Workshop MaterialDigital2019 on May 14th and 15th, 2019 in Freiburg - more info
Through advances in networking, integration and new services based on information streams as well as the processing of big quantities of data, digitalization promises a competitive edge, and not only for manufacturing companies. This would not be possible without a digital representation of products and processes, which is directly linked to integrated material simulation, the de facto tool for the quantitative description of relations between process steps, material microstructures, material properties and component behavior. Within this tool, material modelling, process and component simulation, material characterization and process optimization are linked in an integrative fashion.
This digital representation of products, a necessity of digitalization, is part of the research conducted at the Fraunhofer Institute for Mechanics of Materials IWM.
Through shifting paradigms in the development, production and utilization of materials, new value chains can be created. The key to this is the consistent digital representation of materials and their behavior along their value chain or life cycle. Digital representations enhances safety, reliability, functionality and flexibility during production by permeating processes of development and production with materials intelligence and by enabling developers and engineers to comprehend and use materials in their respective development and production steps as variable systems with adjustable characteristics.
The condition for using digital representation effectively across scales and disciplines is for it to become an integral part of the feature or component in question, allowing for the processing of data from diverse origins without it being associated to one developer, one team or one corporate function. This requires a multidisciplinary linking of methods and concepts from material sciences, process technology and informatics.
The consistency in using material data streams in value chains allows target properties and functions of materials to be translated more effectively into the necessary material microstructure and the material itself to be more economically produced. By digitally connecting production steps and their respective effects to a component’s performance while in use, component functionality, including its operational lifespan, can be enhanced.
The digitalization of materials produces added value by allowing not only the material itself, but also its representation (e.g. material condition, properties and models) to become a part of business models.
Automated experimental material data creation
Automated 3D microstructure analysis
Integration of material data in digital twins via virtual testing
Creation of material data spaces for the hierarchical depiction of real and virtual materials, samples and components along interconnected material histories
Development of material models
Deep learning and statistics
Deducing process/structure/property relationships
A digital twin is the digital image or snapshot of actual materials, test samples and components, which includes its entire historical background (material history). In order to be able to manage a digital twin and its material history, creating a material data space is necessary.
Constant availability of material property information, which may change from the production stage to the end of a component’s lifecycle, creates completely new design possibilities regarding the technological performance and energy and resource efficiency of components and systems. This allows for the implementation of untapped potential of materials to further enhance their reliability, functionality and productivity.
Creating digital sets of data and digital twins for materials
Digital imaging and analysis of material histories
Design and development of material data spaces
Hierarchical descriptions of materials via data determined in experiments and simulations
Integration of material data in a digital twin via virtual testing (Virtual Lab)
Automated material data creation (3D microstructure analysis, tribofarming)
Top-down/bottom-up informed modelling of material properties, ranging from the atomistic to the component scale (multiscale information exchange)
Data analysis for the development of process/structure/property relationships (deep learning and statistics)