Plasma-sprayed yttria-stabilized zirconia (YSZ) coatings are critical to enhancing the performance of thermal barrier coatings in gas turbines and aero-engines; however, their service life is significantly constrained by microstructural evolution and multi-mechanism coupling effects. Focusing on plasma spraying process routes (atmospheric plasma spraying, APS; suspension plasma spraying/solution precursor plasma spraying, SPS/SPPS; low-pressure plasma spraying, LPPS) and key process parameters as primary input variables, this review systematically analyzes their regulatory roles in microstructural characteristics such as porosity and crack density. Available studies indicate that distinct process routes give rise to pronounced structural differences: the porosity of APS coatings is 10%–20%, that of SPS/SPPS coatings is 15%–30%, and that of LPPS coatings is 1%–8%. After thermal exposure above 1100 °C, the porosity decreases to 6%–12%, 8%–18%, and 0.5%–3%, respectively, while the thermal conductivity increases to a maximum of approximately 2.5 W·m−1·K−1 and the Young’s modulus rises to 60–220 GPa. Further analysis reveals that mechanisms such as sintering densification, phase destabilization, thermally grown oxide (TGO) interfacial stress accumulation, and calcium–magnesium–alumino-silicate (CMAS) infiltration exert coupled amplification effects through microstructural evolution, thereby accelerating coating failure. On this basis, emerging regulation strategies are evaluated: the CMAS penetration depth of high-entropy oxides at 1300 °C for 5 h is only about 1/7 that of conventional YSZ, the thermal cycling life of self-healing coatings is enhanced by up to 4.2 times, and the crack density is reduced by approximately 35%. Finally, it is proposed that a quantitative prediction model integrating “structural parameters–evolution kinetics–service life” should be established, and that anti-sintering design, gradient structures, and functionalized systems be combined to enable the transition of YSZ coatings from empirical optimization to predictable design.
Ding et al. (Fri,) studied this question.