, before I get into how to design sustainability into additive manufacturing-- also known as "3D printing"-- I want to ask the question, is additive manufacturing even something that is sustainable? And, with that, I'll hand it over to Ryan.

RYAN ABEL: Thanks, Kieran. To help set the stage here, and in an effort to ward off the worst and most dramatic effects of climate change, the IPCC has set the goal of limiting the average global temperature rise to 1.5 degrees above preindustrial levels. To do this, the solution proposed in 2015 and targeted by a global community is to reduce the global emissions by 45% by 2030. And this helps achieve that net-zero-emissions goal by 2050.

Something more tangible for designers and engineers in the audience is the understanding that approximately 80% of all product-related costs-- and, in this case, even environmental impacts-- are determined during that initial, critical design phase of a product. Throughout today's session, we will discuss and demonstrate how to make upfront decisions that help the environment, especially in a world where, despite our needs to reduce emissions, we are being asked to continue to grow businesses and produce more. Industries will only change by the hard work of many people doing the right thing-- and not just doing the easy thing or the profitable thing to meet short-term goals.

Products go through basic life cycles, from raw materials to manufacturing to the use of those products in the field and eventually disposal. And this has traditionally been a very linear process-- and one that the company that made the product didn't really participate too much in, towards the tail end of the life cycle. It's becoming more important to think of these products in larger contexts. You know, what are the environmental impacts at every stage?

Energy and waste are consumed and generated at each phase of the product life cycle. These things cost money. And if we can work with less material, we are already ahead of the game.

Enter the concept of circularity or circular economy. Companies that consume their own products as part of the raw materials to build new products will lead, in this new economy. This includes designers working on products that are designed to be disassembled, repaired, and returned to the manufacturer, where they can reuse it, and even paying the consumer for participating in the economy-- in a sense, buying the materials they need to build tomorrow's product from the customers that bought yesterday's.

This will create a two-way marketplace and give the consumer and the customer some skin in the game. And this will help, in the end, grow brand loyalty for the company, as well. Back to you, Kieran.

KIERAN MAK: Perfect. Thank you, Ryan. So while there is an urgent need to reduce emissions, there's still demand, as Ryan talked about, for economic and industrial growth. One example of that kind of growth is in the additive-manufacturing industry. Last year this industry was valued at about $13.9 billion, and current projections have it continuing to grow at a roughly 20% year-over-year pace, to reach an estimated $52 billion by the end of the year 2028.

So with this in mind, the goal of this talk is to be able to provide you with guidance towards sustainable best practices in the additive industry. Top professionals like you make these kinds of product decisions that are right for both the environment as well as the needs of businesses. So now, how do we do that?

Well, the goal is to have you walk away from this talk being more equipped to identify applications that provide business and sustainability benefits, assess additive-manufacturing design trade-offs and avoiding common pitfalls, applying DfAM, or Design for Additive Manufacturing, techniques such as generative design, latticing, and part consolidation, and ultimately using those to drive more-sustainable design workflows and higher-performing part designs.

My name's Kieran Mak, and I am a solution engineer on Autodesk Advanced Manufacturing Team. My focus is on additive manufacturing as well as our generative design technologies. And what this means is, I work with our customers to help them leverage these tools to improve their design and manufacturing processes.

RYAN ABEL: And I'm Ryan Abel. I work on the same team as Kieran. And I'm a mechanical engineer as well. I've been in the field for over 22 years and really hail from the product-design world, where I worked at Lutron Electronics and, after about six years, really got into upfront simulation. And that's where I've spent the next bulk of my career.

So my specialties are really the automation, design-automation, insight you get for a performance to make upfront design decisions, driving new product introduction. And I've been spending more time getting up to speed on this new topic of remanufacturing and how it can benefit our customers and Autodesk.

