Beyond Bird Bones: Exploring Generative Design in Traditional Manufacturing

IGNACIO MEDINA: Hello, and welcome to today's session, Beyond Bird Bones, Exploring Genetic Design in Traditional Manufacturing. My name is Ignacio Medina, and with me we will be Josh Reader. And we are going to be both your presenters for today.

Before we start, safe harbor statement. We may do forward-looking statements. No purchase or investment decisions should be based upon the features. They should be only made on the features that are already existent in Fusion. Also, AU content is proprietary. Do not copy, distribute without express permission.

So with that said, let's speak about today's agenda. First, you're going to meet your presenters, then we're going to speak about the generative design fundamentals. After that, we're going to analyze our case study.

George is going to speak about the design for manufacturing conditions and what's important there. And then, we're going to analyze the outcomes do our machining setup for those. And then, finally, summarize everything we learned today. So about your presenters, I'll let Josh introduce himself. Josh?

JOSH READER: Hello, everyone. My name is Josh Reader. I'm a manufacturing specialist from the Bowman Technology Center in the UK. Yes, we've flown over from the UK to America to do this presentation.

So my background is mechanical and manufacturing engineering with a speciality in manufacturing. I test and validate the Fusion 360 software manufacturing workspace as a daily activity. And I'll hand over to Ignacio to talk about himself.

IGNACIO MEDINA: Perfect. Thank you very much. So my name is Ignacio Medina. I'm from Spain. And I'm a technical specialist here at the Fusion 360 hub in Europe, which is based in the beautiful Barcelona.

And my background is as a mechanical and aerospace engineer. I had the pleasure to work before in the industry. Between those areas I worked, I worked for the biggest aircraft manufacturer in the world where I had the chance to design and develop products that are being used using Autodesk technology, and my speciality is generative design.

So speaking about generative design, I could be actually talking hours about this technology. There's a lot of exciting things to be said here. But I'd like to highlight that it's a thoughtful design. And why I'm saying thoughtful design, because of multiple reasons. First, it is allowing us to improve our product performance. How do we improve product performance?

By having a better understanding of our parts, and our products, and designs that we are making. We also have the necessary tools to optimize those and increase their efficiency and change those. And this all without having to do investments in complex setups, and complex testing rigs, or prototypes that at the end of the day will add a lot of cost. We also increasing the product innovation at the end of the day.

Because we are going-- and today this is the case also-- we are going to simultaneously generate multiple outcomes without having to over-complicate the process and use-- creating so many different outcomes that I wouldn't be able to come up with myself without these tools. And we also reducing the product costs, because we are boosting the design process. But not only that, we are also exploring various material and manufacturing alternatives. That means even if at the given moment I don't have access to a specific machine or material, I will still be able to generate an outcome that fulfills those requirements.

And seeing all those points, there is a lot of points here already speaking for the generative design. So why we are not using it more? And the reason is a lot of people look at this, and they see this bird-bone structures. However, the reality doesn't look like this. They have a reality that's closer to this other one, where they have a traditional manufacturing setup, where they're using maybe 2-axis, 2.5-axis, up to 3-axis milling.

And there's this conception that if you're using this manufacturing method, then generative design is not for you. You're not going to be able to benefit of those improvements. It's like two parts of a puzzle that don't fit together and just sticking to it. And we want to demystify this today.

That's exactly the objective of today's session. We want to use Fusion 360 generative design of Fusion 360 in order to create ready-to-manufacture results. So if I got your attention, let's go into it and see our study.

So this is a very exciting case study. We used a sports car pedal, and especially this part in green, which is highlighted there. It's obviously a part with a lot of requirements. And we are going to assume that we know the starting geometry. We have an understanding of it, the loads, which are going to act on it.

And we are going to aim for the maximum weight reduction and also the safety. We are ensuring a safety on the part. So for that, we are going to look at the safety factor, but we are also looking at the displacements on local areas.

And since we also want to explore the weight reduction on the alternatives, we are going to see if we should manufacture it in steel or we can go for a more affordable option, which is aluminum. And then, we are going to prioritize the cost efficiency. That means we are going to use in-house production technologies. In our case, 2-axis and 2.5-axis milling. And then, we are going for a serial production of 500 parts.

