DeMeSSFiD – Development of Methods for the time-efficient Simulation of Shrink Fits with Deviations

• Applicant: Professor Dr.-Ing. Sandro Wartzack
• Subject Area: Engineering Design, Machine Elements, Product Development
• Term: since 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 559100751

Press fits are used in many applications in mechanical engineering such as shaft-hub connections and assembly of rolling and plain bearings. Due to ever-increasing quality and efficiency requirements as well as rising cost pressure, computer-aided simulative validation of product properties is essential in order to be able to predict operating behavior even before the start of production and thus reduce expenditures for experimental validation, production and later com-plaints. In the case of press fits, in addition to parameters such as wall thicknesses and elastic moduli, the fit surface geometries in particular have a significant influence on the resulting de-formation and stress state and thus also on the function of the press fit. The current calculation methods for press fits are based on geometrically ideal joining surfaces, which consequently requires narrow tolerances. Consideration of deviations is currently only possible by employing numerically complex finite element analyses. However, simulation methods with short runtimes are required to be able to perform statistical tolerance analyses during development of complex technical systems such as transmissions. The main objective of the first phase of this proposition is therefore to provide time-efficient methods for the simulation of shrink fits taking geo-metric deviations into account. To this end, a reference model based on the finite element method will be built to enable investigation of the effects of geometric deviations of different wave-lengths as well as other relevant parameters on the deformation behavior and functional proper-ties of shrink fits. The created model will subsequentially be experimentally validated. Using a database created through systematic evaluation of the reference model, methods for accelerating the simulation are then developed. For this purpose, possible simplifications of the reference model as well as metamodeling approaches for data-driven prediction of system behavior will be investigated. Finally, the efficient simulation methods developed in the course of the project will be transferred into a CAE framework that enables product developers to ensure fulfilment of individual functional requirements to shrink fits while taking into account the effects of manufacturing-related geometric deviations.

• DFG Programme: Research Grants