The three main pillars of sustainable development include economic growth, environmental protection, and social equality. To support the pillars, the UN commission rallies around "meeting the needs of the present without compromising the ability of future generations to meet their own needs." The commission has received some pushback that this is vague and hard to measure and to act on. So let's take a closer look at what that might look like in a graphical form.

In preparation for this presentation and just learning more about the challenge in front of us, there's a lot of data out there, and trends, and predictions. And this is my summary of the information I was pulling from the industry. And you're looking at where we are today.

We're on this big downslope of mass consumption of resources. Population's towards the tail end of a big incline. And pollution is still on the rise, as well.

What's important to realize is that, due to low birth rates in the current generations, there is a prediction of population drop that will continue for quite some time. But what you notice, if we hit that 2050 goal, our resource drain, our 2050 goal of net zero, our resource drain can be very well controlled with the implementation of a circular economy. These are solutions to achieve this type of a forecast.

It's also important to realize that what happens after 2100 is really an unknown. Population can go back up again. So the needs of the future aren't really even understood, there, which makes the importance of being a good steward for the resources that we have even more important. But I like to make the point that we are facing a short-term challenge, and our ability to achieve progress in the short term really does set the population up for long-term sustainable industry.

In 2015, the United Nations member states adopted the 2030 agenda for sustainable development. At its heart are 17 sustainable-development goals which are an urgent call for action by all countries in the global partnership. Today, the division of sustainable development goals plays a key role in the evaluation of UN system-wide implementation of the 2030 agenda and drives advocacy, outreach, awareness, and sharing of best practices across industries.

For today's class, the goals that are most important are goals number 9 and 12. 9 covers industry innovation and infrastructure-- and specifically, point 4.1 covers CO2 emissions per unit value add. This is kind of vague, but it does give a way to measure product versus performance metrics that is focused on impact of greenhouse gas vented to the environment.

So it's a way to look at, what do we gain? What's the value we gain from these emissions? And it's a way to really track our progress and evaluate across industries. Whereas number 12 covers responsible consumption and production.

There's a few here that are really interesting. 2.1 covers material footprint per capita. This helps us right-size our decisions, based upon the size of the market and the growth of population. So it adjusts with the changes in population, as well.

4.3 covers hazardous waste. And this is important for additive manufacturing. And Kieran will cover some of the hazardous waste that's created. Because not all additive processes are the same, as far as what the waste or energy consumption is.

And finally, my favorite is the 12.5.1, which is establishing national recycling rates. And this is something I know the United States can really improve on.

So, moving away from the UN goals and numbers and language, let's get back to engineering-and-design basics. Because what you're thinking about is exactly what's important. We need to use materials better, while limiting waste. We need to monitor energy consumption better, while reducing emissions.

And then, finally, to wrap up the introduction-- Kieran and I spent some time coming up with our own definition of sustainability for this class that we'll come back to a couple of times and revisit. The way we look at sustainability is, "Use materials in the most productive way, use less, and reduce carbon footprint through the material life cycle." Kieran, back to you.

KIERAN MAK: Perfect. Thank you, Ryan. So, before we look at how to design sustainability into additive, I want to first start off by covering some of the trade-offs that we have to consider when looking at this manufacturing method. So, while there are a lot of potential benefits that we're going to look at, there are also traps and pitfalls that you can fall into. So what we're going to do in this section is identify what these traps and pitfalls are, so that you can learn how to avoid them.

As more organizations seek to utilize additive manufacturing, it's becoming a more significant and potentially impactful manufacturing process. So, significant for the benefits it can provide-- things like ultra lightweight parts for electric vehicles will reduce material consumption-- but also impactful, just some of the negative effects it can have on environments. So, things like high energy use, material waste, and health-and-safety hazards.

If you Google "3D printing" and "sustainability," there's no shortage of articles promoting additive as a sustainable alternative to conventional methods. And I think, at first glance, this somewhat feels natural. Rather than removing material from a billet of solid block of metal, for example, you create parts by addition, using only what you need, without waste. Like I said, you can create ultra-lightweight, complex parts that reduce energy consumption. And you can consolidate assemblies into single parts, reducing the need for extra material in the form of fasteners or adhesives.

