“This course already changed our vendor interactions.”

Working as a particle accelerator engineer, Curt Preissner ran into the limits of their design philosophy. Which is why he and a colleague took the Mechatronics system design (metron) – part 1 course at the High Tech Institute. This allowed them to introduce a new design approach into the synchrotron community, and better talk to vendors. ‘You need to be able to communicate what keeps you up at night.’

When Curt Preissner took the Metron – part 1 course at the High Tech Institute in Eindhoven, he was impressed by the local expertise, but also the amount of bikes riding around. ‘I bike to work here in the United States, but it’s not at all like in The Netherlands’, he looks back fondly.

Preissner is a mechanical engineer at the Advanced Photon Source (APS), a U.S. Department of Energy (DOE) Office of Science user facility at DOE’s Argonne National Laboratory in Illinois. This synchrotron, a type of circular particle accelerator, generates radiation in the form of x-rays. These x-rays in turn can be used to, for example, make images of the nanostructure of materials. Preissner is designing a very specific component in that system.

‘In a particle accelerator you accelerate electrons with the use of radio-frequency energy’, explains Preissner. ‘They then oscillate back and forth between the north and south poles of magnets, which produces what we call synchrotron radiation. In our machine, that radiation is in the form of x-rays. The energy of the x-rays we produce ranges from a few KeV all the way up to 100 KeV, so it’s highly penetrating. We take those x-rays and use something called a monochrometer to select a particular wavelength. The instrument I’m designing is an x-ray microscope called the PtychoProbe. This will be a unique, world-class instrument, and it will focus the x-rays down to five nanometers, which doesn’t exist right now. So it will be a world’s first. The x-rays will be focused on the sample and diffract off of it. The diffracted x-rays from the sample will then be collected by a detector, from which we process the data to generate an image that shows the structure of the sample.’

Curt Preissner, credit: Mark Lopez Argonne National Laboratory

New engineering philosophy

Preissner and his colleagues realised that this new design, which demands high degrees of precision, would require them to adopt a new engineering philosophy. ‘The specifications we work with can be very challenging’, says Preissner. ‘Generally, our system is static. Yet on the side of the beamline, things are moving. We have to scan our samples in a different way because the new beams are much more bright. This brightness will allow us to see our samples in greater detail. However, this high photon flux can actually damage the sample, and prevent us from seeing these details . So, we want to do this quickly. We don’t need to work as fast as some semiconductor manufacturing equipment. We scan around seven millimetres per second, which aren’t extremely high velocities. But for what we’re used to, this is quite high. The sample and the x-ray lens, called the zone plate, also needs to maintain registration on the order of 1,25 nanometres. That’s pretty tight. We do that over length scales of about 10 millimetres. This is new territory for us. Which is why we’re looking for new engineering approaches to achieve this.’

Integrated approach

After some research, they realised that mechatronics could offer an answer. ‘We first started in the synchrotron community, which isn’t that big’, explains Preissner. ‘There are a countable number of synchrotron instrumentation engineers, probably around a few hundreds, less than a few thousand for sure. The community is not that big. So when we didn’t find the answers we were looking for, we started researching other fields with similar performance specifications. This is how we ended up with semiconductor manufacturing equipment, and in turn the mechatronics approach.’

This approach, while common in some fields, is new in the synchrotron community. Mechatronics, however, might be what they need to keep pushing the technology forward. ‘In the last generation of instruments, ten to fifteen years ago it wasn’t uncommon for a mechanical engineer to sit down with a beamline scientist and just design the mechanics, connect a motion controller, maybe some interferometry, and achieve results that got the scientific job done. The only consideration to dynamics in the design was vibrations, and there was certainly no system-level approach.’, says Preissner. ‘But now the advancements in the accelerator and x-ray optics technology are really forcing us to  push the limits of what we can do. That old approach will not work.,. We need to look ahead; science does not allow us to stand still. The instrument I’m designing will need to be scientifically productive for at least the next ten years. A mechatronic approach is very interesting here. It’s great to think about things like error budgeting and dynamic models from the get-go. It’s a more integrated approach.’

Ending up in the Netherlands

Which is how Preissner and a colleague ended up in The Netherlands taking a mechatronics course at the High Tech Institute. For them it was the ideal way of being quickly plunged into the field. ‘At the APS we don’t always have the luxury to be able to do a huge amount of R&D’, says Preissner. ‘We’re in a time crunch with this project. We need to gain knowledge fast, so we can work with vendors or do our own design. If you look at for example the wafer scanners of ASML, their performance is very impressive. But an important thing to remember is that there’s roughly forty years of development behind them. When we’re designing these instruments we don’t have that time. We need to learn as fast as possible.’

Vendors

One important thing they learned in the course was a new type of language, which allowed them to better speak to their vendors. ‘We’re not just going out and buying something’, says Preissner. ‘We’re proposing things, and deciding whether a vendor can make certain designs. So knowing techniques like error budgeting is important, besides being able to look at designs with a mechatronics view. Getting some formal training accelerated our ability to talk to vendors. There’s certain key issues in this design that keep me up at night, and we need to be able to communicate that. After the course I could go to a vendor and ask them to, for example, show us their error budget. Or I could talk to them about the controller dynamics overlaid with the mechanics dynamics.’

Short timeframe

The course taught them this in a short timeframe. This is important for an engineer like Preissner, who is working on a time-sensitive project for a government-funded organisation. ‘We’re under a high amount of pressure, so we were eager to learn, and did so quite fast. We looked hard for a course that could quickly package this knowledge for us. APS is also a government institution, so we’re using tax dollars. We need to be mindful of how we spend them. We’re always looking at ways to achieve goals in an effective manner, and this course taught us what we needed to know very efficiently.’

All of this is a work in progress according to Preissner. ‘The synchrotron engineering community has been operating in a certain way for a long time. But now people realise that we need to do things differently. This course enabled us to take that different approach.’

Model in a holistic way?

So far the new mechatronics knowledge has mainly been used in contacts with vendors. But Preissner notes that going forward, they want to also use it to design new instruments from the ground up. ‘It’s on the drawing board’, he says. ‘We are wondering if we can take this new approach, and apply it in a more systematic way. Can we model the instrument, the control system model and the influences in a holistic way? What knobs do we need to turn? What control approach would make sense?’

For now, however, Preissner and his colleagues want to expand their knowledge of mechatronics. They’re already looking forward to taking more courses. ‘If you don’t use it, you lose it. So we’re feeling some pressure to apply what we learned as regularly as possible. The first part of the training was good, and now we’re thinking about taking additional courses. When you learn new engineering techniques it takes a bit of time. You have to work with it. It has already changed our vendor interactions. The next step will be changing our own designs from the ground up.’

This article is written by Tom Cassauwers, freelancer for High-Tech Systems.