So before we go into our study and analyze a bit more what we did, I want to highlight it. We are going to analyze it from a manufacturing point of view. So we are going to speak about the conditions from that we need to take into account when manufacturing this part or when generating this part, and then we are going to analyze from manufacturing point of view. However, if you're looking into or you're interested into understanding how to do a setup infusion in generative design, then I recommend you to go to this link, where you're going to have generative tips and tricks. And you can learn and understand how to do it-- set up simple setup infusion.

So staying to our topic, design for manufacturing, Josh is going to speak what considerations are important when designing a part for manufacturing. Josh.

JOSH READER: Yeah, so to start off with, we're just going to talk about general things we need to think about even if we're not using generative design. So what machines do we have available? What is our CAM subscription capable of? And what is the cost of manufacture going to be? And these three things are vital, especially when choosing your method of manufacture in your generative design.

As we can see in the top, the top part of the small box that we've got there are configuration. And this is where we choose whether it's going to be a 2-axis, 2.5-axis, 3-axis, or 5-axis outcome, as well as additive options. But today, we're going to be focusing on 2.5 axis. And that's because that's more of a traditional part to be made. And we want to come away from the idea of this bird bone structure.

So what machines do we have available? We're looking at 2.5 axis only. So these are our basic routers or milling machines. What is our CAM subscription capable of? Can we actually program a 2.5-axis part? Is it just a laser cutter or is it something like that?

And just making sure that we don't need 5-axis. So our basic effusion can machine a part into a 2.5 axis. And then cost of manufacture, what else are we going to be looking at? The size of the material, does it fit in our machine? And the material itself, that changes a lot of things. That changes how much the parts going to cost, how long it's going to take to make.

Tolerances. Tolerance is also massive. If you set really tight tolerances on a component, it's going to cost a lot more to make. And that's because machinists and programmers have to put a lot more effort in getting to that certain size.

Wall thicknesses as well. So our minimum wall thickness, we can input. And that just makes sure that when we're machining the component, we don't have thin walls that can vibrate, which can cause poor surface finishes and geometrical issues.

And then, we've got tooling required. The smaller diameter we have, the smaller tool we're going to require. And we're going to have to step down in those tool sizes when we're manufacturing the component, because we can't rough out a full component with a very small tool. So when we look at this, we can look at the minimum tool diameter that we set, and this is going to mean that we can choose tools that we have in our workshop. If my tools only go down to 10 millimeters, we need to make sure that we're using a minimum tool diameter of 10 millimeters or bigger.

And then, we've got to think about the minimum wall thicknesses. As again, we don't want it too thin, because then we're going to have vibration during the milling process. And then tool direction. So tool direction, this decides where the material is going to be removed from.

So we want to make sure that we're picking a correct section of the part for the material to be removed from for the generative design to do its thing. And this is important, because we need to think about the setup of the components. So how are we going to hold it in the first instance? And where do we want those features to be added? And over to you, Ignacio.

IGNACIO MEDINA: Thank you very much, Josh. So now that we've understood the conditions we need to take into account when creating the design, the next important step will be to analyze the outcomes that we were able to create based on that setup and we established. So we got a lot of different outcomes. And it might be overwhelming to understand how to classify this and analyze this.

We're going to follow three easy steps. We are first going to study outcomes groups, classify them in different groups and study those, then we're going to select the most suited from a manufacturing point of view and see what fits our requirements or what Josh is looking for when manufacturing, and then we are going also to give you some tips for the right model selection. So in case you are starting off with this, you have some very interesting tools inside Fusion that help you to select the right model.

So looking at our outcomes, there's a few interesting outcomes. The first one, if we go over down to the right, we see it's our famous bird bone structure, the one I've been speaking all the time and the most manufacturers look at it and think to yourselves, what I'm supposed to do with this? We're going back to this model, but let's analyze the other ones that generative design gave us. So the generative design gave us two multiple 2.5-axis milling parts, and then we also have 2-axis cutting parts.

Going into the 2.5-axis milling parts, we decided to do two different setups, as Josh was mentioning, to see how the tool diameter is going to impact the outcomes. So we see here, we have an outcome with an 8-millimeter tool diameter. While the other one below that, it's an 80-millimeter tool diameter. And we can see from designer point of view, I can really see one is going to offer me more pockets, because obviously, it's using the small radius of the tool to make pockets. It's going to have a slightly bigger weight reduction.