And so what all the examples here on the screen are great examples of where these benefits can add value, if we look more closely at these claims there are potential pitfalls to watch out for and to avoid. So firstly we'll talk about material reduction. The claim generally goes something like I just said-- that, by additive nature of 3D printing, you only add the material that exactly you need to make a part.

And in some cases, this may be true. But for anyone who's actually printed parts, they'll be quick to point out that there are some potential flaws in this claim. What about all the failed prints?

It's not uncommon that it takes many iterations to get a successful part. And this can be caused by a lot of different factors-- things like hardware failures, material issues, but often poor design is a culprit here-- and specifically poor design for additive. You can't just take a part that was designed for injection molding or machining and throw it on a printer and expect it to be successful.

What about all the wasted material in the form of used powder or filaments or resin that can't easily be recycled? For powder-based processes, for example, the continued heating of the powder in the build degrades the powder over time. And this means that, even if this powder doesn't actually get made into a final part, it eventually will need to be disposed of.

Finally, what about all the support structures? Many processes in part designs require these sacrificial supports that add structure and strength to the part during the printing process. But these structures are sacrificial and they need to be removed and are disposed of. But they often represent a pretty sizable percentage of the final part volume.

Talking a little bit about the energy-reduction side of this, here the claim is generally twofold-- first that you can leverage complexity and optimization to create lightweight parts, and secondly that the printing process itself is actually very energy-efficient. In the case of a metal additive process, for example, there are though some things that-- it does require a lot of energy, to create these kind of complex designs.

If we look at the whole process, raw metal has to be heated up and atomized and blasted with an inert gas, to form an extremely fine metal powder. That powder then needs to be shipped from specialized-material facilities to specialized-manufacturing locations that might not be colocated. It needs to be heated up and melted with a laser, a high-powered laser, that traces the contour and infill of every single layer of the part, over several hours. And then these parts often then need to go into an oven, in order to relieve stresses that build up during the process. So all of this to say that this can be a really highly energy-intensive process, which in turn comes with an energy cost.

This chart compares the greenhouse-gas emissions for various metal manufacturing processes-- machining, casting, and additive. And it looks at it on the basis of CO2 equivalents or kilograms of CO2 equivalents emitted per kilogram of material processed. It also includes the embodied carbon upstream of the actual manufacturing process-- things like creating the powder or creating stock material. And as you can see, the energy to print parts in, say, aluminum is anywhere from 2 to 14 times higher than a casting process, for example.

And that can have some consequences. If we consider this ultra-lightweight aluminum aircraft-seat frame that was designed and manufactured by our Autodesk research team-- you know, this part is great. It's 30% lighter than a conventional seat in the same material. And on a large aircraft that has hundreds of seats, that can save over 100 kilograms of weight and 15 tons of fuel per year.

Now, if our teams had decided to print this large aluminum part, it would have emitted 14 times more CO2 than had it been manufactured with casting. And so, while we do get the benefit of the fuel savings, this adds back over 100 tons of embodied CO2 per aircraft. And this might be a worthwhile trade-off, if the part lasts, say, 30 years, the whole life of the aircraft. But it's not uncommon for changes in aircraft interiors to be refreshed every five years or seven years. And that gives us another reason why we don't want them replacing seats with something like this.

Now, as a last example of an additive-sustainability claim, we're going to look at part consolidation. By consolidating the 14-piece assembly on the left here to a single part on the right, we can reduce the number of fasteners-- which themselves add extra material as well as reduce the assembly costs. But is this kind of consolidation always the most efficient and the most sustainable way of doing this?

