Seurat Technologies was founded off the back of a problem CEO James DeMuth [JDM] ran into while at Lawrence Livermore National Laboratory’s National Ignition Facility (NIF).
In the NIF, there are 192 of the world’s most energetic lasers collectively delivering over two million joules of energy to the same single target. The aim of the research project that was doing this was to create the same nuclear fusion reaction that powers the sun. As such, the stresses on the walls of this space are significant, with ever laser shot said to cause a multi-hundred-degree fluctuation in temperature that causes cracks to form in the walls.
While a specialised steel that could withstand such conditions was found, it was unweldable. It could, in theory, be 3D printed. But to build the two-metre-thick walls that would comprise the 12-metre chambers would probably take a couple of lifetimes.
It has since been DeMuth’s mission to hurdle that manufacturing obstacle, as well as a few more on the way. Already, Seurat has raised $79m, secured multiple letters of intent from clients and has got the ball rolling on its first application projects. Here, we speak to DeMuth to find out more.
JDM: The original intention was ‘let’s print big objects really fast,’ whether it’s a fusion chamber, whether it’s a car, and we initially thought, ‘hey, let’s print an entire layer of material at once,’ and that seemed like a great idea. It’s not. So, we evolved, and we got to a more integrated, optimised systems model of what makes sense to best use the capital equipment of the lasers and crank that knob to 11, make it so that lasers have a fantastic scaling curve. They follow Moore’s Law scaling in terms of cost, but you have to get to volumes that require it, and lasers are super expensive. So how do you get to volume with something that’s really expensive? There needs to be a big enough problem for it to solve to get it there.
The motivation was, really, on solving the problem of energy. I listened to a talk when I was younger about ‘here’s ten of the world’s problems: hunger, drought, poverty…’ But all of the problems can be solved if you solve the energy problem, because energy is literally, by definition, the force of change. So, if you can get cheap energy, you can solve all your other problems and it turned out in order to get cheap energy, we first had to have a new way of manufacturing things. Hence, I’m in manufacturing.
JDM: We take a really high-powered laser, we generate a pulse of energy – so very different from what’s done today, as far as either a continuous beam or a really high frequency rate of pulses, like in tens of kilohertz – and we do this 40 times a second, so at 40 hertz, so relatively slow from that perspective. And then we send it through a whole bunch of optics to format it, we’ve literally got hundreds of optics, we use to manipulate the light in a variety of different ways to do different actions.
For formatting the light, at some point in there, we put it through this specialty device, and this device imprints a pattern into the beam, but it does so in an absorptionless way, which is super critical. When you’re talking tens of kilowatts, if you absorb even a little bit, you’re toast. The micromirror rays that you use for stereolithography, large area stereolithography, they can only take at most a few 100 Watts, right, we're talking 10s of kilowatts, so we're multiple orders of magnitude and power above them. And you needed different technology to handle that sort of power, right? So, we patterned the laser light, we then project it down to the powder bed, and we essentially print a tile in an instant. So that tile, you essentially bring the material up to its melting point, you melt it, and then you control its cooling rate, we actually have the ability to control on a per pixel basis, as part of our process, how we let that material cool, because we can control really the grayscale value of every pixel in the beam while we're printing. So huge level of control.
And because we're essentially imprinting a pattern on the beam, you can think of every pixel in that pattern as its own independent laser beam. So, we're effectively printing millions of points all at the same time, we're taking a process and making it massively parallel. By doing that, we have more time freedom to focus on quality. So that's a really interesting part that we're just starting to turn the knobs on, in terms of making that happen. But we see incredible potential there, especially for talking about printing materials that you can't print today in laser powder bed fusion, like every standard aluminium alloy, many Inconel alloys, etc, that suffer from stress cracking, because they're cooling too fast. So, let's slow the roll.
JDM: Ultimately, we’ll be able to print any material that can melt, but you’ve got to qualify one material at a time, so where do you start? We started with 316L stainless steel, because of a variety of reasons. One, it’s nonreactive, non-toxic, and easily approved by the local fire department to have in a facility that’s owned by a start-up. We also had a lot of history doing work with 316L from Livermore. There’s a lot of literature on it as a laser powder fusion material, and it’s also fairly representative of a tough thermal material, so from a laser perspective, it’s a tough one to do.
As of a couple of weeks ago, we folded Inconel 718 in and we’re doing our first qualification runs on that. It’s not qualified yet but it looks very promising [and is] accelerating, actually, further ahead than steel at some level in some ways. That’s our first two. We are starting with that material because we’ve started our first customer engagement programme and so to that extent, we are customer driven on materials. Our customer engagement programme is what we call our Area Printing Production – or APP – programme. This is a multi-year programme to qualify their materials, their parts, and then demonstrate ramping to production.
Next, we’ll probably be doing either some more Inconels or steels and then likely aluminium, and then titanium, and then broadening from there. But, again, we’re customer driven. So, if there’s the right application that comes along, that could shuffle things up.
JDM: With the folks that we’re talking to right now, it’s less so aerospace and medical, and more so consumer electronics and automotive; industries that have historically been underserved by additive as it exists today. We’re fitting into a need that has not been satisfied, because we’re looking at how do you actually make parts, we talk about by 2030 having parts made for less than $25 a kilo. No one else can get to that point and maintain quality in the process. So, we see an interesting opportunity to solve that need. Initially, there’s what we call the industrial, automotive, consumer electronics, this might include gas turbines, heavy machinery, a variety of components like that. Initially, we’re looking at applications that have relatively short qualification timelines, because it’s important to us as a business. But as we grow, we’re going to be broadening that up, we have to make an initial launch point.
