Page 51 - Fiber Fracture
P. 51

36                                                    M. Elices and J. Llorca

               Table 1. Numerical and experimental results for tensile strength of copper (100)
               Method                      amax @Pa)  (100)   References
               Orowan-Polanyi (Eq. 1)      25                 Kelly and Macmillan (1986)
               Non-self-consistent KKR     55                 Esposito et al. (1980)
               Self-consistent ASW,        32                 Esposito et al. (1980)
               Non-empirical potential     36                 Esposito et al. (1980)
               Johnson et al. potential    41                 Johnson and Wilson (1972)
               Morse-function lattice model   7               Milstein and Farber (1980)
               Experimental values (whisker)   I .7           Brenner (1956)
               Experimental values (whisker)   1.7            Kobayashi and Hiki (1973)
               Flat crystal under shock wave   17             McQueen and Marsh (1962)



               (1981) are: 10.4 GPa in (1 lo), 12.6 GPa in ( 11 1) and 15.2 GPa in (100). Tensile strengths
               at 10.9, 22.3 and 15.3 GPa for A1203 whiskers grown in the (Owl), (1 120), and (1100)
               crystal directions were measured by Soltis (1965) and estimates of 98,87 and 87 GPa for
               the ideal uniaxial tensile strength parallel to (OOOl), (1 120) and (Ioio), respectively, were
               computed from the third-order elastic stiffnesses by Gieske (1 968). The Orowan-Polanyi
               equation gives a value of 46 GPa for the ideal (0001) tensile strength.
                  Metals  exhibit  the  maximum  stress  only  in  whisker  form  because  they  permit
               dislocation glide at low stresses, and whiskers are almost free of  dislocations. In  bcc
               metals, improved  potential  models  will  lead  to  a  better  understanding  of  the  ideal
               strength than has so far been gained from either the Orowan-Polanyi  approach or the
               use  of  the  Morse potential; predictions, for  example, of  the fracture stress for  a-Fe
               whiskers in the (1 1 1) direction using the Orowan-Polanyi  equation are 46 GPa, where
               the maximum tensile stress obtained by Brenner was  13.1 GPa (Brenner, 1956), at an
               elongation close to 0.05 (see paper by Kunzi, this volume).
                  For fcc metal fibres and whiskers, the Orowan-Polanyi  equation and first principle
               band theory approaches give estimates of the ideal uniaxial tensile stress up to an order
               of  magnitude above the  highest  stresses measured  (see, for  example, Table  1).  For
               ‘perfect’ Cu  fibres pulled  in  tension  in  the direction (IOO),  for example, predictions
               of  the theoretical tensile stress using the Orowan-Polanyi  equation bring 25 GPa. Ab
               initio calculations (Esposito et al.,  1980), based on non-self-consistent KKR using the
               muffin-tin approximation or using the self-consistent augmented spherical wave (ASW)
               method, provided values of  55 and 32 GPa, respectively. Other methods, based on  a
               non-empirical pair potential that can be expressed in terms of the cohesive energy and
               can be evaluated when the energy is a function of the nearest-neighbour distance, givc
               values of  36 GPa  (Esposito et  al.,  1980) and 41  GPa  (Johnson and Wilson,  1972).
               Other computations based upon a simple Morse-function lattice model give values of
               7 GPa (Milstein and Farber,  1980). Experimental results on whiskers show values of
               1.7  GPa (Brenner,  1956; Kobayashi and Hiki,  1973). Brenner noted that Cu  did not
               cleave but sheared apart, and Kobayashi and Hiki suspected that surface defects were
               present. McQueen and Marsh (1962) reported a tensile strength of  17 GPa, but Esposito
               et al. (1980) noticed that this value, based on shock-wave intensity, involves techniques
               whose validity is difficult to assess.
   46   47   48   49   50   51   52   53   54   55   56