Often, due to the part geometry, you might be constrained on how the part can be oriented relative to the build direction. And because of this limitation, I might not be able to pack my parts as efficiently as I could. Here, I can only fit eight parts in the build. And you can see there's a lot of wasted space between all these parts. And if I don't have parts to fill that space-- say, if I'm on a deadline and have to get these parts out the door-- I'm just going to go ahead as a manufacturer and print these today, and that will lower my machine utilization as well as increase my energy consumption.

So all of this brings us to the first takeaway which is, don't assume that, by default, additive manufacturing is sustainable. Now, this isn't the most optimistic outcome, but I think it's really important to know how to identify these pitfalls as well as what they are, so that we can avoid them, going forward. Now, despite these pitfalls, there are many practical benefits that additive provides for achieving more-sustainable outcomes. I'll even start with the last two examples that we looked at.

So if you recall that aircraft-seat frame, our team actually didn't print this in metal. Instead, they took a really brilliant approach-- to print it in plastic first. That print was then used to create a positive and then coated in a ceramic, to make a mold. So this in turn allows the part to be cast in aluminum or magnesium-- really, any metal. And it leverages a lower-impact printing technology or lower energy by printing in polymer, to get a complex shape that is optimized for additive, while using an efficient casting process to get the material properties or to get this part actually made in metal.

Now, in the case of that consolidated part, what can we do about solving that packing challenge? Well, if we rethink the design and segment it and add in joining features that allow the part to be keyed back together, we can double the amount of parts that fit in the build. So, as opposed to eight parts, you can print 16 in a single build. And this approach has the advantage that, if one section is damaged and needs to be replaced, you only need to reprint one portion of it as opposed to replacing the entire assembly.

So, all this to say that seemingly subtle design changes and decisions can have a significant impact on the efficiency and the sustainability of a particular part in the process. And this is why it's important, when designing anything for sustainability, to consider not just the manufacturing process that Ryan alluded to but the end-to-end life cycle of this part.

There actually is a methodology for doing this. It's called Life Cycle Assessment, or LCA. And it accounts for environmental impacts all the way from raw-material extraction through the end life of the part.

Now, I want to make it clear that, until you do a full life-cycle assessment of a manufacturing process, you can't definitively say whether or not a particular design is more sustainable than another. For additive, one of the challenges with that is that these databases for LCA assessments are kind of incomplete. There's limited information available to conduct these kinds of studies for additive. And because it's a relatively fast-changing industry, materials either don't or can't make this information easily accessible.

And so that brings me to the second takeaway, which is to advocate for better life-cycle-assessment data for additive manufacturing processes. This data is the only way that we can really, truly know if one design is more sustainable than another. So I recommend pushing suppliers and customers you work with and OEMs that you work with to provide this information about their machines, materials, and processes, so you can make informed decisions.

However, just because it's hard to precisely quantify these impacts doesn't mean that we can't do anything. There are practical steps we can take today, to achieve more sustainable outcomes. And it starts with going back to what the basics of additive is really good at.

So, after all, there needs to be some reason why you would print this part rather than make it be another method. So these are things that are well-established-- improve performance, mass customization, flexible manufacturing. But let's not just limit ourselves to an additive lens. And let's figure out what needs to be done, when we refer back to our UN sustainable-development indicators-- so, things like reducing CO2 emissions per unit of value added, reducing the material footprint, and increasing recycling rates. It's the overlap between these two areas is the space that we can operate within, to find practical and sustainable solutions.

Now, if we go back to our sustainability definition from earlier, we can break this down into a few key areas. One, using materials in the most productive way. Two, using less. And three, reducing our carbon footprint throughout the material life cycle.

Now, "using materials in the most productive way" is really just a way of saying "improving product performance," which we've established additive can be really good at. Using less means reducing waste throughout the manufacturing life cycle. And as we saw, there's definitely some room to improve there. But we're going to look at that. And then third, reducing the carbon footprint during the material life cycle means shifting towards more bio-based and recyclable materials. So, with these in mind, we'll walk through five examples of practical solutions that relate to these categories.

