Page 156 - 3D Fibre Reinforced Polymer Composites

Braided Composite Materials 145
6.4 FATIGUE PERFORMANCE

In their comprehensive investigation of the mechanical properties of two, 4-step braided
composites, Gause et al. (1987) measured the fatigue performance of the 3D braided
materials in tension-tension (T-T), tension-compression (T-C) and compression-
compression (C-C) loading and compared it to a baseline 24 ply tape laminate. The data
was highly scattered but at tests running to a million cycles it was clear that the baseline
laminates had significantly better fatigue performance than the 3D braids. The
maximum (averaged) fatigue stress, as a percentage of their ultimate static strength, that
was carried successfully to one million cycles by the tape laminate specimens was
found to be 73% (T-T), 50% (T-C) and 78% (C-C). This is compared to 57% (T-T),
37% (T-C) and 43% (C-C) for the 1x1 braid, and 56% (T-T), 37% (T-C) and 52% (C-C)
for the Ixlx%F braid architecture. The improved fatigue performance of the 2D
laminates over the 3D braids was attributed to the fibre waviness that is intrinsic to the
braided architecture. This waviness allows the fibres to bend in addition to deforming
axially under load, thus working the matrix more severely. In T-T and T-C fatigue
conditions both braided architectures behaved identically. In C-C conditions the authors
stated that the lxlx'/zF braid architecture showed greater life capability then the 1x1
architecture, which they credited to the presence of the fixed 0" yarns providing greater
resistance to catastrophic fatigue damage. However, the scatter in results that is evident
from the published data makes it unrealistic to draw this conclusion.
Similar fatigue results were seen by Gethers et al. (1994) in their tension-tension
testing of 4-step braided carbodepoxy materials. Although the behaviour of the .3D
braids was not compared to 2D laminates, the average maximum fatigue stress at one
million cycles was approximately 55% of the 3D braids static tensile strength, very
similar to that recorded by Gause et al. (1987). Those specimens that survived one
million cycles of testing were tested to failure statically and found to have a residual
tensile strength that was 80% of the original tensile strength.

6.5 MODELLING OF BRAIDED COMPOSITES

There have been a number of models developed to predict the mechanical properties of
3D braided composites and, in a similar fashion to the other 3D textile composites
described in this book, these models first depend upon an accurate description of the 3D
braided yarn to be made. This description is accomplished through a geometric
modelling of the yarn topology that is based purely upon the braiding procedure itself.
Each particular braiding process has specific, characteristic equations that govern the
topology of the yarn structure within the preform. These characteristic equations are
explained in greater detail for 4-step braiding by Wang et al. (1994) and for 2-step
braiding by Byun et al. (1991b).
Once the geometric model of the 3D braid has been established the process of
modelling its mechanical properties is carried out in a similar fashion to other 3D textile
composites. A Representative Volume Element (RVE) of the braid is identified and the
properties of this RVE are established through application of analysis techniques such
as classical lamination theory (Byun et al., 1991b) or an elastic strain energy approach
(Ma et al., 1986). The classical lamination theory was also used by Yang et al. (1986) in
the development of their Fibre Inclination Model. The properties of the overall