Dr.
Michael Krahl
Contact
Publications
No publications added for this person yet.
Research
C05 Generative design for actively controlled mould systems with high system complexity and functional density
Summary
The DediGrad subproject C05 investigates how generative AI and data-driven design methods can be integrated into the development process of complex mould systems for graded components. The aim is to produce automatic or mixed-initiative design suggestions for mould geometries and for the placement of mechanical, thermal, and fibre-control elements. To achieve this, CAD and simulation data are systematised, training data are generated, and suitable neural-network architectures for mould design are developed. In this way, the subproject establishes the basis for an adaptive, learnable mould-development process that becomes progressively more capable as knowledge within the CRC/TRR grows.
Generative Tool Design
- Establishment of an ANN-based design method for the automatic generation of complex mould systems
- Integration of the method into conventional mould-development workflows as human–model interaction
- Development of efficient data pipelines for model updates with new CRC/TRR technologies
- Application and validation using the mould of the EFS demonstrator
AI-Driven Design Processes
- How can existing knowledge together with CAD and simulation data be systematised, labelled, and prepared as input for generative AI?
- Which ANN architecture and training strategy are suitable for three-dimensional mould design?
- How can engineers interact meaningfully with AI-based design suggestions?
- How should uncertainty, imprecision, and newly emerging knowledge about changing CRC technologies be handled?
Related Project Members
Eugen Boos
Publications
No publications added for this person yet.
Daniel Haider
Contact
Publications
No publications added for this person yet.
Prof.
Steffen Ihlenfeldt
Position
Deputy Speaker Management Committee Subproject Leader
University
Fraunhofer IWU Dresden Technische Universität Dresden TUD Dresden University of Technology
Publications
M. Klaerner, L. Kroll, R. Svidler, R. Rinberg, K. Paetzold-Byhain, N. Modler, W.-G. Drossel, S. Ihlenfeldt (2026) . Potentials of natural fibre-reinforced composites for dynamic applications, Procedia CIRP Volume 138, 2026, Pages 803-808 https://doi.org/10.1016/j.procir.2026.01.138
Dr.
Lars Penter
Publications
No publications added for this person yet.
C01 Simultaneous optimisation of geometry and graded fibre orientation distributions
Optimisation of structural behaviour and fibre distribution in lightweight deign
C02 Integrative numerical design methods for component and process
AI-based simulation and optimisation of complex manufacturing processes
C03 Model reduction and formalisation for increased efficiency of integrative simulation chains
Rapid prediction of local material properties using data-driven models
C04 Methodical derivation of process-structure-property relations based on hybrid data spaces
Linking experimental and simulation data spaces for hybrid process models
C06 Collection and context-sensitive distribution of process data
Context-based utilisation of process data across the entire production chain
Explore publications and materials from the TRR 402
Learn more through scientific publications, posters, presentations, and open-access papers and gain insights into current research results and developments within the collaborative research center.
Get in Touch for Collaboration
Do you have questions about the TRR 402 or are you interested in collaboration? Feel free to contact us – we are happy to hear from you and will gladly respond to your inquiry.