Sage Journals: Discover world-class research

Abstract

In order to reduce serious tendency of the shrinkage porosity in ZL205A alloy, the design of experiment technique, Taguchi method is introduced to optimize the process parameters of cylindrical shell ZL205A alloy parts produced by low-pressure die casting in this article. The effect of pouring temperature, filling time and packing pressure on the shrinkage porosity in ZL205A castings is investigated using the orthogonal array and Advanced Porosity Model in ProCAST to calculate the volume of shrinkage porosity. The effectiveness of selected casting parameters in reducing the shrinkage porosity in cylindrical shell ZL205A casting is verified through analysis of signal-to-noise ratio and analysis of variance. Numerical results indicate that the optimized process parameters, including a pouring temperature of 725 °C, filling time of 20 s and packing pressure of 175 kPa, can be used to reduce the formation of shrinkage porosity in ZL205A castings.

Keywords

Optimization ZL205A low-pressure die casting shrinkage porosity Taguchi method

Introduction

ZL205A aluminum alloy is widely used in the aviation, aerospace and military industry because of its excellent mechanical properties at both room and elevated temperatures.^1,2 Due to its wide crystallization range and mushy solidification mode, this alloy exhibited a high tendency of shrinkage porosity at a temperature of 633 °C–544 °C.³ Therefore, in order to achieve the design performance of the castings, it is very important to predict and reduce the shrinkage porosity of ZL205A alloy castings to an acceptable level in the process design stage.⁴

Taguchi is one of the widely used robust design methods, which focuses on the effective application of engineering strategies. The effective investigation of individual and combined effects of various design parameters can be obtained by this method from orthogonal array and estimate indexes.^4–7 Compared with conventional orthogonal design related to the production quality, Taguchi method not only concerns the effect of process parameters on the variation of the production quality characteristics but also minimizes the variability of the quality characteristics by avoiding the effect of uncontrolled factors.^8–10 In the past decades, Taguchi method has been widely applied to optimize process parameters and reduce variations in the product quality, which has been proved effectively by several researchers. DH Wu used Taguchi method to optimize the process parameters for the die casting production of thin-walled magnesium alloy parts and establish the optimal combination of design parameters.⁵ Balasubramanian et al.¹¹ adopted modified Taguchi method (MTM) to improve the ultimate tensile strength, yield strength and notch tensile strength of pulsed current gas tungsten arc welding. Li et al.¹² used Taguchi method to establish the optimal conditions for the best roughness in chemical mechanical polishing (CMP). Ullmann et al.¹³ used Taguchi method to investigate the effect of various rolling parameters on the ductility of magnesium alloys. Aliofkhazraei et al.¹⁴ used Taguchi method to control the coating process factors and optimize the nanocrystalline plasma electrolytic nitrocarburizing to promote surface hardening and corrosion protection of austenitic stainless steel. Basavarajappa and Chandramohan¹⁵ used Taguchi method to study the influence of wear parameters on the dry sliding wear of metal matrix composites and established the correlation between dry sliding wear of composites and wear parameters by multiple regressions. Ömer and Ramazan⁴ used Taguchi method to investigate the relationship between microporosity and process variables in the sand cast A360 aluminum alloy and proposed the significant effect of local solidification time and dissolved hydrogen level in melt on microporosity. Razani et al.¹⁶ developed out-of-roundness (OOR) optimization for flow-formed AISI 321 stainless steel tube based on the Taguchi method. Sun et al.^17,18 designed the optimal gating system with multiple performance characteristics for cylindrical magnesium casting by Taguchi method and simulation results. Lee and Song¹⁹ adopted fuzzy logic and Taguchi method to estimate submerged-arc welding design parameters based on the simulation results. Dabade and Bhedasgaonkar²⁰ used Taguchi method and computer-assisted casting simulation techniques to reduce shrinkage porosity and improve yield in green sand casting of the ductile iron through the improved gating and feeding system.

Although the Taguchi method can be used to drastically reduce the number of experiments, carrying out all experiments, which is designed by orthogonal array with much parameters or large range of levels, is still time-consuming and cost-consuming. The numerical simulation of casting processes is the most powerful tool to decrease the cost and the time, which has been extensively used to predict the forming process and improve the design of the casting parameters.²¹ Therefore, the effect of casting process parameters on the shrinkage porosity, and the optimal combination of process parameters for the cylindrical shell ZL205 alloy castings produced by low-pressure die casting (LPDC) process, is investigated in this article using Taguchi approach and simulation technique. The contribution of each process parameter is also acquired. And the information proposed by this article would assist casting engineers to design appropriate process parameters in the production of cylindrical shell ZL205A alloy castings.

