A SERVICE OF

logo

PREDICTION OF ELASTIC CONSTANTS AND THERMAL EXPANSION
5-31
Table 5.6 Constituent property input used in code comparisons
(Transverse isotropy assumed in the y-z plane )
Component E
1
E
2
ν
12
ν
23
G
12
α
1
α
2
(GPa) (GPa) (GPa) (10
-6
K
-1
) (10
-6
K
-1
)
Yarn 145 11.7 0.23 0.30 5.52 -0.32 14.0
Matrix 3.45 3.45 0.35 0.35 1.28 40.0 40.0
Table 5.7 Geometric parameters assumed for code comparisons
Yarn Spacing (mm) 1.41
Layer Thickness (mm) 0.26
Yarn Fiber Volume (%) 75
Total Fiber Volume (%) 64
The codes also give plate stiffnesses, computed assuming plane stress conditions
(or using traction-free surfaces) and including bending. Plate stiffnesses are provided in
the form of the A, B, and D matrices of classical laminate theory. Representative elements
are shown in Table 5.10. There is a surprising variation in these results for both the in-
plane and bending stiffnesses. For example, µTex-10 and µTex-20 give much lower
values for the in-plane stiffness than the other codes. The authors attribute this difference
to the way in which the B matrix terms (bending-stretching coupling) are handled. The
effective in-plane modulus of an asymmetric laminate can be much less than a simple
volume average would indicate, because the laminate curves under in-plane load. A
single ply of a plain weave is indeed asymmetric, the upper and lower halves being
orthogonal at any point. Most codes assume that the moment constraints of neighboring
unit cells prevent bending, but this is not assumed in µTex-10 or µTex-20. If many plies
were stacked into a thicker laminate, the discrepancy between the codes would
presumably diminish, since the interactions of adjacent layers would suppress the
development of curvature in any one layer.