钛合金复杂构件熔模铸造缩孔缩松缺陷空间形态定量评估方法与工艺参数波动控制

The shrinkage and porosity is the main defect in investment casting of titanium alloy complex components. Its formation is affected by the fluctuation of the key process parameters in the production process, such as the pouring time and temperature of alloy, the initial temperature and thickness of the mold shell, but the influence mechanism is not clear yet. To this end, this project is intended to carry out the study of quantitative evaluation of shrinkage and porosity defects, to find out the sensitivity and influence mechanism of the key process parameters fluctuations on the evolution of shrinkage defects, and obtain fluctuation control strategy of process parameters in investment casting of Titanium alloy complex components. Firstly, using industrial data acquisition and thermal parameters reverse technology, to accurately obtain the key process parameters including pouring time and temperature of alloy, the initial temperature and thickness of the mold shell, and obtain thermophysical parameters of alloy and mold shell, and then to study the fluctuation characteristics of the process parameters. Secondly, the quantitative evaluation method of spatial shape of the position, size and shape of shrinkage and porosity defects will be established, and the parameter fluctuation-distribution data set will be obtained by orthogonal experiment and numerical simulation combined with the parameter fluctuation distribution model and thermal property parameters. Finally, the relationship model between the key process parameters of the typical investment casting process of titanium alloy complex components and the shrinkage and porosity defect will be established by adopting deep neural network method and the example verification to reveal the influence mechanism of parameter fluctuation on defects evolution, and then multiple key process parameter windows and the process parameters fluctuation control strategy will be obtained by using the multi-objective optimization algorithm. The project will provide theoretical and technical support for the quantitative control of shrinkage defects in the titanium alloy complex castings.

缩孔缩松（孔松）是钛合金复杂构件熔模铸造的主要缺陷，其形成受合金浇注时间与温度、模壳厚度与初温等关键工艺参数波动的耦合影响，而这种影响机制尚不清晰。为此，拟进行孔松定量评估研究，探明关键工艺参数波动对孔松缺陷演变的影响机制，获得工艺参数波动量化控制策略。首先，基于数据采集和热物性参数反求，准确获取合金浇注时间与温度、模壳厚度与初温等工艺参数及合金与模壳热物性参数，建立工艺参数波动分布模型；其次，建立孔松缺陷位置、大小、形状的空间形态定量评估方法，结合参数波动分布模型和热物性参数进行正交实验和数值模拟，获得参数波动-孔松缺陷数据集；最后，采用深度神经网络方法和实例验证，基于数据集建立钛合金复杂构件熔模铸造关键工艺参数与孔松缺陷关系模型，揭示参数波动对孔松演变的影响机制，并基于多目标优化获得多个关键工艺参数窗口与波动控制策略。项目可为钛合金复杂铸件孔松定量化控制提供理论和技术支撑。

缩孔缩松（孔松）是钛合金复杂构件熔模铸造的主要缺陷，其成形受合金浇注时间与温度、模壳温度等关键工艺参数波动的耦合影响，而这种影响机制尚不清晰。准确获取实际合金浇注时间与温度、模壳温度等大量工艺数据是分析工艺参数波动耦合影响的基础。为此，本项目以航发百慕高科公司钛合金复杂机匣铸件的实际生产质量数据为切入点，对铸件的孔松缺陷位置、大小、形状的空间形态进行定量评估，再结合数值模拟和神经网络阐明了多个关键工艺参数窗口与控制策略。首先，采集获取了实际工艺参数，分析建立了参数波动特征，确立了数值模拟工艺参数范围，设计了更符合实际、覆盖性更好的铸件数值模拟正交实验。其次，采用热物性参数反求技术进行现场测温试验，获得了更符合实际的模壳热物性参数；采用动态压强孔松预测模型，开展了准确的接近实际生产的单件化数值模拟，获得了大量的钛合金机匣铸件仿真数据。再次，采用层次分析法建立了孔松缺陷三维空间形貌特征量化表征方法，根据仿真数据结果和部分实际数据，揭示了孔松缺陷位置、大小、形貌的演变规律；使用多元线性回归建模分析了钛合金复杂构件凝固过程关键工艺参数与孔松缺陷（缩孔数、缩松数、缩孔总体积、缩松总体积、缩孔中心距最近表面的距离）之间的定量关系。最后，采用神经网络方法，构建了机匣铸件关键工艺参数与孔松缺陷内禀关系模型，计算获得了大量参数波动-孔松缺陷数据集，结合实际与计算的数据集探明了多个关键工艺参数波动对孔松缺陷演变的耦合影响规律及敏感性规律；基于优化算法获得了关键工艺参数优化窗口与波动控制策略。本项目可为钛合金复杂铸件孔松定量化控制提供理论和技术支撑。