Experimental procedure

In order to identify the process parameters that may affect the density of castings produced by LPDC process, an Ishikawa cause–effect diagram is constructed, as shown in Figure 1. In this study, the parameters such as casting modulus, pouring temperature, pressure rate and packing pressure are selected, and their levels are shown in Table 1. However, for a more reliable design to produce high-quality ZL205A castings, the effect of more casting parameters on forming shrinkage porosity should be considered in future studies. The nominal composition of ZL205A alloy is given in Table 2.

Figure 1.

Cause–effect diagram of die casting defects.

Table 1.

Levels of process parameters.

Level	Parameters
	Pouring temperature, A (°C)	Filling time, B (s)	Packing pressure, C (kPa)
1	695	10	75
2	705	20	100
3	715	30	125
4	725	40	150
5	735	50	175

Table 2.

Chemical compositions (wt.%) of ZL205A alloy.

Cu	Mn	Ti	Zr	B	Cd	V	Zn	Al
Chemical composition
4.6–5.3	0.3–0.5	0.15–0.35	0.05–0.20	0.005–0.06	0.15–0.25	0.05–0.30	≤0.1	Bal.

Taguchi method for analysis of casting defects

During the design process, the errors caused by the variation of characteristic cannot be eliminated. Taguchi method focuses on a signal-to-noise (S/N) ratio instead of porosity simulation results to evaluate characteristic in the optimum setting analysis to minimize the influence of the variation of shrinkage porosity on the analysis of simulation data, because S/N ratio can reflect both the average and the variation of the quality characteristic.^14,16 Then the stability of the quality characteristic of products can be improved by controlling process parameters with significant influence.^9,10 There are three types of quality characteristics including lower the better (LB), nominal the best and higher the better.^4,5,15,22 This study is focused on producing minimum shrinkage porosity in ZL205A castings. Thus, the LB characteristic is implemented, and the S/N ratio for the LB characteristics is given by

S/N = - 10 \log (1 / n \sum_{i = 1}^{n} y_{i}^{2})

(1)

where n is the number of simulation experiments, y represents the ith response value and subscript i indicates the number of simulation design parameters.

In addition to the S/N ratio, the analysis of mean (ANOM) statistical approach can be used to estimate the optimal levels of process parameters. The means effects plots for S/N ratio can be used to determine the preferred levels of process parameters by the largest S/N ratio among all the levels of process parameters.¹⁹ This method can also be used to indicate the significant effect of process parameters on the formation of shrinkage porosity. But the ANOM approach has less strict logicality of the mathematics;⁸ therefore, analysis of variance (ANOVA) should be used to confirm the reliability of statistical significance of process parameters on the formation of shrinkage porosity.^4,13,16,19 The ANOVA statistical method can be given by the following equations⁶

{SS}_{T} = \sum_{i = 1}^{n} y_{i}^{2} - {(\sum_{i = 1}^{n} y_{i})}^{2} / N

(2)

S S_{p_{i}} = [\sum_{j = 1}^{r} \frac{p_{i} {(j)}^{2}}{N_{p_{ij}}}] - {(\sum_{i = 1}^{n} y_{i})}^{2} / N

(3)

DO F_{T} = N - 1, DO F_{p_{i}} = j - 1

(4)

V_{p_{i}} = \frac{S S_{p_{i}}}{f_{p_{i}}}

(5)

F_{p_{i}} = \frac{V_{p_{i}}}{V_{e}}

(6)

{SS}^{'}_{p_{i}} = {SS}_{p_{i}} - (V_{e} \times f_{p_{i}})

(7)

P_{p_{i}} = \frac{SS'_{p_{i}}}{S S_{T}} \times 100

(8)

where SS_T is the total sum of squares, n and N are the number of experiments, $S S_{p_{i}}$ is the sum of squares due to parameters, subscript i indicates the number of simulation design parameters, j is the level number of parameters given, r is the number of levels for parameters, degree of freedom (DOF) is the number of DOFs, $V_{p_{i}}$ is the mean square or variance of parameters, $F_{p_{i}}$ is the F ratio of parameters, SS′ is the corrected sums of squares and $P_{p_{i}}$ is the percentage contribution.

