Thanks for visiting Scientific Computing World.

You're trying to access an editorial feature that is only available to logged in, registered users of Scientific Computing World. Registering is completely free, so why not sign up with us?

By registering, as well as being able to browse all content on the site without further interruption, you'll also have the option to receive our magazine (multiple times a year) and our email newsletters.

A breakthrough in fractures

Share this on social media:

Topic tags: 

Peter Gumbsch, professor at the university of Karlsruhe and head of Fraunhofer Institute for Mechanics of Materials

When engineers build things they are usually happy when it works and look no further. Peter Gumbsch is a scientist who wants to know why one thing works and another does not. This involves looking, at a fundamental level, at the materials the engineers use and trying to work out why they have the bulk properties that they do.

His work on the simulation and modelling of materials was quickly recognised as groundbreaking in many respects and his career has taken off accordingly. He has rapidly risen to a full professorship at the University of Karlsruhe with a joint appointment as head of the Fraunhofer Institute for Mechanics of Materials. This year he was awarded the most prestigious prize for a scientist in Germany, the Leibnitz Prize. This comes with a research grant of †2.5m.

It surprises many that someone whose work has been so highly theoretical has ended up heading an institute known for its applied research, but Gumbsch is highly driven by the possibilities his fundamental research could open up for the real world.

Professor Oliver Kraft, from the Institute for Materials Research, in Karlsruhe, says: ‘I have known Peter since we were graduate students and I have always been impressed by the way he interacts with people. Of course he has a strong background in computational material science and I am an experimentalist. I have found – with other theoreticians and modellers – that it is sometimes difficult to interact with them because they often do not understand the limitations of experiments, but Peter is very good with that. In fact the institute he leads is mostly doing experimental work, so he has been able to broaden himself in that way.

‘We both have families of the same age and we often used to get together with our families and have fun. He is very much a family man and, while he works long hours, he makes sure he is up to make breakfast for his children. He is very good at interacting with everyone, even his students, with whom he has a casual relationship that is unusual in Germany.

‘He is probably the leading person in Europe for the atomistic simulations that he works in and he is co-ordinating some very large European projects. He has become extremely well known. He has had many opportunities to move to the US and other European institutions and it is his choice to be where he is now.

‘His most important scientific contribution has probably been in really understanding fractures. Fracture mechanics has traditionally treated materials as a continuum and it was only about 20 years ago that people started studying it at the microscopic level. Peter’s contribution has been to take it down to the atomistic level. Of course there are others in the field, but he is leading in understanding fracture at an atomistic level. More recently he has been modelling deformation and these two things are his major contributions.

‘The Leibnitz Prize is probably the most important scientific prize in Germany. It’s a lot of money and, in principle, you can stop writing proposals for three or four years and just do pure research. The nice thing is that you can try some of your crazy ideas that would never get funded and hope that one or two work. He is always looking to try new things, so I think he might try a few of these.’


Peter Gumbsch

Gumbsch was born in the Black Forest region of Germany and grew up in the Stuttgart area. His father worked for the German postal service and his mother was a full-time mum. He did well at school and, from an early stage, he found maths and physics his easiest subjects. He was a keen swimmer, competing in regional championships.

A good all-rounder, he found it hard to chose what to study at university, as he also had an interest in economics. In the end he took physics as his main subject at Stuttgart University with a minor in economics, which he studied by distance learning. The economics degree was never finished, however, because he could not get to grips with the law side.

Gumbsch took seven years to get his degree, as he was working as a teaching assistant while he studied. He also got to spend some time at IBM Laboratories, which inspired him to work in industry rather than academia.

He says: ‘I liked this much more than teaching and my main aim became to go into industry. I was working on an electron microscope and analysing materials, so when it came to deciding what I should do for my diploma thesis I decided to go for material science. During that thesis I became totally fascinated by the electron microscope and looking directly into materials and trying to make sense of the structures. I really got into research at this stage and, shortly after starting my thesis, I was sure I wanted to do a PhD and postponed my ambition to go into industry. I continued to study at the Max Plank Institute for Materials Research in Stuttgart.

‘At this point the Max Plank Institute started a research group in atomistic modelling, so I switched from electron microscopy to do modelling and simulation for my PhD thesis. It was still the early days of computer modelling and it was hard to get the computing resources to do something meaningful. The modelling itself is very similar to what we do today, but in those days we worried more about boundary conditions. We started as just four graduate students working on this and it was only when I was near the end of my PhD that we were joined by an experienced group leader.

‘First, we developed a technique of coupling atomistic simulations to finite element techniques, so we made some progress on something that is still a hot topic today. Second, we applied this technique to fracture; real atomistic detail of bond breaking at a fracture. This was also ground breaking because there was not much done on this topic before. We did not publish many papers, but those we published were well cited and people liked the work.’

Gumbsch’s work on cracks attracted a lot of attention at various conferences and, in particular, it resulted in an invitation to study at Imperial College London from David Pettifor. He stayed there for a year and then spent six months in Oxford.

Gumbsch says: ‘I had worked on coupling atomistics to larger-scale simulations and investigated crack tips. By doing this I found out that the atomistic models that we used were really questionable. David Pettifor was working on much improved models and I had met him very early in my PhD thesis. He said that if I was interested he would find the money for me to do the same work using his much better models. I wanted to learn how to do this modelling better and to apply more fundamental techniques. I also wanted to stay in research and I thought it would be good to go abroad and learn something. I had also spent some time abroad working at Livermore. It was quite unusual to go abroad while still a PhD student.’

