The effects of Cr, Al and B addition on the microstructure and high-temperature oxidation behaviors (at 1200, 1250 and 1300 ​°C) of Nb–Ti–Si based alloys were investigated. The results showed that the addition of Cr stabilized α-Nb₅Si₃, while Al promoted the formation of β-Nb₅Si₃ and adding B promoted the formation of γ-Nb₅Si₃. Among the three elements, Al and Cr were beneficial to oxidation resistance at 1200 ​°C, and B was favorable to the oxidation resistance at 1300 ​°C. At 1250 ​°C, Al and B had the same effects on the improvement of oxidation resistance. The ratio of these alloying elements might play an important role in enhancing oxidation resistance. The oxidation resistance of the three kinds of silicides was compared, and the sequence was: γ-Nb₅Si₃> β-Nb₅Si₃> α-Nb₅Si₃. To predict the effects of the investigated alloying elements on the oxidation resistance of Nb–Ti–Si based alloys in a wider range of concentration, an artificial neural network (ANN) model was established, showing excellent accuracy and generalization ability. With the instructions of the ANN model, the oxidation resistance can be optimized with less additions of alloying elements.

研究了 Cr、Al 和 B 添加量对 Nb-Ti-Si 基合金组织和高温氧化行为（1200、1250 和 1300 ℃）的影响。结果表明，Cr的加入稳定了α-Nb ₅ Si ₃，Al促进了β-Nb ₅ Si ₃的形成，B的加入促进了γ-Nb ₅ Si ₃的形成. 这三种元素中，Al和Cr有利于1200℃的抗氧化性，B有利于1300℃的抗氧化性。在1250℃时，Al和B对提高抗氧化性的作用相同。这些合金元素的比例可能在提高抗氧化性方面发挥重要作用。比较了三种硅化物的抗氧化性，顺序为：γ-Nb ₅ Si ₃ > β-Nb ₅ Si ₃ > α-Nb ₅ Si ₃. 为了预测所研究的合金元素在更广泛的浓度范围内对 Nb-Ti-Si 基合金抗氧化性的影响，建立了人工神经网络 (ANN) 模型，该模型显示出优异的准确性和泛化能力。根据ANN模型的说明，可以通过添加较少的合金元素来优化抗氧化性。