Efficient fatigue life estimation of structural assemblies based on cyclic test data of individual parts

Description

... Secondly, a statistically appropriate sampling of the design space of different adapter end and tube geometries is used to obtain simulation fatigue data of the various components. These data represent the sampling points for the stochastic meta-models resulting in an analytical description of fatigue data as a function of adapter end and tube geometries. Therefore, these meta-models are able to represent the whole design space of tie-rods within a specific parameter range. Thirdly, the analysis method is finalized by cross-checking the results of the meta-models with test data. Our method is also capable of calculating fatigue life data of assembled tie-rods based on the developed meta-models of single parts. Within this third stage, the calculated fatigue life of assemblies is again compared to the corresponding fatigue test data in order to ensure robustness of our method and further optimization of the meta models and their interaction where required. In general, this method provides a continuous optimization as well as further development of meta-models and finite element models by updating and adding test data. Hence, the design space can be extended to cover a larger range of tie-rod geometries. Consequently, fatigue life predictions of tie-rods with new dimensions, not covered by test data, can be performed using the verified meta-models only. The proposed powerful characterization of all meaningful tie-rod configurations enables a very fast, accurate and test data driven fatigue life estimation. Moreover, due to its analytical conception, our method is also applicable within an efficient optimization process, helping to identify the strongest but also lightest tie-rod design for a given load environment.

Period	29 May 2018
Event title	12th International Fatigue Congress
Event type	Conference
Location	FranceShow on map