(In January, just before the arrival of the high field magnet in Berlin, I was interviewed by Kerstin Hoppenhaus for the #HZBzlog. This is the English version of the complete interview.)
There really is a whole team working on the High-Field Magnet. What brought this team together?
The High-Field Magnet is not the average project for us here at HZB. And if you do a project beyond your own area of expertise, then there have to be people from outside as well. They all had to be found. For very specialised tasks. And for that reason they are quite a varied bunch. Everyone has very specialised tasks and very specialised talents.
What were you looking for in particular?
Actually, there were two different things. We have only a few, but important administrative areas in the group. For that reason, one of the first hires was our Controller. It is always important to have a comprehensive view of the finances.
And the second important area is certainly the engineering work. We have several physicists in the group also, but the engineers are almost even more important for us. This is because we couldn’t just simply discuss the finer points of the design, but rather had to actually get the thing built. And then someone had to do the calculations in detail for it. Starting with Matthias Hoffmann, he is a mechanical engineer who can design and lay out the complete details of his cryostats in 3D with the accuracy within a hair’s breadth. That is important – so that everyone understands what their area is and how these interface.
Or when you look at Stephan Kempfer, who is an electronics engineer who looks after basically all the technology in the infrastructure building. He works closely there with Jochen Heinrich, who is responsible for the helium liquefaction. They have to coordinate things really closely.
Robert Wahle joined the group a bit later and is somewhat closer to the science end. He is building what are called sample cryostats, the devices that make sure the samples can later be held at various temperatures necessary for the measurements.
And there is our technician Christof Fritsche. A technician probably does not have such an easy time of it in a group like this. Everyone pulls him in different directions – give me a hand here, help me out there... And we also travel a great deal, which many technicians normally don’t care for that much. We had a stroke of luck in that Fritsche could already speak excellent English, which helped make our traveling easier.
Hartmut Ehmler also had a lot experience traveling. He was with our colleagues at MagLab in Tallahassee even more frequently than I was, weeks at a time, and made sure that we really understood everything they were building for us there. Nothing would have been worse than not knowing exactly what we were taking delivery of! At some point, we will be standing there having to operate the thing by ourselves.
The High-Field Magnet is a hybrid magnet, which is rather unusual. Why are you building it this way?
“Normal” magnets today would actually be the superconducting types. You can do a great deal with them and we have already accumulated a lot of experience with them here at HZB and previously at HMI. But the way things are now, superconducting magnets have several physical limitations.
So then you consider utilizing resistive magnets - at enormous cost, because they require huge amounts of electrical power and result in unbelievably high operating costs. Or you have to increase the complexity of the thing. You have to combine superconducting magnets of a particular type, referred to as cable-in-conduit coils, which are available, then add on as many resistive magnets while keeping the operating costs within reason. Then you get a considerably higher magnetic field.
This is not that easy. You very quickly have to contend with various technical limitations. These are, for example, the very high forces that can arise between the individual coils, but primarily it is the high complexity. All at once you no longer have just ten, but a hundred things that have to work simultaneously. And ‘simultaneous’ really does mean that. If one thing doesn’t work, everything shuts down. And that is a real challenge. Which is why a hybrid magnet is rather rare.
The expenditure for the magnet is pretty large. Development and construction have taken years, the whole infrastructure needs a separate new building – how has that happened?
The keyword as to why it is all so complicated is hybrid magnet. You have to meet quite different technical requirements for the superconducting and resistive coils simultaneously. This produces large-scale systems in the end. And there are not many people in the world who want to go to the trouble of building a device like this. But it offers the means of creating very strong magnetic fields while at the same time keeping the operating costs within reason.
You need to keep in mind that in creating such strong magnetic fields, we are still not very far along actually and that we are really working at the edge of the envelope. We are improvising with things that we have – the superconducting and resistive magnets – even though they are not very compatible with one another, to enable us to carry out leading edge research. In the hope we learn enough in the process that things with the next magnet look a lot better.
What is it exactly that makes hybrid magnets so complex?
Well, there are two aspects. Firstly, it is frankly always better to have just one coil, rather than two. This is because you do not have as many forces working against one another, and you do not have to be so extremely cautious and careful with the many mechanical aspects, or with the electrical ones either. In normal operation, everything continues working fine with two, but quickly becomes dangerous if small perturbations occur.
The other point is that superconducting and resistive magnets are worlds apart. You have one magnet that you must cool to very low temperatures using liquid helium. You have to put it in an evacuated container and go to a lot of expense and trouble to prevent thermal radiation from penetrating the container. And then you have to combine it – with as little spatial separation as possible – with an installation that creates a great deal of heat, which you need an enormous quantity of cooling water for. So you therefore have one unit operating at very low temperatures in immediate proximity to another unit which you have to pump several dozens of litres of water each second through in order to get rid of the heat. And that creates vibrations, among other things, which naturally you don’t want. Incompatible operating characteristics simply have to be made compatible.
The reason is simple – they are two standalone systems that separately evolved over a period of years. Without one system considering the other. Then someone comes along and says “Now we have to co-locate and jointly operate the two”. That doesn't work at first, of course. Then you try to find compromises. There have to be compromises. And that is what makes it so difficult probably.
You developed and built the High-Field Magnet in close collaboration with the National High Magnetic Field Laboratory in Tallahassee, Florida. How did this collaboration develop?
At the moment, there are just four hybrid magnet projects world-wide. That is not very many. And there is actually only one single hybrid magnet at the moment that is being used in what we call scientific user-service, i.e. where scientists can go with their samples and make their measurements. And that is the hybrid magnet in Tallahassee.