However, the 80-millimeter one is also going to make thinner elements. It is going to be a slimmer design. And it's removing the pockets in this case. And it's just doing the contour on the part. That's on the first view.

And Josh is going to speak about the manufacturing on this. And we're going to have a deeper look on what challenges we might have in this. We also see the 2-axis cutting tool part. Nothing much to add to this. It is a simple cutting part.

It's giving us a very good contour on what we want to achieve. And then, coming back to a famous bird bone structure, it is still a very interesting design. Even if we are not going to directly be able to use it, it is still giving us a very good understanding of what's going on in the part. We see the load parts. It is being represented on that structure.

And we can use this very good guide in order to maybe design all part using those elements as a reference, but removing those that are less critical. We see, for instance, beams that are maybe not that thick and then important. And we can assume that they are less critical for the design. And by this, we are able to create a design, which is manufacturable with the traditional manufacturing method without having to rely completely on maybe the outcomes that we have created with the automated generation of 2.5 or 2-axis milling.

So now that we analyzed it from a designer point of view, Josh is going to give us a more deep analysis on the manufacturing side. And, Josh, if you want to start, we see we have four different outcomes very similar between them. And I see the two on the bottom, they are very, very similar. Is there any issues on those? Can you explain us a bit?

JOSH READER: Yeah, so as you can see on the bottom two, they're very similar compared-- very similar, but different to the top two. I would categorize these as, right now as a machinist, I wouldn't want to be making that. We've got some sharp corners. We've got some upstands.

And those can really cause issues with the tooling, so it could break the tool. It could damage the tooling. And they're just not ideal.

But the good thing about Fusion 360 is that small upstand can be edited. We know they're structurally not causing any issues. So we can just remove that simply just going into the CAD workspace getting rid of it, and then we've got a nice open pocket to machine. But compared to the top side, they look much nicer. They look more structurally sound and something that would be easier to machine.

IGNACIO MEDINA: OK, and speaking about the top two designs, I see one has a more normal pocket while the other one is a more open pocket. For manufacturing standard point, for milling point of view, what will be the considerations or what way will this affect the milling of those parts?

JOSH READER: Yeah, so the one on the left-hand side looks more structurally sound to me. But the issue is it's got a deeper pocket. So because it's a closed pocket, we'll have to use smaller tooling to ramp into that pocket, which can take a little bit longer.

So it's going to take longer ramping into the pocket. We're going to have to finish those sidewalls. We're going to have to finish those bottom walls.

But the one on the right-hand side, it's got a bit of an open pocket than another smaller pocket. The open pocket is really useful, because we can come from the outside of the part onto it, which reduces the amount of time we have to spend ramping onto the component. And just a quick example of-- ramping is the way we move the tool onto the surface. So ramping onto the surface takes a little bit longer than just coming down to the side and coming onto the component.

IGNACIO MEDINA: OK, and what other things from a manufacturing engineer standard point of view will you also consider when selecting a generative design outcome of all those we have here?

JOSH READER: Yeah, so we can't see some of the issues here, but there's things like thin walls, but we can control that in generative design. So if there is thinner walls, we can choose whether we want to avoid those due to the vibration. Because it's allowed to almost do what it wants in terms of creating geometry, we want to make sure that we can hold on those geometries. That affects the workholding standpoint. So if there's better areas to hold on, which means we're going to have a more rigid setup, so we're going to have a better surface finish.

And we're going to be able to achieve those dimensional tolerances easier. Yeah, as long as there's no undercuts, we're going to have to require less tooling. So there's a few little things that we can think about when choosing our designs.

IGNACIO MEDINA: OK, great. And moving to the outcomes we select to the right, the ones we picked, how do they differ from a manufacturing standpoint? One looks, for me, easier to manufacture than the other one. I'm right or can you tell us a bit more about that?

JOSH READER: Yeah, so obviously, with the 8-millimeter outcome, we're going to have to actually step down in tools. So we're going to have actually more operations. We're going to have to do something called rest billing, which I'll show you later on. But it's going to take longer to machine that 8-millimeter outcome than it is going to be the 18, because that means you don't-- because with the 18-millimeter, we can go with a bigger tool. We don't have to do as many tool changes.

