Abstract
Taking a stainless steel three-way valve body as the research object, this paper addresses the issues of shrinkage porosity, shrinkage, and deformation that frequently occur during investment casting production. By leveraging ProCAST software and grey correlation analysis, the production process was studied. The causes of defects in the original scheme were analyzed by improving the gating system. Orthogonal experiments were designed with shrinkage porosity, shrinkage, and deformation as quality objectives. The theoretical optimal process parameters were determined through grey correlation analysis, and the significance of different factors affecting the castings was analyzed. The results demonstrate that grey correlation analysis is suitable for optimizing investment casting process parameters. It was found that the pouring system has the greatest impact on casting quality, followed by pouring time, shell roasting temperature, and pouring temperature. The optimal process parameters obtained are a pouring temperature of 1,610 °C, a shell roasting temperature of 1,050 °C, and a pouring time of 6 seconds.
Keywords: shrinkage porosity; deformation; numerical simulation; process optimization; grey correlation analysis
1. Introduction
Investment casting is an important manufacturing process for near-net-shape forming of three-way valve bodies. It is characterized by its applicability to the production of complex structural castings and high process yield, but there are numerous processes that affect casting quality [1, 2], and castings are prone to defects such as shrinkage porosity, shrinkage, sand inclusion, and deformation. These defects can reduce the mechanical properties and reliability of castings. Optimizing the structure of the gating system and process parameters are effective measures to avoid the above defects [3].
Currently, casting process optimization is mainly based on a combination of simulation and experimental exploration. Researchers such as DONG J G [4] have used ProCAST software to conduct numerical simulations of the LPDC process for A356 alloy wheel hubs, studying the forming quality of wheel hubs under different cooling conditions. By comparing the tensile properties of wheel rims under different cooling conditions, it was found that wheel hubs cooled for 140 s at a rate of 400 L/h exhibited higher tensile strength and elongation, which were 178.9 MPa and 6.6%, respectively. YUWEN X X et al. [5] conducted numerical simulations based on finite element theory using Fluent software to analyze the flow of molten metal and temperature field during the pouring process, setting boundary conditions and initial conditions such as pouring speed, temperature, and heat transfer coefficient, and evaluated the feasibility of using Fluent for casting simulations. They observed the flow of molten metal, recorded temperature distribution data at the end of pouring, and provided initial conditions for further exploration of the solidification process of castings.
2. Material, Casting Structure, and Process Optimization
2.1 Casting Material and Structure Analysis
The casting has a mass of 2.42 kg, a total length of 150 mm, a maximum outer diameter of 78 mm at the connection, a maximum wall thickness of 24.96 mm, and a minimum wall thickness of 4 mm. The casting has a three-way structure. The casting has uneven wall thickness, characterized by being thicker at the top and thinner at the bottom, thicker in the middle and thinner at the end face, with thin inner walls and a complex structure. The inner cavity requires a smooth and flat surface with strict dimensions. The original process is prone to defects such as shrinkage porosity and deformation during the solidification of the casting.
2.2 Defect Cause Analysis
The main reasons for the defects are related to the casting structure and process parameters. The uneven wall thickness of the casting leads to uneven cooling rates during solidification, resulting in shrinkage porosity and deformation. In addition, inappropriate pouring temperature, shell roasting temperature, and pouring time can also lead to casting defects.
3. Process Optimization Scheme
In addition to being related to the casting structure, appropriate pouring temperature, shell roasting temperature, and pouring time are guarantees of casting quality. A high pouring temperature facilitates casting formation, but coarse grains and deformation and shrinkage porosity are prone to occur [6]. A slow pouring speed can reduce shrinkage porosity in castings. A low shell temperature is prone to cold shuts and inclusions, but an excessively high temperature can result in coarse casting grains and reduced mechanical properties. Based on the above reasons, the following optimization scheme was formulated: Firstly, the number of ingates was increased to feed the thicker parts of the casting, and the improved gating system model. Secondly, orthogonal experimental method and grey correlation analysis were used to screen out the best process parameters.
The volume of shrinkage porosity and deformation are important indicators for evaluating casting quality [8]. Pouring temperature (A), shell roasting temperature (B), and pouring time (C) are important process parameters in the investment casting process. The influence of the above three factors on the volume of shrinkage porosity and deformation was studied. The orthogonal experimental factor level table and orthogonal experimental scheme are shown in Tables 1 and 2, respectively.
Table 1: Orthogonal Experimental Factor Level Table
| Level | Factor A (°C) | Factor B (°C) | Factor C (s) |
|---|---|---|---|
| 1 | 1,610 | 1,000 | 4 |
| 2 | 1,640 | 1,050 | 6 |
| 3 | 1,670 | 1,100 | 8 |
Table 2: Orthogonal Experimental Scheme Table
| No. | Factor A (°C) | Factor B (°C) | Factor C (s) |
|---|---|---|---|
| 1 | 1,610 | 1,000 | 4 |
| 2 | 1,610 | 1,050 | 6 |
| 3 | 1,610 | 1,100 | 8 |
| 4 | 1,640 | 1,000 | 6 |
| 5 | 1,640 | 1,050 | 8 |
| 6 | 1,640 | 1,100 | 4 |
| 7 | 1,670 | 1,000 | 8 |
| 8 | 1,670 | 1,050 | 4 |
| 9 | 1,670 | 1,100 | 6 |
4. Experimental Results and Analysis
4.1 Grey Correlation Analysis
Grey correlation analysis is a statistical analysis method suitable for quality objectives influenced by multiple factors. Grey correlation analysis has low requirements for data regularity and can have a sample size as small as four, which can reduce the number of experiments and thus reduce production costs and improve R&D efficiency. The core idea is to first dimensionless the original quality objectives, then calculate the correlation coefficient and correlation degree, and analyze the significance of factor influences and rank them based on the correlation degree.