Which means we did not have very many choices when it came to partners for the collaboration. What made Tallahassee attractive in addition is that the working group there is large enough. They have brought together breadth and depth of expertise in a group of forty or fifty people, so if you have to deal with many different issues, you can get them resolved through various people within the group. Despite the great complexity. It is doable there.
The magnet consists of many different components. How do you make sure that everything works together in the end?
There are a lot of opportunities to run tests during initial development, and you have to use every opportunity, because a magnet is indeed so very complex. But you are correct – we only get the really crucial result at the very end. And you have to make this risk as small as possible. For a superconductor, that means you cannot be 100.0% certain as long as the superconductor has not been cooled and then hit with power. And that can only happen once the coil has been finished. Naturally, you test the conductor by itself, at low temperatures as well. You run the power through and see if the conductor withstands the forces. But you can only be absolutely certain that the coil you’ve wound from this conductor also really works at the very end. That is the difficult point and really also the tough demand you have in working with superconductors…
Things are not so terribly complicated with the equipment and technical facilities, but here as well, all components are individually tested prior to commissioning the magnet.
That means you have to live with considerable uncertainty over quite a long period of time.
I don’t think there is any other way. It’s like this. You think about how it would be if life could be a little simpler – you would always like to have as much certainty as possible. But you also have to accept that at some point, a balance has to be struck between the efforts you go to, which all costs money that you have to get from somewhere, and the result. Risk assessment is a very important, but also very difficult task in almost every project. And I believe ... it is almost a philosophical matter, because zero risk is impossible anyway. We simply have to deal with it. The question is how close to zero can you reduce risk? Physicists recognise this. They call it approaching something asymptotically...
One of your most important tasks in this project was establishing and maintaining communication between the various partners. Were you prepared for this?
Actually I wasn’t. Communication is difficult at the best of times. People always experience this in their careers, once they have been around for awhile – that things don't work due to human shortcomings, even though they could work technologically. When you then notice how surprisingly large the differences are that can be seen between people, when you are working in an international arena, then I must say it is experience you have acquire yourself. I didn’t have this experience at the beginning.
And how did you deal with that?
You have to have the determination to find a solution, to find a way, and then also be flexible at the same time. To be willing to discuss things, even when difficult, and find areas of agreement. When various highly-qualified people work together, they all have to be convinced, they have to believe unanimously that the right way has been found. When only one person leads the way and the others say "Well, I will go along with it, but it is not going to work anyway” – that’s no way to proceed.
It becomes really difficult once there are no more technical arguments left. You try to get back to the engineering, to the calculations, that always works as a last resort. Then things become objective and you can come up with criteria. Saying that “I just don’t get along with this person well enough” is not a criterion.
The partners in this collaborative effort on the High Field Magnet are spread halfway around the world. How do you manage to communicate reliably?
That is an important point. We are working with very complex things and live permanently with the risk, sometimes a bit more or a bit less, that misunderstandings occur, that something hasn’t been explained well enough. You produce a lot of paper, you write a lot of emails, you meet via video conferencing, you communicate over the phone ... but despite all this, real certainty only ever comes through a conversation in person. It is only then that you can examine all the details in depth and in case of doubt, check whether your counterpart understands everything that you want. Because I will not get what I need if I haven't explained it well enough. And that doesn't happen without personal contact.
It is always enormously important that you can watch someone doing the work, and what is being worked on where, and how. They are such simple things like how the work place is organised, workplace safety, whether there is any risk that the work performed will be defective because the tools are not up to the task - there are lots and lots of little details. The worst is the more you think about things, the more things occur to you. And you can really only check them if you are on-site in person.
Perhaps we shouldn’t complicate it with reasons. It is simply a matter of conscientiousness. You have to be convinced that what is meant to happen actually will happen. That is a question of personal responsibility. Just writing up a contract and a couple pages of technical specifications – that would be unimaginable.
With as many things that can go wrong in a project like this, have you been able to sleep well the last few years?
Not always. You wake up lots of times... But you go back to sleep again too. Psychology is sometimes a difficult topic. Even if you have assembled as much information as humanly possible, the thought always surfaces: have I overlooked something, is there any kind of risk still, is there anything else that can be done? Because in the end, you simply want to get it right. And that leads to waking up in the middle of the night occasionally.
What do you do in case of doubt?
I have to articulate that doubt. A well-reasoned question has to form in my mind that I can pose to someone. And then I reach for the phone. We have various people with expertise in various areas, and I can discuss it. That is a most important point, if something is unclear: try to talk about it.
And also the other way around, in case someone comes to me with doubts. Not to shrug them off. Listen. And try to find an answer. It is possible that the answer is simply “impossible”. But it can also be possible, and necessary, to think something over, re-calculate something, or possibly even take action.
The magnet has now arrived in Berlin and this is the final stretch, so to speak. What is the schedule for the next steps?
Scheduling is always a difficult subject ... my aversion to talking about scheduling is a characteristic that has grown with the increasing duration of the project!
We think we will begin the cool-down in summer. But a hundred things can still go wrong that cause delays, because repairs are needed, etc. I hope that we do not have to chase down a leak again for a couple weeks like in Chivasso. And then we might just get started in summer... But we also do not know precisely how long it will take for the coil to cool down. A couple of weeks...
And then we will carry out all possible electrical tests very, very cautiously. Because when operating a superconducting magnet containing this much energy, you really have to have everything under control. All the control systems, the emergency shutdown systems, they all have to be tested in turn. That takes time... We are still writing the operational testing protocol, and I assume that the protocol will continue to be written even during the operational tests. And things will continue to occur to us that we want to test. It is just too important. We have invested too much effort to permit any risk during commissioning!