Project Details
Projekt
Solution Spaces for Optimization of Structures Subjected to Impact Loads Considering Polymorphic Uncertainties
Applicant
Professor Dr.-Ing. Fabian Duddeck
Subject Area
Mechanics
Lightweight Construction, Textile Technology
Lightweight Construction, Textile Technology
Term
from 2016 to 2020
Project identifier
Deutsche Forschungsgemeinschaft (DFG) - Project number 310971106
Solutions spaces are a new deterministic approach (proposed in 2012/13) to enable design of very complex structures in early development stages via simplified mechanical modeling, component decoupling, optimization of functional requirements, component optimization, and final system validation. Although currently matured enough to find first industrial applications, it lacks a method for concise consideration of related uncertainties. Beside classical probabilistic and possibilistic uncertainty modeling, lack-of knowledge and modeling uncertainties in a broader sense have to be addressed additionally. The method proposed hence here for non-deterministic solution spaces, focusing on applications related to structural impact, consists of the following modules:1. New simplified crash models considering probabilistic/possibilistic uncertainties.2. Component decoupling with non-deterministic solution spaces considering lack-of-knowledge uncertainties (e.g. unknown future developments of the system, changes in requirements, changes in other components) and modeling uncertainties (incompleteness, impreciseness, discretization and simplification etc.).3. Integration of expert knowledge into the non-deterministic solution spaces approach.4. Optimization algorithms for a new robustness approach based on flexibility requirements to handle lack-of-knowledge.5. Adaptive reliability methods to respect differences in component requirements and related lack-of-knowledge including maturity degree of development.6. Probabilistic and possibilistic optimization (topology, shape, size, and material) of components into the non-deterministic corridors derived from the solutions spaces. Hence, an overall approach is proposed to allow efficient development in early design stages with their inherent polymorphic uncertainties.
DFG Programme
Priority Programmes
