Was there any particular moment or concrete problem that guided your scientific path? How did your interest in Mechanical Engineering in particular arise?
I do not have a particular passion for mechanical engineering. In fact, I think I could have pursued other areas such as journalism, medicine or architecture. What I truly enjoy is the academic environment. I realised this when I had a few interviews in industry, where I felt that the objectives and the way of working were not aligned with my interests. My first contact with academia was through a European project related to fatigue in aeronautical structures during my undergraduate degree, and I very much enjoyed the experience. At university I can carry out research, teach, write papers, travel and present my work at international conferences, meet people from different cultures and perspectives, have a very flexible schedule and work at my own pace, and so on. That is what I really enjoy. At the beginning of my career, books by David Lodge and Philip Roth, which revolve heavily around the academic world, also helped me define my path.
Joining processes are often invisible to a layperson, yet absolutely crucial for the safety and durability of structures. What makes this area particularly challenging from a scientific point of view?
Joining structures and materials means joining different things. Otherwise, they would not need to be joined. And joining different things is always a challenge. Even with ourselves, it requires reconciling differences. Differences in shape, behaviour, composition, and so on. The other challenge is that it involves very different areas of knowledge. For instance, in adhesive bonding, one must master chemistry, as the adhesive is a chemical product; physics, in terms of surface properties; fluids, since the adhesive is applied in liquid form; manufacturing, in terms of how to fill the bonding area; solids, to understand how the joint will behave once the adhesive has cured; and so on. It is a multidisciplinary field. But that is also a positive aspect, as it allows interaction with people from different fields using very diverse scientific methods. This helps one become a more well-rounded researcher.
The transition to more sustainable mobility requires multi-material solutions and new joining methods. Which advances do you consider most promising in this area, and what major challenges remain unresolved?
More sustainable mobility brings to mind electric vehicles. And when one thinks about electric vehicles, one thinks about range. Achieving a long range requires efficient batteries and a lightweight vehicle. This can be achieved by using lighter materials such as aluminium or fibre-reinforced plastics, in combination with steel, which remains widely used. This creates additional challenges in terms of dissimilar materials. Welding, which is commonly used to join steel, is more difficult to implement with aluminium, as aluminium loses many of its properties when melted. Therefore, processes that do not involve melting the material are widely used, such as solid-state welding, plastic deformation processes like riveting, and adhesive bonding. It is estimated that within five years, adhesive bonding will be the main joining method in a car. However, the technology still requires further development in terms of durability, particularly under cyclic mechanical loads, temperature variations and humidity.
After decades studying how structures are joined, and sometimes fail, which principles or ideas do you consider fundamental to testing and ensuring reliability?
Designing a structure to withstand a load is one thing. Designing it to withstand many loads over many years is another. That is the issue of durability. It is very difficult to predict how a joint will behave in the long term. Ideally, tests should be carried out on the structure under real conditions. For example, testing a car for ten or fifteen years and observing what happens. But clearly that is not feasible. Instead, accelerated tests are carried out under very severe conditions, and the results are extrapolated to milder conditions typically encountered in practice. On the other hand, it is also important to have a backup plan in case the structure fails. The aeronautical industry makes extensive use of this approach, known as fail safe design. For example, having a double joint using both adhesive and a rivet. If the adhesive fails, the rivet remains to carry the load and prevent catastrophic failure. It is also important to have some form of monitoring that allows the detection of problems, such as cracks, without destroying the joint. One of the most widely used methods is ultrasound. Waves are sent into the structure and are reflected differently when there is a discontinuity.
The detection of weak adhesion is one of the most complex challenges in this field. What makes this problem so difficult to solve, and which approaches do you consider most promising?
I ended the previous answer by mentioning non-destructive methods such as ultrasound; indeed, detecting weak adhesion in adhesive joints is almost impossible precisely because there is continuity between the adhesive and the substrate. It is one of the major research topics in adhesive bonding. Interestingly, the aeronautical industry was a pioneer in the use of adhesive joints because they allow greater strength through better use of the bonded area. However, that same industry remains very sceptical about their use because of concerns regarding weak adhesion that cannot be detected. For that reason, adhesives are never used on their own in a primary aircraft joint, as if it fails, the aircraft falls. They are always combined with an additional joining method, meaning that if the adhesive fails, there is always an alternative load path. Some promising methods exist, such as laser inspection, but the device can damage the joint if not properly controlled. In our group, we have developed a promising method based on high-frequency vibration and data processing using machine learning, although it still needs to be validated at full scale.
What role do you foresee machine learning playing in advanced engineering modelling in the coming years?
Machine learning and AI in general are indeed tools that may completely change the philosophy of joint design. They may be what is commonly called a game changer. Traditionally, we rely on physical concepts to understand and predict the behaviour of structures. We are not always able to solve the equations generated by the models we devise directly, and therefore resort to approximate methods such as finite elements, but the entire process is grounded in fundamental physical concepts. Machine learning changes this paradigm entirely. It is possible to arrive at a prediction, a result, bypassing all theoretical foundations. It is a kind of shortcut that allows the same problem to be solved much more quickly. Provided there is a large amount of data to train the algorithm, it can then handle any situation. For a highly competitive industry such as the automotive industry, this can make all the difference, for example by reducing the number of joints and optimising their placement.
You have collaborated with major industrial companies and led dozens of international projects. In your view, where does the relationship between academia and industry work best, and where do frictions still exist?
Our group began exclusively as a research group over twenty years ago. We were not particularly motivated to work with industry. The situation gradually changed, mainly because companies began contacting us due to our publications. They realised that the group possessed knowledge that could be transferred and used to solve concrete problems. At present, our activity is divided between fundamental research and consultancy for companies. It is very rewarding to know that the knowledge we generate is useful to industry and to society as a whole. We have had very positive experiences. Fundamental research is generally funded by public bodies that scrutinise our work in detail, and rightly so, as we are using public funds. One might expect companies to be even more demanding, as the economic aspect is more pressing, but paradoxically we generally have more freedom with companies. There are some exceptions, but overall they are very understanding and we are able to achieve the objectives easily. To date, we have never experienced moments of friction with companies.
Beyond research, you have consistently invested in the pedagogical dimension of your work, from open access teaching materials to being a strong voice on inclusion and diversity. What changes and developments would you like to see in engineering education?
I would like to see more passion in teaching. With passion comes everything else. Passing on knowledge is the most noble thing there is. I feel incredibly privileged to be in this position. I also feel a great sense of responsibility. That is why I want to give everything I have so that students learn as much as possible, and with enjoyment. I want teaching to be a positive experience, not a chore. It is a Herculean challenge given the level of distraction in our society, but we must at least be prepared to try. To try new approaches, new teaching tools, to make students feel that we genuinely want what is best for them, to convey passion. That already goes a long way. And I would also like to see more diversity. Our society includes people of all kinds, yet at university I see the same type of people. We must create more mechanisms to achieve a more universal and inclusive university.
|
