For maximal lightweight potential and optimal manufacturability, part geometry, material and process must match in mutual regard and thus, part and process conceptualisation impact most on final part performance. Besides rigorous analysis of process dynamics, simulations enable an automated search of optima during part design.

We couple our elaborate and detailed simulation models with optimisation algorithms. Apart from definition of suitable optimisation objectives we also focus on computational efficiency: For instance, AI-Models can guide the optimiser’s search and concentrate costly simulations on the most promising variants. This reduces the computational effort by up to 70%.


Research focus
  • Laminate optimisation
  • Topology optimisation
  • Optimisation of local reinforcements (patches)
  • AI-assisted manufacturing process optimisation
  • Development of efficient process models


Research projects

Dr.-Ing. Clemens Zimmerling
Tel.: +49 721 608-45409
Email: clemens.zimmerling∂kit.edu


M.Sc. Louis Schreyer
Tel.: +49 721 608-45380
Email: louis.schreyer∂kit.edu


Example of AI-integration into optimisation: An AI-algorithm learns with a set of predefined set of process examples, which part feature requires which process configuration.

Selected publications in the research field

Forming process optimisation for variable geometries by machine learning – Convergence analysis and assessment
Zimmerling, C.; Kärger, L.
2023. Material Forming 26th International ESAFORM Conference on Material Forming (ESAFORM 2023) Krakau, Polen, 19.04.2023–21.04.2023, 1155–1166, Materials Research Forum LLC. doi:10.21741/9781644902479-126
Machine learning algorithms for efficient process optimisation of variable geometries at the example of fabric forming. PhD dissertation
Zimmerling, C.
2023, January 18. Karlsruher Institut für Technologie (KIT). doi:10.5445/IR/1000154623
Formability Assessment of Variable Geometries Using Machine Learning - Analysis of the Influence of the Database
Zimmerling, C.; Fengler, B.; Kärger, L.
2022. Key Engineering Materials, 926, 2247–2257. doi:10.4028/p-1o0007
Zeit- und kosteneffiziente Prozess und Produktentwicklung für den Hochleistungs-Faserverbundleichtbau unterstützt durch Techniken des Maschinellen Lernens
Zimmerling, C.; Kärger, L.; Carosella, S.; Middendorf, P.; Henning, F.
2019, May 20. 6. Technologietag Hybrider Leichtbau (2019), Leinfelden-Echterdingen, Germany, May 20–21, 2019
Optimisation of manufacturing process parameters for variable component geometries using reinforcement learning
Zimmerling, C.; Poppe, C.; Stein, O.; Kärger, L.
2022. Materials and Design, 214, Art.-Nr.: 110423. doi:10.1016/j.matdes.2022.110423
Deep neural networks as surrogate models for time-efficient manufacturing process optimisation
Zimmerling, C.; Schindler, P.; Seuffert, J.; Kärger, L.
2021. ESAFORM 2021 - 24th International Conference on Material Forming, ULiège Library. doi:10.25518/esaform21.3882
Rapid Determination of Suitable Reinforcement Type in Continuous-Fibre-Reinforced Composites For Multiple Load Cases
Zimmerling, C.; Fengler, B.; Wen, H.; Fan, Z.; Kärger, L.
2020, September 1. 23rd / 6th Joint Event: International Conference on Composite Structures - International Conference on Mechanics of Composites (ICCS / MECHCOMP 2020), Porto, Portugal, September 1–4, 2020
Virtual Product Development Using Simulation Methods and AI
Zimmerling, C.; Poppe, C.; Kärger, L.
2019. Lightweight Design worldwide, 12 (6), 12–19. doi:10.1007/s41777-019-0064-x
Manufacturing uncertainties and resulting robustness of optimized patch positions on continuous-discontinuous fiber reinforced polymer structures
Fengler, B.; Schäferling, M.; Schäfer, B.; Bretz, L.; Lanza, G.; Häfner, B.; Hrymak, A.; Kärger, L.
2019. Hospital physician, 213, 47–57. doi:10.1016/j.compstruct.2019.01.063
Application and Evaluation of Meta-Model Assisted Optimisation Strategies for Gripper-Assisted Fabric Draping in Composite Manufacturing
Zimmerling, C.; Pfrommer, J.; Liu, J.; Beyerer, J.; Henning, F.; Kärger, L.
2018. 18th European Conference on Composite Materials (ECCM 2018), Athen, GR, June 24-28, 2018
An approach for rapid prediction of textile draping results for variable composite component geometries using deep neural networks
Zimmerling, C.; Trippe, D.; Fengler, B.; Kärger, L.
2019. Proceedings of the 22nd International ESAFORM Conference on Material Forming ; Vitoria-Gasteiz, Spain, 8–10 May 2019. Ed.: L. Galdos, Art.-Nr.: 020007, American Institute of Physics (AIP). doi:10.1063/1.5112512
A meta-model based approach for rapid formability estimation of continuous fibre reinforced components
Zimmerling, C.; Dörr, D.; Henning, F.; Kärger, L.
2018. Proceedings of the 21st International ESAFORM Conference on Material Forming : ESAFORM 2018 : Palermo, Italy, 23-25 April 2018. Ed.: L. Fratini, Art.Nr. 020042, American Institute of Physics (AIP). doi:10.1063/1.5034843
Forming optimisation embedded in a CAE chain to assess and enhance the structural performance of composite components
Kärger, L.; Galkin, S.; Zimmerling, C.; Dörr, D.; Linden, J.; Oeckerath, A.; Wolf, K.
2018. Composite structures, 192, 143–152. doi:10.1016/j.compstruct.2018.02.041