It's a pretty basic shape. So it's going to be easier to machine in a general aspect. So the cycle time, I would expect the right-hand side to be much, much quicker than the left-hand side, because we don't have to ramp into any pockets or change tooling size.

IGNACIO MEDINA: OK, great. That gives us a very good idea of what things we have to look at. But as I was mentioned before, some people are looking at this video or joining us in this session, they might be not as experienced on the manufacturing side and they might be having some challenges trying to understand and see those parts. However, Fusion and generative design tool is giving us the necessary tools to understand which outcomes might be better.

So this is exactly what I want to show you here. You can select a filter out the recommended outcomes based on your own criteria. That means you can select which things are most important for you. For instance, for us, low manufacturing complexity was our highest priority, because we wanted to show simple outcomes. And we can filter those out.

And based on that, Fusion is not only going to give us suggested outcomes, it is also going to list all the outcomes, telling us which ones are having better chances or are meeting more of our conditions and which are less ideal for criteria or base we selected. If that's not enough selection criteria, you have also a chance to filter by mass, price, and stresses. You have various ways of filtering those parts out. You can obviously also select the outcomes and compare both of them between. So like we were doing there in the previous screen, we were selecting two different outcomes and analyzing both in comparison.

And finally, you can also have a scatter plot view, where you can analyze two components based on-- multiple components based on multiple features, so price compared to the loads or to the mass we have on the part. So it is giving us the instruments to understand those parts. And not only in relying on what your charge is telling you, you have also automated ways of doing that. So now that we told you about the outcomes and we analyzed those, I think the next step will be to machine those. And I think Josh has a very nice setup now.

JOSH READER: Right, so the first thing I'm going to look at today is what kind of programs are we going to be using, how easy it is to make these parts in Fusion 360, and how Fusion helps us create multiple of these parts. So first thing we're going to look at is how we're going to set up the component. So the first option is in a vise using a basic block.

And that's kind of the method I'm going to be using today. But we could do other things, like this part can be casted or it could be 3D printed before we actually mill it out. But for a basic guide, I'm using a single billet of material. And I'm going to walk you through how we program the component.

So I'm going to go through a normal milling method, which is roughing, rest roughing, finishing. And they're just the ways we go around the component. So if we look at setup 1, I've chosen this side, because this side gives me more area when I come to my second side of the component to hold on. So we can see that there's more material here to hold on when I'm using a [INAUDIBLE].

So first program is going to be an adaptive toolpath. So an adaptive toolpath utilizes the flute length of an end mill with a smaller step over to have a higher material removal rate. And what you'll see with a lot of these toolpaths, their model-aware toolpath. So what I can do is I can copy this straight over to another model using templates. And that will create the exact same toolpath, but follow the same model.

So we've got our rest roughing toolpaths to start with. I'm then going to finish the profile, so finish the outer profile of the component using a silhouette style. So what that does is it picks up the silhouette of the component. It's going to finish the whole outside of the component. And then, this is what I talk about when I talk about rest milling.

So rest milling is where we use a smaller tool to get into the areas, the original roughing tool couldn't get into. So we see this area here are 16 millimeter tool couldn't get into there without just plunging straight in, which we don't want. So we've gone with a smaller tool, in this case, an 8-millimeter tool, to rough out that counterbore that we see there.

And these red lines here are exactly what I call the ramping moves earlier on. And these are slower and just getting to the depth instead of plunging in, because that's not what we want. And then, we've got to think about how are we going to [INAUDIBLE] the next side. So the [INAUDIBLE] is our work offset. So it's this little coordinate system here, so our machine knows exactly where we're coming from.

And if you think about it on this hand side-- this side, we're going to just have a piece of stock. We want to make sure that we've got a precision area to come off as a [INAUDIBLE] on a second side. So I've just put a bored hole straight through the middle, so that we're able to access it on that side.

Then, I'm using a flat finishing strategy. So flat finishing, it automatically identifies all the flats, the flat bottoms, and flat walls to ensure that we can machine those to the right size. So this is something which we will call a finishing pass.

