Page 290 - Engineered Interfaces in Fiber Reinforced Composites
P. 290

Chapter 6. Interface  mechanics and,fracture toughness theories   27 I

                configuration.  In  the  case  of  interlaminar/intralaminar  fracture where  the  crack
                propagates parallel to the fiber direction, the fibers are peeled off or fractured rather
                than being pulled out. The fiber bridging in interlaminar fracture arises mainly from
                the misalignment of fibers across the main crack plane,  localized concentration  of
                fibers and matrix material (i.e. fiber rich and resin rich regions) and the growth of
                the crack on more than one plane.
                  Based on the 'inherent  flaw model'  (IFM) proposed  earlier by Waddoups et al.
                (1 97 I),  Caprino et al. (1 979, 1980) predicted eo to be approximately 2 mm for double
                edge  notch  (DEN)  and  CN  tension  specimens  of  continuous  epoxy  matrix
                composites  containing  carbon,  glass  and  boron  fibers.  The  IFM  was  originally
                developed  to  predict  the  notch  strength  of  laminates  with  finite  width  and  an
                inherent flaw of length, 2co. eo is found to be sensitive to the variations in ply layup
                sequence, fiber orientation and type of fibers used. Slightly higher values of eo  have
                been obtained for CN and SEN tension specimens of carbon-poxy  system based on
                the 'damage zone model (DZM)': eo  = 2-4 mm (Hillerborg et al., 1976; Aronsson
                and Bachlund, 1986). In the DZM, the closure stresses due to the bridging fibers act
                on the entire crack surfaces in accordance with a particular closure stress-crack  face
                relationship  which  is  consistent  with  the  damage  mechanisms  taking  place.  The
                predictions of eo  and the notched strength, on, using the DZM are shown (Aronsson
                and Backlund,  1986) to agree better with experimental measurements than the IFM
                model. For example, the experimental value of co for three-point bend specimens of
                short glass-polyester  systems is about 6 mm which is almost the same as the DZM
                prediction  of 6-7  mm, while the IFM model predicts a lower value of 4.1  mm. A
                summary  of  all  these  previous  models  including  the  well-known Mar-Lin  model
                (Mar and Lin 1979) and a recent effective crack growth model to predict the residual
                strength  of  composite  laminates  with  notches  and  other  forms  of  discontinuities
                have been given by Afaghi-Khatibi et al. (1996a,b).
                  Because cracks in multi-angle ply laminates seek the easiest paths to propagate
                preferentially along the fiber-matrix  and laminar interfaces, the shape and size of
                the damage zone in these composites depend strongly on the loading configuration
                relative to the fiber orientation within the individual plies. The damage zone sizes are
                compared between the laminates of different layup configurations for varying notch
                lengths in Fig. 6.26. It is noted that the [90°] sub-cracks in [0°/900], carbon-epoxy
                laminates are approximately 10 times longer than  [45"] sub-cracks in [O0/  f 45"/0"],
                laminates, while  [O"]  sub-cracks for the former laminates are about twice those for
                the latter laminates. The complex appearance of the damage zone together with the
                interactions between laminae make the quantitative characterization of the damage
                zone size in multi-angle laminates extremely difficult.
                  In  short  fiber  composites,  energy  absorption  mechanisms,  such  as  interfacial
                debonding and  matrix cracking, most  often  occur  at the fiber ends  (Curtis et  al.,
                1978). The damage model proposed by  Bader et al. (1979) assumes that short fiber
                composites  fail  over  a  critical  cross-section  which  has  been  weakened  by  the
                accumulation of cracks, since the short fibers bridging this critical zone are unable to
                carry the load. In fatigue loading, sudden fracture takes place as a direct result from
                the far-field effect of the composite, rather than due to the near field of the crack tip
   285   286   287   288   289   290   291   292   293   294   295