Data–Physics Fusion-Driven Defect Predictions for Titanium Alloy Casing Using Neural Network

Peng Yu,Xiaoyuan Ji,Tao Sun,Wenhao Zhou,Wen Li,Qian Xu,Xiwang Qie,Yajun Yin,Xu Shen,Jianxin Zhou

DOI: https://doi.org/10.3390/ma17102226

IF: 3.4

2024-05-10

Materials

Abstract:The quality of Ti alloy casing is crucial for the safe and stable operation of aero engines. However, the fluctuation of key process parameters during the investment casting process of titanium alloy casings has a significant influence on the volume and number of porosity defects, and this influence cannot be effectively suppressed at present. Therefore, this paper proposes a strategy to control the influence of process parameters on shrinkage volume and number. This study constructed multiple regression prediction models and neural network prediction models of porosity volume and number for a ZTC4 casing by simulating the gravity investment casting process. The results show that the multiple regression prediction model and neural network prediction model of shrinkage cavity total volume have an accuracy of over 99%. The accuracy of the neural network prediction model is higher than that of the multiple regression model, and the neural network model realizes the accurate prediction of shrinkage defect volume and defect number through pouring temperature, pouring time, and mold shell temperature. The sensitivity degree of casing defects to key process parameters, from high to low, is as follows: pouring temperature, pouring time, and mold temperature. Further optimizing the key process parameter window reduces the influence of process parameter fluctuation on the volume and number of porosity defects in casing castings. This study provides a reference for actual production control process parameters to reduce shrinkage cavity and loose defects.

materials science, multidisciplinary,chemistry, physical,physics, applied, condensed matter,metallurgy & metallurgical engineering

What problem does this paper attempt to address?

This paper mainly discusses the shrinkage defects (such as pores and looseness) in the precision casting process of titanium alloy casings. The author proposes a defect prediction strategy driven by data-physical fusion, using a neural network model to predict and control these defects. The study constructs a multivariate regression prediction model and a neural network prediction model by simulating the gravity investment casting process to predict the influences of pouring temperature, mold shell temperature, and pouring time on defect volume and quantity. The study shows that both the multivariate regression prediction model and the neural network prediction model have a prediction accuracy of over 99% for the total volume of shrinkage cavities, with the neural network model having higher accuracy. Sensitivity analysis shows that the degree of influence of key process parameters on defects from high to low is: pouring temperature, pouring time, and mold temperature. By optimizing the range of these key process parameters, casting defects can be reduced and the quality of titanium alloy casings can be improved. The paper also introduces the use of orthogonal experimental design to reduce the number of simulation experiments in order to determine the optimal process scheme, and constructs a neural network prediction model with inputs of pouring temperature, pouring time, and mold shell temperature, and outputs of defect volume and quantity. The model adopts activation functions Tanh and ReLU to handle nonlinear problems. In summary, the paper aims to provide guidance for controlling the casting defects of titanium alloy casings in actual production by combining data-driven methods with physical simulation, in order to improve the safety and stability of aerospace engine components.

Data–Physics Fusion-Driven Defect Predictions for Titanium Alloy Casing Using Neural Network

Prediction of Fatigue Life for Automotive A-Alloy Wheel with Integrated Cast Defects

Numerical Simulation for the Investment Casting Process of a Large-Size Titanium Alloy Thin-Wall Casing

A Self-Training-based Approach for Aluminum Alloy Casting Quality Prediction

Prediction method of shrinkage cavity and shrinkage porosity in solidification process under low-pressure casting

Deformation Intelligent Prediction of Titanium Alloy Plate Forming Based on BP Neural Network and Sparrow Search Algorithm

Energy-based efficient residual stress prediction of titanium alloys with lack-of-fusion porosity in powder bed fusion

Simulation Study on the Investment Casting Process of a Low-Cost Titanium Alloy Gearbox based on ProCAST

Research on prediction of microporosity in Ti-Al alloy investment casting

Artificial neural network approach for mechanical properties prediction of as-cast A380 aluminum alloy

Optimization of vacuum counter-pressure casting process for an aluminum alloy casing using numerical simulation and defect recognition techniques

Ductile fracture prediction of HPDC aluminum alloy based on a shear-modified GTN damage model

Data-driven prediction of keyhole features in metal additive manufacturing based on physics-based simulation

Casting Design Optimization Using Numerical Simulation of Precision Forging Die for Large Titanium Alloy Integral Bulkhead

Machine learning-enabled predictions of as-built relative density and high-cycle fatigue life of Ti6Al4V alloy additively manufactured by laser powder bed fusion

Digital Twin and Data-driven Quality Prediction of Complex Die-casting Manufacturing

Research on numerical simulation of investment casting process for gamma titanium aluminide turbocharger

A physics‐informed neural network framework based on fatigue indicator parameters for very high cycle fatigue life prediction of an additively manufactured titanium alloy

Effect of centrifugal casting process on mold filling and grain structure of K418B turbine guide

Computer-Aided Engineering (CAE) Simulation for the Robust Gating System Design: Improved Process for Investment Casting Defects of 316L Stainless Steel Valve Housing