Skip to main content

A MICROMECHANICAL APPROACH TO INCREASED CONFIDENCE IN THE DESIGN OF LOAD BEARING COMPOSITE STRUCTURES

Objective

The objectives of this work is to develop micromechanical finite elemen t modelling techniques for composite material behaviour prediction. This micromechanical modelling will be developed for material behaviour at the fibre/matrix level and linked to laminate behaviour at the macromechanical level.
The composite used in this project was T300 carbon/BSL914 resin. A finite element (FE) based unit cell approach successfully predicted the elastic constants for the unidirectional composite (UDC). To investigate the mechanisms of failure under static loading, the UDC was tested in longitudinal shear, transverse tension and transverse compression. Angled ply lamina were also tested in tension to investigate shear behaviour. To establish interaction between plies, crossply laminates were investigated under static and cyclic loading. Creep behaviour was investigated on the pure resin, on the UDC loaded in transverse tension and on the 45 degree lamina.
Under longitudinal tensile loading the failure process was shown to be controlled by the progressive breakage of fibres. A statistical model was developed which could predict the tensile strength. Under longitudinal shear, transverse tension and transverse compression, failure was shown to be dominated by the behaviour of the matrix material. However, an interphase layer, formed between the fibre and the parent matrix during the curing process has a strong influence on behaviour (particularly in transverse tension).
A damage model was developed which could predict the transverse tensile strength of the UDC. In transverse compression and longitudinal shear, the observed strain to failure of the UDC was much greater than that in transverse tension.
Testing and analysis of crossply laminates showed that significant damage occurred within the inner layers of the laminate, particularly under cyclic loading, prior to failure. Load recovery models were developed to predict the observed behaviour.
Nonlinear FE analysis, using the unit cell model to investigate stress relaxation due to matrix creep, showed that as the composite cooled from cure to room temperature, the temperature at which residual stress buildup outweighs matrix creep is around 140 C. Thus, to calculate the residual stresses in the composite, relaxation due to creep effects in the matrix any be neglected if a stress free temperature of 140 C is assumed. This avoids the need for full nonlinear analysis.
THE PROJECT INCLUDES THE FOLLOWING INNOVATIVE ASPECTS:

- PREDICTIVE CAPABILITY FOR COMPOSITE MATERIAL BEHAVIOUR BASED ON CONSTITUENT FIBRE AND MATRIX PROPERTIES-BEHAVIOUR.

- EARLY IDENTIFICATION OF POTENTIAL PROBLEMS.

- UNDERSTANDING OF THE IMPORTANCE OF INTERNAL STRESSES, CRACK DENSITY, ETC...

- DEVELOPMENT OF A DESIGN METHODOLOGY

THE EXPECTED BENEFITS ARE:

- INCREASED CONFIDENCE IN THE USE OF COMPOSITE MATERIALS FOR CRITICAL APPLICATIONS

- REDUCED WEIGHT AND COST DUE TO LESS CONSERVATIVE DESIGNS

- REDUCED DEVELOPMENT COSTS AND SHORTER LEAD TIMES.

Funding Scheme

CSC - Cost-sharing contracts

Coordinator

Rolls Royce plc
Address

BS12 7QE Filton
United Kingdom

Participants (3)

RISO NATIONAL LABORATORY
Denmark
Address
Frederiksborgvej 399
University of Cambridge
United Kingdom
Address
Trumpington Street
CB2 1PZ Cambridge
École des Mines
France