Casting simulation for analysis of casting defects

An actual casting with an outer radius of 1380 mm, an inner radius of 1350 mm, a height of 790 mm and a number of stiffeners inside is employed for simulating. The three-dimensional geometrical models of the castings, gating system and mould are implemented using the Unigraphics NX4.0 software of UGS Corp. The geometrical models are converted into *.x_t files to generate the finite element method (FEM) meshes by the MeshCAST module of ProCAST software. Once the FEM meshes are automatically generated by this module, the casting process parameters are regarded as initial boundary conditions for the simulation and defined in the preprocessor.

The Advanced Porosity Model (APM) module (ProCAST, 2006) is used to obtain the accurate volume of shrinkage porosity and calculate the packing pressure on the casting and the pressure drop within the mushy zone during solidification. This postprocessor can be used to graphically predict the defects of microporosity and macroporosity and the total volume of porosity. The numerical simulation results of shrinkage porosity ratio can be calculated using the following equation

y_{i} (‰) = \frac{V_{P o r o s i t y}}{V_{C a s t i n g}}

(9)

where $V_{Porosity}$ is the volume of all the pores in the casting and $V_{Casting}$ is volume of the whole casting.

Results and discussion

Each of the three process parameters investigated in this article has values at five levels, and L₂₅(5³) orthogonal array with 3 columns and 25 rows is selected for analysis. Compared with the classical combination method with full factorial experiment, which requires 3 × 5 = 243 sets of experiments to capture the influencing parameters, the selected orthogonal array can be used to obviously reduce the number of experiments. The process parameters and levels were arranged, as shown in Table 1, and the orthogonal array, the response results and the S/N ratio are given in Table 3, respectively, in which A, B and C indicate the pouring temperature, the filling time and the packing pressure, respectively.

Table 3.

Computation of S/N ratio for y_i (‰).

Experiment number	Parameter/level			y_i (‰)	S/N ratio (dB)
Experiment number	A (°C)	B (s)	C (kPa)	y_i (‰)	S/N ratio (dB)
1	695	10	75	8.003859	−18.066
2	695	20	100	7.46504	−17.4606
3	695	30	125	7.059216	−16.9751
4	695	40	150	6.635326	−16.4372
5	695	50	175	6.28835	−15.9707
6	705	10	100	7.419912	−17.408
7	705	20	125	6.920107	−16.8023
8	705	30	150	6.455726	−16.1989
9	705	40	175	6.199351	−15.8469
10	705	50	75	7.903549	−17.9564
11	715	10	125	6.813052	−16.6668
12	715	20	150	6.07395	−15.6694
13	715	30	175	6.057812	−15.6463
14	715	40	75	7.789645	−17.8304
15	715	50	100	7.368822	−17.348
16	725	10	150	6.311709	−16.0029
17	725	20	175	5.954552	−15.497
18	725	30	75	7.522659	−17.5274
19	725	40	100	7.097785	−17.0225
20	725	50	125	6.687127	−16.5048
21	735	10	175	5.995705	−15.5568
22	735	20	75	7.562218	−17.573
23	735	30	100	7.033366	−16.9433
24	735	40	125	6.684246	−16.501
25	735	50	150	6.786176	−16.6325

S/N: signal-to-noise.

Main effects

With process parameters combined in the selected orthogonal array, simulation experiments are run with 25 ZL205A alloy castings produced by LPDC process. Based on equation (1), the shrinkage porosity ratio in the values of S/N ratio is shown in Table 3. And the S/N ratios are employed as a response index to compare the shrinkage porosity ratio of different process parameters instead of directly using the values of shrinkage porosity ratio. The ANOM statistical approach referring to S/N ratios of various process parameters at different levels is given in Table 4, and the S/N response graphs are also plotted in Figure 2. It can be seen from Figure 2(a) that the S/N ratio increases as the pouring temperature increases from 695 °C (level 1) to 725 °C (level 4), and the S/N ratio decreases when the pouring temperature increases to 735 °C. It can also be seen from Figure 2(b) that the highest value of S/N ratio presents at the filling time of 20 s, and a shorter or longer filling time is harmful to the quality of casting. And it can be seen from Figure 2(c) that the S/N ratio increases with the increasing packing pressure, but the slope of this line decreases from 0.022 to 0.019 because of the mushy solidification of ZL205A alloy. This observation on the response of S/N ratio to packing pressure is consistent with the result reported by Hou et al.²³

Table 4.