So when it comes to improving product performance, focus improvement efforts on enabling and enhancing clean technologies. Here, think about leveraging additive to produce complex, optimized parts that don't just deliver improved performance but also do so for technologies like heat exchangers for carbon capture or hydrogen reactors.

Now, remember all those stale builds and wasted powder that we looked at in the first section? Well, in order to reduce those and eliminate that waste, we can leverage simulation. So, by running process simulation of metal-powder-bed printing, it's possible to actually predict part failures before they happen. And because AM is a very digitized process, we can make changes to the part geometry, the support design, the process parameters, to correct these and prevent them before they happen.

And while we did disprove a little bit the claim that additive only adds exactly the material needed to make a part, it is really good at selectively adding material, with certain techniques. Direct energy deposition is a great example of that-- or DED. And it's a process that actually also has a very comparable CO2 emission to a lot of the other conventional methods. It's very fast, and it can selectively add material to areas of parts like turbine blades or propellers or molds that need to be repaired. So think about remanufacturing. This extends the service life of these parts, as well as amortizes the amount of time that the embodied energy that went into making them gets amortized over.

Digital warehousing is another great example. It can enable parts to be produced on demand. And oftentimes, companies have whole teams of business planners that need to be able to predict production quantities that a company needs to produce in a given time. And the reality is that, despite the best predictions, you'll almost always miss it a little bit. You'll either underproduce, which means that you're losing out on sales, or you'll overproduce, creating excess product waste.

Finally, material manufacturers are increasingly offering materials that are bio-based or fully recyclable, which hasn't really been the case in the past. And if you can explore switching from a material that's not renewable or nonrecyclable to one that is, it helps extend that material life cycle and promote circularity.

So what I want to leave with on this topic are these five examples that can be used as a jumping-off point for identifying practical solutions for designing sustainability into additive. These can help internal advocates at companies like yours start the conversation and filter through applications that provide real benefits while avoiding the pitfalls that we saw in the previous section. This brings us to that third takeaway, which is, building a business case that aligns economic and environmental benefits to the unique strengths of AM.

In the next sections, we're going to look at three workflows that are based on these examples. And we'll start here by talking a little bit about shifting to bio-based and recycled materials. And so with that, I'll hand it back over to Ryan.

RYAN ABEL: Thanks, Kieran. To tackle minimizing the amount of material we're using-- so we're going to reduce the amount of material use and look at our bio-based materials-- to do this, we're going to tackle the workflow that you'd use inside Fusion 360 to set up, review, select, validate a design. In this case, it's going to be multiple designs.

The design-and-print workflow is divided between generative design and then using Fusion for the additive build space. For this demonstration, we're going to focus on the generative-design portion, setting up our goals and constraints, solving on the cloud and interacting with it real-time, exploring the results, and onward.

So the question is, can we use a different material? Or can we use less material? And using a different material is a common question. And a lot of times, the question's asked when you're looking at the same design-- meaning that, can we use this design and just change the material? Very often, that's just a very constraining question. And with automation tools like generative design, you don't have to think of changing the design as a heavy lift for engineering. It's becoming very easy.

And in this case, what I'm going to show you from the recording is a process that took me two and a half hours, to go from nothing to over 180 different results, where we picked three that we could then build. So this is very much a process of creating three designs, in the end, that would traditionally have taken weeks, if you could even build them or even conceptualize them. And then, now we're back down to a process under three hours, to get them on a printer and getting ready to prototype the next day.

People that have seen generative design before might be thinking, in their mind, here come the bony structures-- the minimum material conditions. That's not where we're going. I'll show you what the final designs look like. And they're color-coded by the different materials we're going to explore.

And one thing you'll notice is that they are all unique. They do have similar features. And all of them reduce the mass of the original design. So again, if you're to look at, can we change material, I would say, "can we change material and design" is something that needs to be paired together to get the most out of the process.

