1) At large scale, c can be reasonably small particularly for relatively low yield strength materials such as single crystals or coarse-grain alloys. This can provide large toughness at intermediate temperature in even those materials prone to brittle fracture. At low enough temperatures low energy brittle fracture still prevails in such materials. 2) This does not preclude a microscopic fracture mode transition as temperature increases and yield strength decreases. All this implies is that with greater shielding and lower yield strength, σysc1/2 in the denominator of Eqn. (2) decreases with increasing test temperature. This raises Kc exponentially. By elimination of decohesion along either a cleavage plane or an intergranular or bi-material boundary because the necessary Kc is too high (!), the alternative fracture mode is ductile fracture. The same concept applies to hydrogen embrittlement but with additional complexity. Consider stainless steels or superalloys. Does hydrogen reduce kIG, increase σys but leave c unaffected? Clearly this does not happen in many stainless steels that fail by microvoid coalescence but perhaps it does in high strength superalloys that fail by intergranular fracture. An intriguing aspect of this is that slip localization would tend to enhance shielding by reducing c. Even if kIG were reduced this could lead to microvoid coalescence as the preferred fracture mode. 0.1 1 10 100 1000 0.01 0.1 1 10 W/Cu/Ti without hydrogen Cu/Ti without hydrogen Cu/Ti with hydrogen 0.1 1 10 100 1000 G ( J/m^2) Film thickness (µm) GIC (J/m 2) Figure 5. Effect of film thickness on Cu/Ti/SiO2 interfacial fracture resistance with (dashed line) and without hydrogen(solid line). 3) Finally, with respect to small scale, we hypothesize that the values of c can be considerably larger at the same yield strength in thin films compared to larger scale microstructures. The limited shielding available due to restricted lengths of slip bands and numbers of dislocations will tend to give larger values of c and form lower toughness and higher susceptibility to hydrogen embrittlement. This is particularly the case for thin nanocrystalline films with high yield strengths (fine grain size) weakly bonded to a substrate. For such cases it is seen that all three parameters in the exponential, a lower kIG in the numerator and high yield and a larger stand-off distance in the denominator, would favor embrittlement.
RkJQdWJsaXNoZXIy MjM0NDE=