Working Hours:
Monday - Friday, 9:00 - 17:00
ASML takes on the role as a driving force for high-tech AM
“It’s fun, it brings freedom, we want it, but the quality must be good.”
What role will additive manufacturing play in the production of advanced machinery, like metrological instruments and lithography scanners of ASML? We asked Arno Sprengers, team lead precision mechanics at ASML, about his vision.
ASML has emerged in recent years as a driver of innovation in manufacturing processes with the great promise of additive manufacturing (AM). Due to the great potential for both cost reduction and functionality, various suppliers are currently investing heavily in 3D manufacturing capacity. Are we currently witnessing a new phase in the maturation of AM?
When we ask Arno Sprengers, he initially chooses a more relativizing tone. Additive manufacturing has not yet reached the level of conventional metallurgical processes. ‘It took two thousand years to come to a block of material with known properties.’
Sprengers took the initiative ten years ago with Summa, Fontys, and the TUE to establish a center for knowledge sharing. Within the Knowledge Sharing Centre (KSC), knowledge about 3D printing is one of the focus areas.
At ASML, he is currently co-establishing a competence center for additive manufacturing. There, he works with a handful of people who focus, among other things, on a General Standard of ASML for AM. These guidelines for suppliers cover a wide range of areas for the lithography company. In addition, he gathers knowledge about manufacturability with the team to support ASML’s designers in making good designs.
There are quite a few challenges to bringing additive manufacturing to the level of materials that the high-tech industry has been familiar with for years, such as aluminum, nickel alloys, steel, and titanium. Sprengers: ‘A designer gets a specification and starts thinking about the behavior and possibilities of the material in the application. If he or she takes titanium or stainless steel, a machinist works with certified material. The designer knows the properties of the part when it comes out of the factory. The problem with 3D printing is that we do not yet know that as well. We and the entire market are working hard to guarantee the quality of a material in the end product.’
While metal powders for 3D printing do come from bulk processes, this base material then enters a micro-environment that is still insufficiently predictable: every 3D printer thereby creates its own version of the material. This inherently gives rise to many uncertainties. On the market, there are 3D printers from many manufacturers, and even identical printers can produce products with differing properties.
What exactly is the challenge?
Thinking in processes. The machines are, by definition, different, even 3D printers of the same brand. But the user sets the parameters and must measure over time what the stability of his product is.
“We think we can make a difference with additive manufacturing for parts that need to move very quickly.”
This is much more about thinking in processes than turning or milling. It is not comparable.
That means we are now having discussions with suppliers to conduct statistical process control (SPC, ed.) on their additive manufacturing processes, which might involve fifty parameters. They have to measure and maintain all of these constantly. The goal is to achieve an outcome with predictable quality.
Which parts of your machines are the first to be considered for 3D manufacturing?
Then you assume that we will manufacture existing parts differently. That is too simplistic. You have to start thinking in additive manufacturing. That means combining functions to achieve a much better effective outcome. Much of what you can’t achieve with conventional technology, you can achieve with AM. That makes it mega-interesting. We think it brings us to the future. But in the meantime, we don’t want to lose that quality. We want to have predictable quality.
Challenging: “ASML always goes to the extreme. We push the boundaries of manufacturing methodology. With AM, we want to achieve a wall thickness of half a millimeter, but with the assurance of a specific quality. If we can make that match, then 3D printing is a great tool. But we are not there yet. It is not that we will no longer need conventional techniques. We do want to apply AM where necessary.”
But the question remains whether it is even possible?
“Again, ASML is ASML. We trust that something will be possible. We don’t think too much in terms of impossibilities, but in terms of possibilities. We now need the whole market to think along with us. The challenge is to think in processes. That is really something very different from conventional machining. Most suppliers we currently work with are conventional machining companies. They start with a block of material that has already undergone a process. They give it a shape that makes the part usable. Until now, they did not have to worry about processes.”
“Just like conventional parts, you can measure the shape of AM products. But for the internal structure, you need something like computer tomography. That is expensive and difficult. We do not want to rule out CT scanning, but it is not feasible for every product. We need to use it in the chain to control quality. We believe in it one hundred percent. It brings ASML into the future, it’s just not mature enough yet to be used everywhere.”
There are already many printed parts in an ASML machine, such as cooling channels.