So the materials we're going to look at are two of the PA 11s, which are the castor-bean or bio-based materials that can be compared to our PA 12, which is our petroleum-based material. I put the PA 12 on there just because it's one of those materials that's been engineered specifically for additive. They control the particle-size variation, like you see on the left, a lot more, and just the shape distribution. This helps with print finish, color durability, shape-contour resolution. And it's being used a lot as bridge production, to ship parts that would traditionally be injection-molded.

Again, I put it on here because material properties are interesting. So it helps you understand-- if you grab a material that's oil-based and it has a higher performance, can you get enough material out of it to make it, financially, and maybe even economically or sustainable decision.

So when we're looking at the material properties, the three properties that I'm interested in, when doing a generative-design study, is primarily the Young's modulus and ultimate strength. And you'll notice that, as we go from left to right here, we're generally getting lighter materials, more elastic, and stronger. And it's that elasticity and strength that is making the design challenge both interesting and something that requires the design to change. And I hope that comes through inside this presentation.

So we're going to start at the very beginning. Like I said, this process took about two and a half hours, and it's been streamlined down to a couple of videos here that we'll cover in under five minutes. It does move through pretty quick, but I hope you appreciate the workflow.

So we start blank with an existing design. And we're going to kick this off and just look at the design, review the feature history at the bottom, see how this was built in a traditional process-- not with additive in mind. Now, we're going to generative and set the problem up-- which is completely different Now, we're going to look at defining geometry as red for "material cannot go here," so we need to be able to get some type of wrench in here, to tighten it. And we'll have green shapes-- in this case, washer shapes-- where you're going to put your fixed constraint for mounting and loads for the pivot hinge.

Now, the loading on the pivot hinge, this can move across a wide range of angles. So we have to represent that as specific load cases. So we try to determine the worst-case positioning or sideways load or twist, and we use those to capture what loads it needs to withstand. Which means it's going to be the loads that are used to optimize and design the part.

So in this case, we have three load cases. And then we go into the manufacturing method for printing material. And going to select all the different additive materials that we want and then run a quick preview locally before sending it to the cloud, just to see what the general shape looks like, and start to see how it's going to get smaller and smaller. Generative design is really good at both minimizing material but also moving material to where it's needed.

So now we've sent the run to the cloud. And now while it's solving in the cloud, you have full interaction with the results. You can start to view them real-time and compare them side by side here.

This is an iterative process. For every model, it's going to take 20 to 30 iterations to get to the final one. But that doesn't mean that some of the earlier ones don't provide value. So you can download them and evaluate them whenever you like.

This is one of my favorite ways to look at it. Looking at our displacement, here, versus mass, we can pick some of the highest-performing ones and, again, go right from the graph to looking at the geometry, to really understand the value.

Now, what's important to realize is, performance is being measured by how stiff this is-- how much the ears move, when you twist and torque this thing. And the result is telling us that we can get the mass down really, really low, but our displacement's going to be over 2 millimeters Now, I know that the performance of the original design was closer to 1 millimeter, so this might adjust to performance. So we might want to build it and test it and find out. But even better, we can go back into our setup. And while this thing continues to solve, we're going to expand our design study.

So let's copy that one. We're just going to make one small change. We're going to limit the displacement to 1, so we can compare apples to apples as far as stiffness is concerned. So the we can send this model off to the cloud to solve, as well. And it gets added to the same Explore environment as the first set of results.

So now we have 135 runs, running in the cloud all at the same time. You can see the variety of different shapes being created. But again, going to this x-y plot, we can pick the high-performing ones very quickly. In this case, there's four that I'm interested in. So let's download those and get back to work inside of Fusion, to understand its performance and inside the assembly better.

There we go. So the other thing that we're going to do inside here is, as part of this exercise I noticed that the parts could be more symmetric. So there is another capability in here, to force symmetry in the solve. So we do have a plane of symmetry here that we want to activate. So I'm going to turn that on, as well, and send those to the cloud.