Then, I'm going to use hole recognition. So hole recognition automatically finds all the holes on your component. And you can then decide what kind of strategy you want to use.

In this case, I wanted to make sure that it's spot drilled and drilled the holes. So all I did was click the hole recognition, it identifies the holes. I can choose how I'm going to identify those holes, and then it creates these toolpaths for me.

And then finally, deburring. Deburring just takes a little edge off that side of the component to make sure that we've got no sharp edges. And this is, again, it's automatic.

So first of all, that's a lot of programming. So there's one off, which is going to be difficult. But what we can do now is I can create this as a template. So I'm then going to go into store as template. So I'm going to call this 2.5-axis generative design guide 1.

And then, what I can do after that is-- what that's done is it's taken all of those toolpaths. And then, I can go over to a different design, so we can see that this is originally-- this is the 8-millimeter component. I'm going to come over to my 18-millimeter component. And I've already created the setup just to make this a little bit easier, is I can then come in here and create from a template that I've made. So we can see that we've just come from one side and just put all of the tool paths into this component.

So you can see again, correct, that's what I want. We're going to be finishing the profile. We're then going to be rest milling this area again. But this part has an error on it.

And that's because this is a 2D toolpath. All I have to do, though, is just reselect the hole that I want to machine, and I'm ready to go. So we can see that all these toolpaths are validated now. So we can see it's all good.

And then, I'm just going to quickly show you how hole recognition works. So we can go on to hole recognition. So this is where it identified the three holes that I want to machine. I want to ignore this one. I want to spot drill and drill this one.

And in this case, what I want to do is I want to split the hole signature. And that's because I've already machined that counterbore and I don't want to machine those holes. So we're then now looking at this bit.

And I want to spot drill and drill the hole. So all I've done there is just a couple of clicks, and there we go. We've got four toolpaths ready to go.

Now, if we go back to our other toolpath, I'm going to do the same on this side. So we can see that we've already got all of our toolpaths ready, we've machine them, and we've finished them. So all I can do here is create another template. So I'm going to store this as template again.

I'm going to call this 2.5 axis [INAUDIBLE] design 2. Then I can go again and paste it into the second setup again. So I've done that. We can then create from template again.

So all we have to do here is make sure that we can select the bottom of the component or a certain area. So I'm going to select this area, but I don't want it to go further down. So all it does is just reduces the number of clicks that we have to do when creating the toolpath. And then, we go generate that again. So we now have another complete component.

Here we go. So that's how easy it is to take-- you can fully program one. And let's say we have six different outcomes and we want to machine all of them, we generally only need to get one done, and then we can use that template on all of those parts and save some time, save clicking, and generally save time on programming. So with that, I'm going to hand back to Ignacio.

IGNACIO MEDINA: Thank you very much, Josh. Now that Josh showed us how to do the machining setup and we learned how easy it is to use templates and reuse those, let's go to a summary and see what we learned on today's session. So our main objective was to learn how to use generative design via traditional manufacturing. I think we achieved that goal with excellence.

And we were able to recognize by that to identify the critical design conditions, what is important when creating a setup and conditions on generative design. A part of that, we were also able to discover how those parameters might impact-- those constraints might impact our generative design setup and the outcomes we will encounter. We saw it with the 8-millimeter and 80-millimeter setup. And we were also able to use technology, which is a generative design, which will help us to drive innovation to improve our existing design in being able to use our traditional manufacturing systems without having to do an investment on more advanced technologies.

So next time somebody comes to you and tells you that generative design is something only for space applications, you can prove them wrong and show them how to do it. So now that we learned all of those things today, it's time to call for action. So we have three favors to ask you. The first one, give us an honest review on this class page. That means a lot for us and help us to understand how valuable this kind of classes are.

Then second, go to Fusion and try to use generative design in one of the products. It doesn't matter which product. Use the first one that comes out of your mind. Try it out, and you will see how that helps you to improve the designs and come up with a lot of interesting ideas.

And finally, write a comment tagging us on LinkedIn with your latest generative design. We are excited to see what people are coming up with. And we want to really see where you're taking the limits of generative design.

So thank you very much for joining us in today's session. It was a pleasure. And I hope to see you soon.