Predicting mechanical behavior and microstructure evolution via numerical simulation is an effective way to improve the quality of superplastic-formed products for nickel-based superalloys. Analysis of mechanical and microstructures during the stable superplastic deformation of Inconel 718 indicates that dislocation propagation and grain growth lead to stress increase, while recovery and dynamic recrystallization are responsible for flow softening. Moreover, the δ phase has an indirect effect on the flow stress by fixing grain boundaries and dislocations. The variation of δ phase content during superplasticity is attributed to the precipitate coarsening caused by thermal diffusion and the particle density decrease due to grain boundary movement. This paper proposes a viscoplastic constitutive model that accounts for the impact of internal state variables on the superplastic flow, such as dislocation density, recrystallization, grain size, and δ phase evolution. By describing changes in the average radius and volume fraction of particles, the model reflects the contribution of precipitate as a pinning point to promote dislocation proliferation and restrict grain boundary slip and grain growth. Moreover, the true strain rate varying with time step was introduced into the parameter optimization involving genetic algorithm to realize the adaptability of the model to large deformation. After encoding the material subroutine UMAT, the constructed model effectually forecasted the change in flow stress and internal variables during superplastic stretching under random conditions. Finally, the simulation and experiment of the bipolar plate forming verified the practicability of the model in predicting the thickness distribution and microstructure evolution for superplastic parts.

中文翻译：

---

Inconel 718合金超塑性变形内变量本构模型的建立及应用

通过数值模拟预测力学行为和微观结构演变是提高镍基高温合金超塑成形产品质量的有效方法。Inconel 718 稳定超塑性变形过程中的机械和微观结构分析表明，位错传播和晶粒生长导致应力增加，而回复和动态再结晶则导致流动软化。此外，δ相通过固定晶界和位错对流变应力有间接影响。超塑性过程中δ相含量的变化归因于热扩散引起的析出物粗化和晶界移动导致的颗粒密度降低。本文提出了一种粘塑性本构模型，该模型考虑了内部状态变量对超塑性流动的影响，例如位错密度、再结晶、晶粒尺寸和 δ 相演化。通过描述颗粒平均半径和体积分数的变化，该模型反映了析出物作为钉扎点促进位错扩散、限制晶界滑移和晶粒长大的贡献。此外，将随时间步长变化的真应变率引入遗传算法参数优化中，实现模型对大变形的适应性。对材料子程序UMAT进行编码后，构建的模型有效地预测了随机条件下超塑性拉伸过程中流变应力和内部变量的变化。最后，双极板成形模拟和实验验证了该模型在预测超塑件厚度分布和微观组织演化方面的实用性。