BeNeNe is an innovation project to develop an AI-based solution for optimised loading of autoclaves. In industrial practice, an autoclave is used to cure a wide variety of fibre composite components that differ in shape and size. Four partners - SHS plus GmbH, 3D ICOM GmbH & Co. KG, Institut für Polymer- und Produktionstechnologien e.V. and Faserinstitut Bremen e.V. - have worked together on an intelligent assistance system for a resource-saving and self-optimising autoclaving processes. The project is funded with a total volume of around €863,000 from Germany's Central Innovation Programme for SMEs (ZIM).
High-performance components made of fibre-reinforced plastic for the aerospace industry are predominantly manufactured using the so-called autoclave prepreg process. The polymer matrix is cured under pressure and at an elevated temperature in the autoclave. Since a large number of component variants occur in industrial practice, the arrangement in the autoclave must always be adapted based on empirical values. On the one hand, care should be taken to ensure optimum use of space with the moulds due to the high energy requirements. On the other hand, it is imperative that the quality of the components remains at the same level. Due to the large number of possible combinations when loading the autoclave, a prior, data-based optimisation for different loading scenarios using currently available calculation methods is not economically feasible.
Artificial intelligence optimises the loading of autoclaves
The aim of the "BeNeNe" project is to develop an intelligent assistance system for a more resource- and energy-efficient production of fibre composite components manufactured by autoclaving. The focus is on both the preparatory production steps and the production process itself in order to achieve a holistic and continuous optimisation of the process and the resources used.
The basis of the system is to consist of an artificial neural network into which simulation models of mould tools and autoclaves are implemented. The models provide the neural network with compressed empirical values for components, which, together with the "knowledge" of the neural network, can be used for the optimal calculation of the respective production process. This method reduces the time required for calculations enormously. Sensory test tools are to be developed for measuring different autoclaving processes, which will be individually evaluated with a measuring system.
The idea for the "BeNeNe" project was born within the framework of the ENVIPRO, or Environmental Friendly Production Innovation Network, funded by Germany's federal Central Innovation Programme for SMEs (ZIM). Project owners are actively supported in the realisation of R&D projects and in securing financing. ENVIPRO is managed by the IWS GmbH, which also takes over the application management of the cooperation projects and accompanies the members intensively in the development of new technologies.
via: 3D ICOM
