Consolidation Experiments for Laminate Composites

Consolidation experiments conducted on special prepregs made of constant viscosity oils and aligned graphite fibers are reported. Measurements of the deformation behavior of the fibers in the "drained" state (oil impregnated, but zero pressure in the oil) reveal that, to a first approximat...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Journal of composite materials 1987-07, Vol.21 (7), p.650-669
Hauptverfasser: Gutowski, T.G., Cai, Z., Bauer, S., Boucher, D., Kingery, J., Wineman, S.
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Consolidation experiments conducted on special prepregs made of constant viscosity oils and aligned graphite fibers are reported. Measurements of the deformation behavior of the fibers in the "drained" state (oil impregnated, but zero pressure in the oil) reveal that, to a first approximation, the fiber network can be modelled as a nonlinear elastic network, and a model based on bending beam behavior is shown to accurately fit the data. Further more, a finite load (> 3 psi) is required to compress the fibers to volume fractions equal to, and above the range of about 0.56 to 0.61. Sample to sample variation, and in a few cases inelastic effects are also noted. Measurements of the permeability of the fiber bundle indicate a strong anisotropy as well as a strong fiber volume fraction dependence. It is shown that the Carman-Kozeny equation gives a good approximation to the axial direction permeability with kxx = 0.7. Compared to this, the transverse permeability is smaller by a factor of 25 or more. Furthermore, the transverse permeability depends upon fiber mo bility. Experiments comparing the measured and predicted values of the "resin" pressure during simulated compression molding show reasonable agreement. Similar experiments for simulated bleeder ply molding show that the theory tends to overpredict the rate of decay of the "resin" pressure if the Carman-Kozeny equation is used. This difference is probably due to a slight fiber rearrangement which unloads the fibers and causes a "log jam" effect. The incorporation of a modified Carman-Kozeny equation which allows Szz → 0 for Vf < 1 yields an excellent fit to the data. The implications of the results are discussed.
ISSN:0021-9983
1530-793X
DOI:10.1177/002199838702100705