Researchers at Universidad Carlos III de Madrid (UC3M) have developed a set of innovative methods and algorithms that improve the performance and precision of vehicle design through topological optimisation, a mathematical technique that allows designs to be optimised by distributing materials efficiently. The results of the research, which the scientists have applied to the design of parts for a competition motorcycle to reduce its weight while maintaining performance, could have a major impact on sectors such as the automotive and aeronautics industries.
Thanks to a filter that works directly with the density field using automatic algorithms, the researchers have managed to manufacture parts with optimal material distribution based on the constraints imposed, whether these were weight, volume, stress or the amount of heat that the part had to transmit or withstand.
Advertisement
‘The advantage of using topological optimisation for part design is that you automatically obtain the optimal shape, instead of having to apply different iterations and calculations,’ said Abraham Vadillo Morillas, a doctoral student in the Department of Mechanical Engineering at UC3M. ‘In addition, this process has other advantages, such as direct and indirect cost savings. For example, an airplane manufactured with a lighter part than the original will be more environmentally friendly and will make travel cheaper for the passenger.’
Although the experimental evaluation of these advances has been focused on the manufacture of motorcycle parts, the applications of these methods are much broader, as many sectors can benefit from the reduction in weight of their components, as well as the improvement in their rigidity.
Advertisement
‘Any part subjected to loads or vibrations can benefit from these advances: from vehicle components to industrial machinery elements, architectural structures, jewellery and even fashion,’ explained Cristina Castejón Sisamón, a professor in the Department of Mechanical Engineering and head of the MAQLAB research group.
The methodology used to achieve these advances consisted of adapting topological optimisation to the main manufacturing processes and improving the interpretability and accuracy of the results using new algorithms. In the case of additive manufacturing, the researchers have developed specific methods for 3D printing that have already been used to manufacture motorcycle components with significantly reduced weight.
Advertisement
Recently, the team has begun to incorporate artificial intelligence techniques in order to move toward even more autonomous optimisation processes. ‘Our goal is to develop agents that can adjust optimisation parameters themselves, in an approach we have called AVM (adaptive variable modelling). In this way, we hope that the new AI-based tools will change the paradigm of mechanical design and enable the generation of optimal parts more quickly, accurately and efficiently,’ said Castejón.
The doctoral thesis produced based on this line of research was the first related to the MotoStudent competition project in which UC3M participates with its students and the MOTO-MAQLAB-UC3M association. This line of research began more than ten years ago with the aim of supporting students participating in the competition. The research has already produced prototypes that have participated in several editions of MotoStudent.
The research has been published in Structural and Multidisciplinary Optimization.