Firstly, the original quality objective data was dimensionless using interval value transformation:
Where x_i is the experimental value for the i-th trial; y_i is the dimensionless value.
The grey correlation coefficient calculation formula is:
Where y_i0 is the ideal optimal dimensionless value of the reference sequence for the i-th trial; ρ is the resolution coefficient, ρ ∈ [0,1] is taken as 0.5 to weaken the distortion effect caused by excessively large values.
4.2 Weight Calculation of Indicators
The weight coefficients of each indicator were obtained, as shown in Table 3.
Table 3: Calculation Results of Indicator Weights
| Indicator | Weight |
|---|---|
| Volume of shrinkage porosity | 0.2873 |
| Deformation amount | 0.7127 |
After obtaining the weight coefficients of each indicator, the grey correlation degree of each scheme was calculated using the following formula:
4.3 Experimental Results and Analysis
The experimental results and grey relational analysis calculations are presented in Table 5, where y represents the original data value, y’i represents the dimensionless value, δi represents the grey relational coefficient, and Gi represents the grey relational degree. From Table 5, it can be observed that after adopting the improved pouring scheme, the volume of shrinkage porosity and deformation significantly decreased. Specifically, the volume of shrinkage porosity in the third scheme was the smallest, at 0.0446 cm³, and the deformation in the second scheme was the smallest, at 0.0239 cm. The larger the grey relational degree at any parameter level indicates better corresponding quality objectives. By comparing the grey relational degrees, the experimental schemes for different parameter levels were ranked, as shown in Table 4. The maximum grey relational degree was 0.9208, and the minimum was 0.3333. Therefore, the optimal combination of factor levels was A1B2C2, which corresponds to a pouring temperature of 1610°C, a shell roasting temperature of 1050°C, and a pouring time of 6 seconds.
4.4 Mean Range Analysis of Grey Relational Degree
The mean range analysis of the grey relational degree is presented in Table 6. It can be seen that the mean range of the shell roasting temperature (B) was the smallest, at 0.2033, while the mean range of the pouring time (C) was the largest, at 0.3476. Therefore, the significance order of factors affecting the casting quality was B (shell roasting temperature) < A (pouring temperature) < C (pouring time), indicating that the pouring time had the most significant impact on the casting quality, followed by the pouring temperature and shell roasting temperature, which had relatively similar impacts.
The trend of the grey relational degree with factor level changes. The mean grey relational degree of factor A decreased rapidly from level 1 to level 2 and then increased slowly, indicating that a smaller A factor value was better within the range from level 1 to level 2. To achieve the optimal quality objective, the A factor could potentially be set to an even smaller value. The mean grey relational degree of factor B showed a uniform increasing trend, suggesting that the B factor value could be further increased within the range to optimize the quality objective. Factor C first increased rapidly and then decreased rapidly, indicating that level 2 was the optimal factor level for C, which could significantly improve the quality objective compared to other factor levels.
4.5 Analysis of Casting Filling Test Results
The filling test adopted the optimized pouring system model and the best process parameters: a pouring temperature of 1610°C, a shell roasting temperature of 1050°C, and a pouring time of 6 seconds. During pouring, the molten metal entered the sprue from the pouring cup, then flowed into the mold cavity, and finally into the runner. The molten metal filled the mold from bottom to top, allowing gases to escape smoothly and preventing air entrapment. The filling process was stable and smooth, with no turbulence or metal splashing.
4.6 Analysis of Casting Solidification Test Results
The solidification process can be seen that the thinnest part of the casting solidified first, followed by the rest of the casting from bottom to top, then the runner and ingate, and finally the pouring cup. Before optimization, isolated liquid regions appeared at the thick intersection of the gating system. After optimization, the ingates on the sprue provided feeding to the areas with large shrinkage porosity, significantly reducing the volume of shrinkage porosity and achieving directional solidification.
5. Conclusion
(1) Through orthogonal experiments and grey relational analysis, it was found that the pouring system structure, pouring time, shell roasting temperature, and pouring temperature all affected the quality of a stainless steel three-way casting, verifying the feasibility of grey relational analysis in exploring investment casting process parameters.
(2) The optimal process parameters for the investment casting of a stainless steel three-way casting were determined to be a pouring temperature of 1610°C, a shell roasting temperature of 1050°C, and a pouring time of 6 seconds. Compared to the original scheme, the volume of shrinkage porosity decreased by 97%, and the deformation decreased by 27%.
(3) According to the mean range analysis of the grey relational degree, the pouring system structure had the greatest impact on the quality of the stainless steel three-way casting, followed by the pouring time, and then the shell roasting temperature and pouring temperature.