Now, what you noticed, here, I progressively went through setting up some studies looking at the results and adding more definition to either the physics or the symmetry aesthetics, here, because this is how the tool gets used. It's a very interactive process. So now we have a symmetric design that we like. We can open it up and run the full FEA. But it also gives you the history of how this part was created-- the obstacles that were used, the preserves that were used in the end, this-- or the shape.

Now, this shape is very organic. And it's still editable, using the underlining sub-D surface definition. This is a very common workflow for designers but relatively new to engineers. So we've embedded some automation tools that will create things that look like cylinders. Let's make them real cylinders. Let's smooth things out and control it.

So there's a lot of automation tools in there. Here's a cutout that I don't want. You can just delete it. This was cut out because of the preserve.

But now our simulation results are done. So this is traditional FEA on the final result. This was part of the generative-design process, where you can identify your stress concentrations and ultimately your performance metric of what the displacement is under load.

So that really wraps up the end-to-end workflow. Like I said, this is meant to show you what can be done in two and a half hours, leveraging over 150 runs in the cloud and three different problem definitions.

So, just to recap the results-- our original design was made at a PA 11. And our displacements for our two critical dimensions were between around 1.7 and 1.8. If we were to keep the material the same and just change the design, we can save 14% material. And this might be good enough for a project like this.

But the other thing to observe is that the generative process likes to stiffen things up, as part of its core skill here. So in the end, it used 14% less material and is stiffer. It deflects less. This might even perform better. Again, physical testing is the only way to really determine that.

Now, looking at the high-performing PA 11, you can get rid of 49% of the material, but now our deflections start to increase again. This might be acceptable, but we're deviating from what we've been shipping or from what we've tested in the past, so we might want to look into it. Finally, once we set the displacement limit to match performance, our material savings was 28%.

So what this tells you is not what the final answer is. Because we still have to take this and do some of the economics on what the cost of these materials are and get into more of the print time of these, as well. But right now, after two and a half hours, I have five parts I can print and start to test on and really start to make some informed decisions. And that's the whole purpose of what upfront simulation is about and what design automation with generative design-- how it's changing the perception that it takes weeks to build and test. Today, we can do it in really, sometimes a matter of hours.

KIERAN MAK: All right. Thank you, Ryan. So, shifting gears here a little bit, we're going to look at specific ways that additive manufacturing can be leveraged to improve the performance of clean technologies. So one of those examples or those categories that we talked about. In this particular example, I want to look at the cooling of an EV battery cell.

So data showing that exposure to heat and fast charging actually can promote the degradation of lithium-ion batteries more than actual use. So the battery in a vehicle that does not have active cooling will degrade twice as fast as one that does. And while battery chemistry has an important role to play here, cooling systems like a heat sink-- shown here on the right-- are really essential to improving the performance and extending the life of these batteries. EV batteries actually have a relatively narrow band where they perform optimally, which is why thermal management is so critical to enabling this transition to electric vehicles.

Now, in this particular example, we're going to take a simplified look at how to design a lattice heat sink on a battery module to improve the cooling performance with additive manufacture. Lattices are great for this type of application, because they can provide really high surface area to volume ratios while also being easy to print, self-supporting, et cetera.

So now I'm going to walk through what this process looks like step by step. So we'll start off in Fusion. And we can just import that baseline heat-sink design and use that as our reference, offsetting the profiles, trimming away the fins, so that we keep a comparable design to the original.

Next we'll extrude out a solid body that is separate from the original design. And this is important here, because this separate body is going to represent the design space I want to populate with a lattice, as opposed to the solid volume that I will combine it back with later. Now, the reason I'm doing only midplane here, I'll explain a little bit later. But what we're going to do now is move over to the Volumetric Latticing feature in Fusion 360. And this is a part of the product-design extension.

So what this feature does is it leverages our new volumetric kernel. So this is geometry. It's not a solid body. It's not a mesh. It's a volumetric representation.