Seurat Technologies
TCT: Can you explain the thinking behind introducing Area Printing via a contract manufacturing service, as opposed to selling machinery?
JDM: We get great economies of scale as we grow our machines bigger. Our machines are already massively more productive than what's on the floor today, we're about 10 times the state of the art of what we have right now, at these quality levels. So, as we grow bigger, we get better economics, and it's a great scaling curve, and we can pass those prices along to the customer. As we get bigger, our machines get more expensive. And it becomes a bigger barrier to entry for selling machines. And you see that machine sales are already lumpy. And machine sales also add on a delay in getting customers to series production, right? Once you've convinced them that your machine can do what they want you to do, then they have to prove out to themselves that they can operate the machine to the point where it gets that same quality, right? And to become an expert in anything is often called three-year time lag. So, whether it's three years or not, you're instantly inserting additional delays. Whereas if you've already qualified parts in a production run, just make more of them. Once you've already gone through that qualification process, you can skip that barrier to entry. So, there's lower barriers in terms of costs, lower barriers in terms of time, part sales look more like recurring revenue, instead of lumpy revenue that's from machine sales.
It doesn't have the right architecture for where we see ourselves going, which is ever bigger machines, which means ever lumpier cash flows, which would mean ever higher barriers to entry for machine sale price. And I think we often use the phrase, 'know what you want to be when you grow up and then chart a path to get there.' We're starting with selling parts for customers in our APP programme. That's how that programme works. And we see when we want to grow up, the lowest barriers to entry, the best commercialisation way is to sell parts.
JDM: What post-processing steps we do as part of this, and what we do in-house versus what we do outside, those are going to be business decisions, various contract deal negotiations, I’m sure we’re going to have other outside vendors, and there might be joint partnerships that go on as well. How that all unfolds? Great question. We’re providing a black box solution, where you give us a CAD file and your specs, here’s your parts, and we’ll take care of the other details. When you are at a mature technology level, that’s what customers expect. They want to have parts made that meet their specifications, so that they can make money. It's only when the technologies are immature, and unreliable, as additive generally is today, that they want to really get their own hands dirty, because it's not moving fast enough for them.
JDM: It’s now eight letters and the first one of them has started. That was in the beginning of February. This is the APP programme that I was mentioning earlier. It’s a multi-year programme that we break up into three phases. Phase one: Qualify their material. Phase two: Qualify their parts, demonstrating repeatability, reliability, demonstrate that we’re hitting the economic targets, hitting the target quality targets. [Phase three:] and then we start ramping up volume for them and demonstrating that we can maintain all those targets as we do production ramps, and then at the end of that you’ve gone through your ramp and phase three and you’re entering series production. Again, it’s a per customer, per material, per part way to get those customers to where they’re looking to go. For this reason, we’re really only looking for applications that are of moderately high volume.
The first company [we’re working with] is in automotive and energy-related turbochargers as an example. One of our companies is in consumer electronics. We’ve got a couple more industrial and automotive. Obviously, there’s a lot of applications from the US government, various branches of military and so forth. There’s a fairly broad and diverse set of interests, but that gives a general flavour for what we’re looking at there.
JDM: The number one is we've got increased customer demand that exceeds our capacity to handle it in the near term. So build up our capacity so we can handle that. This year, we're going to be able to handle really four customer slots. We've filled two of them, and we're in the process - I mentioned we have eight letters of intent - so we've got six more that are sort of on their way imminently. And we've got two that are open. So we got to pick which ones are the right ones for us to move forward with. And we also got to extend our capabilities to handle more. These are the ones that have given us letters of intent. As of December, we had 25 in the funnel and that’s a number that we see growing organically, so we need to increase our ability to facilitate customer demand.
And then, supply chain right now is a mess, and one of our biggest concerns as a company is making sure that we can still deliver things somewhat remotely near our schedule. We have areas where we’re using funds to help us secure multiple sources, help us make sure that the work we’re doing is going to be on track. There’s hiring folks, there’s setting up the facilities and equipment, there’s also some small amount that’s going towards the next gen of some really long-term items, but for the most part it’s facilitate those customers.
JDM: I would say democratising manufacturing. I like to use the analogy from Aaron Levie at Box: Envision the car market based on the number of horses and buggies in the early 1900s, you just couldn’t do that, you couldn’t fathom what’s out there. I think we’re at the same point here. When we get additive costs to below that of conventional manufacturing and you have all this design freedom that comes with additive, we’re going to see an explosion of creativity, and an explosion of use cases and applications and just things you never thought were possible.
You can take a micro lattice structure embedded into steel, and make it stiffer and lighter weight than carbon fibre yet, ultimately, unlimited recyclability and very high melting points, all the advantages you have with steel, no fatigue limit. You could make steel behave like aluminium. Aluminium is used in aeroplanes, and so forth. There's only so many times that things can flex before it will break, right? We know this. It's an engineering fact. If you make it as steel, you can design to the point where it never will. So anyway, this is just a couple of examples. But there's just an immense wealth of creativity that we're going to see, as you can democratise manufacturing and allow the designers to really leverage that. Right now, they’re told, ‘no, you can’t do that, because it’s not cost effective.’ So, take away that barrier.
That’s what I’m very excited about. And everything that comes when you do that: the sustainability portion, bringing jobs to where you're manufacturing things instead of sending them off to the other side of the world to be in a far-flung community, do them locally. So, I think that's all part of the big vision of how we do things. Distributed manufacturing, democratise manufacturing, making a world that's better for the people on the planet.
This interview has been edited for brevity.
Want to discuss? Join the conversation on the Additive Manufacturing Global Community Discord.
Get your FREE print subscription to TCT Magazine.
