本研究采用实验与模拟计算相结合的方法，深入探讨了深海流动环境下锰铝青铜合金的腐蚀行为与失效动力学过程。结果表明，在流动腐蚀过程中，存在一个使合金腐蚀速率急剧上升的临界流速(4m/s)。海水流动产生的切应力使MAB合金(111)晶面产生原子级的轴向拉伸应变，显著降低了Cu原子的迁移-溶解活化能垒，腐蚀电位明显负移。在流速与压力耦合作用下，合金表面由原有的“流痕”转变为连续的凹坑状，腐蚀速率显著增加。基于密度泛函理论(DFT)，通过构建海水流速与压力耦合作用下的铜合金微观结构模型，计算得到的腐蚀速率常数与实测值趋势基本一致，实现了深海动态环境下腐蚀倾向性的预测。基于响应面矩阵方差分析(ANOVA)表明，流速、溶解氧、压力和温度对合金腐蚀速率的影响程度依次递减，其中流速与压力的交互作用最为显著。最终建立的腐蚀预测模型R2值为94.59%，证实了模型具有较高的准确性。

The corrosion behavior and failure dynamics of MAB alloy in deep sea flow environment were studied by combining experiment and simulation. The results indicate that there is a critical flow velocity (4 m/s) at which the alloy corrosion rate sharply increases during flow-induced corrosion. The shear stress generated by seawater flow induces atomic-level axial tensile strain on the (111) crystal facet of the alloy, significantly lowering the migration-dissolution activation energy barrier of Cu atoms, leading to a noticeable shift in corrosion potential towards more negative values. Under the combined effects of flow velocity and pressure, the alloy surface transitions from the original “flow trace” pattern to continuous pits, resulting in a marked increase in corrosion rate. Based on density functional theory (DFT), a microstructural model of the copper alloy under combined seawater flow and pressure was developed. The calculated corrosion rate constants aligned closely with the experimental data, enabling effective prediction of corrosion tendencies in dynamic deep-sea environments. Response surface methodology variance analysis (ANOVA) revealed that flow velocity, dissolved oxygen, pressure, and temperature influence the corrosion rate of the alloy in descending order, with the interaction between flow velocity and pressure being the most significant. The final corrosion prediction model had an R2 value of 94.59%, demonstrating its high accuracy.