He returned to the Max Plank Institute to continue his research, but his PhD advisor retired during this time. About three years later he moved to another department and got a junior professorship, but the role turned out to be more that of group leader than professor. He wanted to become a full professor, which meant looking for a university appointment.

Gumbsch says: ‘I was originally a physicist, but I had become involved in materials science and had got my fingers a bit too dirty to really be regarded as a physicist any more. It was difficult to find a materials science chair in Germany that would suit someone with a physics background. Eventually I was offered an engineering position at Technical University of Braunschweig and had another offer from Ohio State University, which was looking for someone to lead a new materials modelling group.

‘Then came the offer to take over the chair of mechanics of materials in Karlsruhe as well as the position of director of the Fraunhofer Institute for Mechanics of Materials. That kind of joint appointment is extremely difficult to refuse. They were looking for someone in the mechanics of materials with a materials background and some knowledge of mechanics, but with a physics background. It suited what they were looking for quite well and I guess they liked me. I don’t know what made them decide in my favour, but it suited me very well.

‘I knew a few people here, but it was important for me to have the split position, because there is a lot of management work in the Fraunhofer position. I had always done fundamental research before at the Max Plank Institute. The Fraunhofer Institute is only 20 per cent state-funded and gets the rest of its funding from industry.

‘When I joined I saw the opportunity to get more modelling into industry. The institute did get the attention I was hoping for and we have grown about 50 per cent since I joined. There were already a lot of people working in modelling and simulations, but it was mostly finite element when I joined.’

Gumbsch is extremely excited about the possibilities presented by the Leibnitz Prize. Its †2.5m cheque comes without strings attached and is designed to encourage its recipients to try something that would not ordinarily get research funding. He wants to broaden his research away from the classical materials engineering questions to look at things like proteins and how they adhere to surfaces like bones.

He also wants to study friction, which is yet to be fully explored at a fundamental level. Gumbsch says: ‘The really fantastic thing about it is that it gives me the opportunity to start something new. I can’t really complain as the research money in my area is quite good. When it comes to modelling and simulations of chemical processes, surfaces under stress, friction and so on, it is really ugly and dirty and not well understood. If I had to justify doing these subjects in a proposal, it would be very difficult. There is a lot of fundamental work that needs to be done before I could put together a sensible research proposal. This is something I can now just try, but I will probably stick to the harder materials for now.’

His research interest has focused on what actually happens inside a material when it is under stress and fractures. It starts with a model of the interactions between atoms in a metal crystal and then models what happens when there is a distortion. The field has moved on to considering a dislocation as a discrete object within a material, considering this to have properties that can be modelled, and considering the bulk material to be simply a medium in which the dislocation moves. One of the aims of this work is to understand processes like metal fatigue, with a model based on fundamental principles. Ultimately the properties of materials could be predicted and, in the long term, new materials could be designed in a computer according to its application.

Another strand is to fully understand the properties of materials as they are used, so that behaviour under all conditions could be predicted. This could lead to lighter cars, which are more fuel efficient, or else some change in the manufacturing process could result in a product that lasts longer, because it suffers less from metal fatigue.

Gumbsch says: ‘If you can go closer to the physical edges – to the capabilities of the material – then you find that there is a lot of room left in existing materials. You might find you don’t need such large tolerances and you can save material and save energy.’

He is also looking to expand the scope of the Fraunhofer Institute modelling into other engineering challenges. At present most of the work done on material coatings is done by chemists using hit and miss experiments. There are plenty of other areas where fundamental research could help, such as modelling friction and wear in an engine, to understand what could be added to lubricants to make the engine last for much longer.

Gumbsch wants the results of his work to be used in the real world, solving real problems, rather than just being of academic interest. He has found that the world of engineering and materials is quite capable of throwing up challenges of fundamental science that can keep a theoretician going for many years to come.

He says: ‘This is very important to me, I find it inspiring. I like to pick my research problems from the real world. I still consider myself more of a scientist than an engineer, but I need the engineering environment in order to find subjects for research. I always want to take it a few steps beyond the engineering solution and find out what is really going on, so I still think of myself as a scientist.’

John Murphy

Peter Gumbsch CV

Education

 

1991 Dr rer. nat. (with distinction) from the University of Stuttgart, Germany

1988-1991 PhD Max-Planck-Institut für Metallforschung, Stuttgart, Germany

1981-1988 Physics (Diplomstudiengang) at the University of Stuttgart, Germany

 

Employment

 

1989 Visiting Scientist at Sandia National Laboratories, Livermore (six months)

1991-1992 Research Associate at Imperial College for Science, Technology and Medicine, Department of Mathematics, London, UK

1992 Visiting Scientist at the University of Oxford, UK (four months)

1993-1996 Research Associate at the Max-Planck-Institut für Metallforschung, Stuttgart, Germany

1997-2001 Head of Research Group, Max-Planck-Institut für Metallforschung, Stuttgart, Germany

2001- present Full professor in mechanics of materials and head of the Institute for Reliability of Systems and Devices (IZBS) at the University of Karlsruhe (TH)

2001- present Head of the Fraunhofer institute for mechanics of materials (IWM) with locations in Freiburg and Halle/Saale, Germany

2006 Visiting Professor at School of Engineering, University of California, Santa Barbara (five months)