A G. Evans et al. /Journal of the European Ceramic Society 28(2008)1405-1419 141 Al-Rid 3.5 Hf1(179) Plastic Zone HfA1(1.74) Wake 2.5 predation segregation SN(0.42 s1+SN(095) 15 (b Interfacial Coverage 3.75 0.36 -- Clean Interfacial Separation(nm) Fig. 14.(a)Trends in the work of separation calculated using density functional /Ro=O ments at a stoichiometric interface, with and without dopants and segregants. The numbers refer to the work of separation on the indicated plane (in J/m") thermo-chemically incompatible with the TGo+u and require a diffusion barrier to preclude deleterious interactions. A su able diffusion barrier is 7-YSZ, which also provides a tougher 0.3 Conventional plasticity underlayer to hinder the propagation of cracks originating near and above the TGO. Unfortunately, solutions for toughening 0.5 the zirconates are not immediately evident, and the challenge remains to be addressed by future research. 6. Research opportunities 002 0.04 006 From the foregoing synopsis the opportunities emerge for D/R research on alumina and zirconia (and related oxides) that might Fig. 13.(a) A schematic showing how traction/separation results(right)obtained further enhance in the performance of hot section components from first principles calculations become an embedded process for a crack exten- in turbines sion simulation that captures the multiplicative influence of plastic dissipation the intertace toughness. (b)The ratio of the toughness to the work of separa- (i) Despite decades of research the thickening rate of the a- tion as a function of the ratio of the dislocation-free zone width, D, to the plastic zone size on the metal side of the interface, Ro ETo/ay, with ay the yield D3(once formed)remains to be understood. There are strength and I the plasticity length scale. dramatic effects of impurities( detrimental) and dopants (beneficial). The grain size is also important because the diffusional fluxes occur primarily along the grain bound aries. The counter-fluxes of oxygen and Al have yet to be accounted forin the context of a viable atomistic model. This is of key importance since the latter are directly responsibleA.G. Evans et al. / Journal of the European Ceramic Society 28 (2008) 1405–1419 1415 Fig. 13. (a) A schematic showing how traction/separation results (right) obtained from first principles calculations become an embedded process for a crack exten￾sion simulation that captures the multiplicative influence of plastic dissipation on the interface toughness. (b) The ratio of the toughness to the work of separa￾tion as a function of the ratio of the dislocation-free zone width, D, to the plastic zone size on the metal side of the interface, R0 ∼ EΓ0/σ2 Y, with σY the yield strength and l the plasticity length scale. Fig. 14. (a) Trends in the work of separation calculated using density functional theory for both stoichiometric and Al-rich interfaces. (b) The atomic arrange￾ments at a stoichiometric interface, with and without dopants and segregants. The numbers refer to the work of separation on the indicated plane (in J/m2). thermo-chemically incompatible with the TGO4,90 and require a diffusion barrier to preclude deleterious interactions.90 A suit￾able diffusion barrier is 7-YSZ, which also provides a tougher underlayer to hinder the propagation of cracks originating near and above the TGO. Unfortunately, solutions for toughening the zirconates are not immediately evident, and the challenge remains to be addressed by future research. 6. Research opportunities From the foregoing synopsis the opportunities emerge for research on alumina and zirconia (and related oxides) that might further enhance in the performance of hot section components in turbines. (i) Despite decades of research the thickening rate of the - Al2O3 (once formed) remains to be understood. There are dramatic effects of impurities (detrimental) and dopants (beneficial). The grain size is also important because the diffusional fluxes occur primarily along the grain bound￾aries. The counter-fluxes of oxygen and Al have yet to be accounted for in the context of a viable atomistic model. This is of key importance since the latter are directly responsible