“It makes a big difference whether you load something statically or dynamically. If you load something statically, you can test if it is leak-proof, and it stays leak-proof. If you load something dynamically, you can do a test, but then you don’t know what will happen in the long term to that part, when it is exposed to varying levels of forces and pressures. That makes it a much more complex ballgame.”
You are currently setting up the ASML competence center for 3D manufacturing. What needs to happen?
“Most competencies within ASML are design-oriented. But for AM, it is end-to-end. We do both design and quality. Otherwise, AM will not achieve the right quality. That means the supplier must take ownership of whether the design is also manufacturable. That means we ask them to help in the design process to make it manufacturable. On the other hand, we as a machine builder have to think along to get the right parameter sets in the 3D manufacturing process. The supplier must secure that data and ensure that his machines remain stable enough over time. We think we can eventually deliver the right quality with statistical process control. But that is quite complex.”
“Don’t forget that working with competent people in the chain is very important. That means training the people in the entire chain is also very important. For this, we want to set up a learning environment together with Additive Center, which will be called Amify. In it, we give the entire industry the opportunity to think along. It is learning from and for the AM industry.”
Are you going to redesign parts of existing machines from scratch?
“That is not the way to go. We come up with a new platform approximately every five years. The mechanics of older and existing generations are completely tuned for stiffness and weight. With each platform, designers agree on volumes with each other. You cannot change the volumes afterward. That would be a hassle. Changes are always significant. Therefore, you would need to start an entirely new project with a large team. You cannot simply say: I’ll just print this part instead of milling it.”
So, 3D is not a tool to reduce costs in existing platforms?
“No, you really need to look at new platforms. For existing machines, it does not save any costs. I do not want to rule it out completely, but you cannot use AM as another manufacturing process. You need to rethink and restart the design process.”
But it fits for a future generation of metrology or lithography machines?
“We think we can make a significant difference with parts that need to move very quickly. Then mass plays a very big role. By combining parts, we can reduce the mass considerably. This allows us to reach a much higher acceleration. That is one of the main points that make AM interesting for the next generation of machines. A lighter design means more speed. These steps are typically taken when developing new platforms.”
ASML already uses welded titanium in carrier structures for wafer stages
“In that case, you machine less material. That is effective and favorable, but that is not the AM I am talking about now. It’s about lighter, better, ideal shapes, and more design freedom.”
Is the material from 3D processes strong enough?
“If you can achieve good material quality, then 3D printing is a great application. Is it the only way to manufacture? No. It is one of the ways, nothing more.”
What are the things that make you hopeful?
“The design freedom is great, but the problem is that you need to collaborate better throughout the entire chain to come to effective, good end products. It transcends boundaries. The 3D manufacturing technique is not yet equivalent in maturity to conventional techniques. As an organization, you must also genuinely want to change and think in processes instead of conventional techniques. You need to do statistical process control.”
Isn’t it just a matter of setting up five printers and letting them run smoothly?
“No, in my opinion, absolutely not.”
How are you going to help your suppliers?
“We want to use our skills, but the responsibility must lie with the suppliers.”
At the same time, you want to know exactly what to do?
“That makes it extra difficult. The responsibility for applying the correct parameter set and the proper control over it lies with our supplier. We want to discuss it with them, but not take over the responsibility. That should remain with our supplier. That makes it challenging, because you are working with different parties. Actually, we should perform SPC without gaining insights into the IP of the machines. That makes me believe in dashboards providing an overview of the delivered quality.”
“We want to continue doing business with multiple suppliers. That means statistical process control over multiple machines. Our suppliers and their machine suppliers will therefore have to compile that dashboard and share it with ASML to convince us that the material is okay, without insight into the intellectual property of the machine builder. That is very open and increases the maturity of the process.”
Will 3D printing make a significant contribution to the cost reduction of your machines?
“Yes, I believe so. How, is hard to tell. But nonetheless, ASML reaches their goal of making machines which are needed. 3D printing offers a freedom that one hundred percent contributes to what ASML needs. Without AM, it can be quite difficult to achieve it, so we are eagerly waiting for 3D printing to work for us. It’s great, it brings freedom, we want it, but the quality must be good.”
“The unfortunate thing is that we need products to learn what quality we can achieve. Where we set the boundaries. What dynamic loads we can tolerate.”
So it wouldn’t surprise you if ASML starts 3D printing parts of wafer stages?
“That is definitely going to happen. I firmly believe in that, but you should ask an ASML employee.”