And that has a number of advantages, but the first one you'll probably be noticing on the screen here is that it's really fast, even in this sped-up video. It's so fast that I can do things that wouldn't be possible with other modeling kernels-- quickly changing my unit cell, scaling the design up or down, or translating it and rotating it. Regardless of how complex it is, those are things that anyone who just tried to design these kinds of structures-- you can't do with a mesh.

I can scale the structure uniformly or nonuniformly. I can adjust the density, so I can predict what kind of [? my volume ?] [? from action ?] will be. And I can adjust that density along a particular edge, so I can control the density gradient, as well as even do this along a sketch.

So say, for example, I want to increase or reduce density along a particular flow path-- make it sparser or less sparse, throughout a particular path. I can also do things like create solid regions, using the Offset command-- so imagine I need to print this part with a flat face or have it made up against another part-- as well as make adjustments to the structure numerically, to adjust printability or control how it intersects with other regions of this heat sink.

Now, I'll go through this process, but really what I'm trying to do is, once I've tuned my design, gotten the volume fraction to where I want it, I can predict with that, knowing what the mass will be compared to the original design-- I'm going to want to do two things, after that. So I'm going to first want to run a simulation, to see how it performed, and as well as manufacture it. So, in order to do that, we can use this volumetric representation to design, but we will need to convert it to a mesh-- and eventually convert it to a solid body.

So here I am going ahead and converting it to a mesh. And this is all done within Fusion, so being able to work between those different modeling paradigms makes it easy to design for additive and take advantage of both mesh and parametric design. And here I'm going to remesh it, just to clean it up and make it a little bit more lightweight and easy to work with. And what I'm going to do next is--

What we just saw was nothing new-- converting a volumetric to a mesh-- not a big deal. But going from a mesh to a solid body, that's a really challenging thing. And with the Organic Mesh Conversion tool that's also part of the product extension in Fusion, this makes it really easy. This is a process that might take someone-- they might have to completely rebuild apart from scratch.

But using this tool, I can convert my mesh into a solid model that I can then run a simulation on. Now, depending on your complex geometry, it might take a little bit of while to run this. But this is why I split it into two halves.

But all I have to do now is mirror this body, merge it back to the rest of my heat sink, and I have a final design that I'm ready to review and eval the simulation or run a simulation on and print. Now, the whole point of doing this was to answer the question of whether or not this particular design performs better than the other one that we looked at-- so, the previous heat sink. And in order to figure that out, I'll move over to the simulation workspace in Fusion. So I'm going, in the same document, from my design to my simulation-- and set up an electronic [INAUDIBLE] study.

In this case, I've already modeled a few extra components that represent an air source and a duct to control the flow. Here, I'm simplifying. And I can modify, using the geometry tools here, to remove elements or combine parts that are not relevant or don't necessarily need to be separate for the simulation. And just like how Ryan set up a gen-design study with targets and constraints and loads, I can do the same thing with a thermal study here.

So I can apply a thermal load to these components, assign a critical temperature that my battery can't exceed-- in this case, we're going to set that to 35 degrees-- as well as apply or assign specific geometry to be my heat sink, apply a flow rate for the air source-- and when I'm ready to do this, I can select the ambient temperature and finally push that geometry and that study to the cloud. Just like we did before, you can also clone and create studies so that you can iterate and create essentially a DOE, so that you run all of these in parallel and don't have to wait for them to solve one by one. When that is done solving, when we get it back from the cloud, I can review the results.

So, looking at this original heat-sink design first, we can see that, for these particular conditions, it did exceed that optimal temperature of 35 Celsius. And I can visualize how the temperature is distributed through the heat sink as well as how the air flows through the system.

Looking at the gyro now, we can see that it does achieve a lower temperature. So it's below that 35-degree threshold. And I can see that it's taking more heat out and air is flowing throughout those lattice gyroid structures in a way that's transferring a lot of heat out of the system.