Uncertainties in the prediction of CFRP laminate properties in the context of a reliability based design approach

Link:
Autor/in:
Verlag/Körperschaft:
Hamburg University of Technology
Erscheinungsjahr:
2013
Medientyp:
Text
Schlagworte:
  • Design philosophy
  • Mechanical properties
  • Reliability
  • Tension testing
  • 620: Ingenieurwissenschaften
Beschreibung:
  • © QinetiQ Ltd 2013. In order to pursue a reliabililty based deseiyg approach, material related uncertainties need to be quantified. For composite structural designers, the evaluation starts at the coupon level where a limited number of layups gets tested. The state of the art stiffness determination process then neglects the non-linear stiffness behavior as well as the spatial variability and aims at conservative estimates. In the present work, a unidirectional laminate is tested in tension and the stiffness is evaluated at different loading points. Monte Carlo simulation technique is used to predict laminate stiffnesses based on this data for two different layups and compared to a similar simulation based on constituent properties. The predictive qualities are checked against measured laminate stiffnesses. Furthermore, polynomial curve fits are generated for each specimen and mean and confidence intervals evaluated for a strain working range of interest and compared to the secant method suggested by the respective Norm. A simple beam problem is used to illustrate the consequences of the different stiffness determination methods. In the considered tension case the maximum computed deviation of a specific stress quantile to the reference stress quantile is around 5%. The evaluation methods investigated lead to significantly higher deviations when comparing the respective deflections, thus suggesting that ultimately a more robust testing and evaluation method needs to be applied. This is especially true since in case of stiffness determined designs the usage of a low estimate of the stiffness is not necessarily a conservative assumption.
Quellsystem:
TUHH Open Research

Interne Metadaten
Quelldatensatz
oai:tore.tuhh.de:11420/7128