| Περίληψη: | The work conducted in this thesis is part of an on-going effort from AML/UoP to be a leader in composite aerostructures design not only in terms of static design and analysis, but also on energy absorption, strain rate dependent phenomena, impact and blast.
Although the area of quasistatic behaviour of composite structures and their joints has advanced significantly, issues pertaining to their behaviour under medium and high load rate behaviour have been significantly lagging.
The complexity of the phenomena involved, and their stochastic nature have deterred certification organisations from including elevated load rate phenomena in their procedures.
The work conducted here aims to practically cover this gap in composite joints modelling and investigation.
The foundation of this work is laid upon three main pillars:
• The first are the standardisation efforts by the AGATE consortium in what regards quasistatic material properties.
• The second is LS-DYNA which has incorporated a material model covering progressive damage modelling and rate dependent behaviour for orthotropic composites
• The third is experimental work conducted in the framework of the VULCAN project by TNO, in the area of shock loading of composite specimens.
These three pillars lead to the definition of the material parameters for a specific composite prepreg, both in terms of quasistatic and transient behaviour. In order to derive the rate dependent part of the material’s model behaviour a set of Prony Series coefficients is required to be defined (at minimum two). In order to achieve this, a parametric study expanded by Neural Networks has been conducted. This work is described in Chapter 2. Due to the fact that the contents of the standardisation database are common for all material systems, the proposed procedure can be applied for all AGATE, NCAMP and Composites Material Handbook 17 standardised materials, requiring only the shock experiments.
Another issue not allowing the direct explicit fastener modelling in larger than specimen models has been the computational costs associated with it. This issue is covered in Chapter 3, where an innovative bolt representation has been developed; the basis of this work involves the separation of the geometrical to the stiffness representation of the bolt. This development is also followed by the validation of the concept against the AGATE experiments conducted per the ASTM D-951 testing procedure, in order to ensure uniform contact characteristics even when the laminate sequence changes. Finally, an assessment of the combined bearing bypass behaviour of a typical skin laminate sequence is concluded.
|
|---|