Main effect for S/N ratio of raw data.

Process parameters	A	B	C
Level 1	−16.9819	−16.7401	−17.7906
Level 2	−16.8425	−16.6005	−17.2365
Level 3	−16.6322	−16.6582	−16.69
Level 4	−16.5109	−16.7276	−16.1882
Level 5	−16.6413	−16.8825	−15.7036
Maximum–minimum	0.331581	0.282029	1.532908
Rank	2	3	1

Figure 2.

Signal-to-noise response graph: (a) pouring temperature, (b) filling time and (c) packing pressure.

According to Taguchi method, for each parameter, a larger S/N ratio means a smaller variance of performance around the objective value, and a high influence on shrinkage porosity corresponds to the maximum S/N ratio difference at five levels. Therefore, it can be seen from Table 4 that the combination of optimal process parameters is A4B2C5, and the corresponding values are a pouring temperature of 735 °C, a filling time of 20 s and a packing pressure of 175 kPa. It can be seen from the difference of S/N ratio that the packing pressure has the greatest effect on the shrinkage porosity, and it is followed by pouring temperature and filling time.

As a matter of fact, a proper pouring temperature keeps the fluidity of melt good, to reduce the shrinkage porosity in a casting in the solidification process. A suitable filling time provides long enough exhausting time, to prevent the entry of oxide film into the melt. A high packing pressure facilitates the interdendritic flow and feeding, to prevent the formation of porosity.

ANOVA

In order to conduct the process parameters contribution more accurately and reliably, variance is analyzed using equations (2)–(8) to investigate the sequence of the important process parameters. It can be seen from Table 5 that the pouring temperature of liquid alloy and the packing pressure have significant effect on the solidification of castings, and factors A and C are of 99% confidence. This also means that filling time has a less significant effect on the shrinkage porosity in castings. The percentage contributions (P%) of each process parameter on the total variation are shown in the last column of Table 5.

Table 5.

Analysis of variance for raw data.

Factors	Sum of squares (SS)	Degrees of freedom (DOF)	Variance (V)	F ratio	SS′	P (%)
Pouring temperature, A	0.235	4	0.112	5.618^a	0.368	3.905
Filling time, B	0.274	4	0.034	1.694^b	0.055	Pooled
Packing pressure, C	9.439	4	2.150	107.948^a	8.521	90.435
Error	0.317	12	0.020	–	–	5.074^c
Total	56.428	24	–	–	–	100

At least 99% confidence.

Pooling, ep.

V_e (pooled) = 0.023; P_e (pooled) = 5.660.

Confidence interval

The optimum value of shrinkage porosity ratio can be predicted for the selected levels of significant parameters. The significant process parameters and their optimum levels have already been selected as A₄ and C₅. The estimated mean of shrinkage porosity ratio can be expressed as⁶

SP = {\bar{A}}_{4} + {\bar{C}}_{5} - \bar{P}

(10)

where $\bar{P}$ is the overall average value of shrinkage porosity ratio, ${\bar{A}}_{4}$ is the average shrinkage porosity ratio at the fourth level of pouring temperature and ${\bar{C}}_{5}$ is the average shrinkage porosity ratio at the second level of packing pressure.

Substituting the values of these terms into equation (10)

SP = 6.833 + 6.715 - 6.099 = 5.930 ‰

The 99% confidence interval of the estimates above is calculated using the following equations^6,22,24

C I = \sqrt{F_{α} (4, 12) \times V_{e} (\frac{1}{n_{eff}})}

(11)

n_{eff} = \frac{N}{1 + \sum DO F_{p}}

(12)

where V_e is the error variance, N is the total number of results and DOF_p is the total associated DOF for the estimated of mean.

The 99% confirmation interval for the predicted optimum shrinkage porosity ratio is 5.930 ± 0.213‰, which means the predicted optimal range is 6.143‰ < SP < 5.717‰.

Confirmation experiment

Two confirmation experiments and simulations are conducted with the optimal and the initial process parameters, respectively. The initial settings are a pouring temperature of 715 °C, a filling time of 20 s and a packing pressure of 100 kPa. It can be seen from Table 3 that the shrinkage porosity ratio simulated with these initial process parameters is 7.107‰, and that simulated with the optimal process parameters is 5.955‰. As shown by the X-ray examination results of casting produced, there is no shrinkage porosity in the casting produced with the optimum setting of process parameters, and the other is incompetent.

Conclusion

Taguchi method is employed in this article to investigate the effect of three well-known process parameters on the formation of shrinkage porosity in ZL205A aluminum alloy castings produced by LPDC process. The following conclusions can be drawn from the simulation results of shrinkage porosity ratios:

The optimal combination A4B2C5 has been obtained using Taguchi method, which means that a pouring temperature of 725 °C, a filling time of 20 s and a packing pressure of 175 kPa can be used to achieve quality ZL205A alloy castings produced by LPDC process.

Packing pressure has the most significant effect on the formation of shrinkage porosity. Pouring temperature also has a significant effect on the formation of shrinkage porosity. Filling time has a less significant effect on the formation of shrinkage porosity volume in castings.

The predicted range of optimum casting shrinkage porosity ratio for the 99% confidence interval is 6.877‰ < shrinkage porosity ratio < 6.451‰.

Footnotes

Declaration of conflicting interests

The authors declare that there is no conflict of interest.

Funding

This study was supported by State Key Laboratory of Materials Processing and Die & Mould Technology, Huazhong University of Science and Technology (09-04) and National Key Technology R&D Program (2011BAE21B02).

References

Jia

Wang

. The key technology of a stand casting of high strength Al2Cu alloy ZL205A. Chinese J Rare Met1999; 23(2): 153–156.

. Casting process of hanging rack on aircraft made of high strength aluminum alloy ZL205. J Mater Eng2001; 1: 43–46.

Zhai

Feng

. Formation law and criterion of nebulous macroscopic segregation in ZL205A alloy castings. China Foundry2008; 5(1): 20–23.

Ömer

Ramazan

. Application of Taguchi’s methods to investigate some factors affecting microporosity formation in A360 aluminium alloy casting. Mater Design2007; 28: 2224–2228.

Der

Mao

. Use of Taguchi method to develop a robust design for the magnesium alloy die casting process. Mat Sci Eng A: Struct2004; 379: 366–371.

Kumar

Shan

. Optimization of tensile properties of evaporative pattern casting process through Taguchi’s method. J Mater Process Tech2008; 204: 59–69.

Chen

. Research and practice on process strategy of robust design. Master’s Degree Dissertation, Zhejiang University, Hangzhou, China, 2003, p.15.

. Research of small batch production process quality control by Taguchi method. Master’s Degree Dissertation, Nanjing University of Science and Technology, Nanjing, China, 2008, p.33.

Hamada

. Using statistically designed experiments to improve reliability and to achieve robust reliability. IEEE T Reliab1995; 44(2): 206–215.

10.

Sun

. Robust design method and its application in the rocket engine design. Master’s degree dissertation, Northwestern Polytechnical University, Xían, China, 2006, p.17.

11.

Balasubramanian

Jayabalan

Balasubramanian

. Process parameter optimization of the pulsed current argon tungsten arc welding of titanium alloy. J Mater Sci Technol2008; 24(3): 423–426.

12.

Zhu

Zuo

. Optimization of polishing parameters with Taguchi method for LBO crystal in CMP. J Mater Sci Technol2009; 25: 703–707.

13.

Ullmann

Saleh

Schmitdchen

. Improvement of ductility for Twin Roll Cast and rolled AZ31 strips by use of Taguchi method. Arch Civ Mech Eng2013; 13: 1–6.

14.

Aliofkhazraei

Taheri

Rouhaghdam

. Systematic study of nanocrystalline plasma electrolytic nitrocarburising of 316L austenitic stainless steel for corrosion protection. J Mater Sci Technol2007; 23: 665–671.

15.

Basavarajappa

Chandramohan

. Wear studies on metal matrix composites: a Taguchi approach. J Mater Sci Technol2005; 21(6): 845.

16.

Razani

Aghchai

Dariani

. Experimental study on flow forming process of AISI 321 steel tube using the Taguchi method. Proc IMechE, Part B: J Engineering Manufacture2011; 225: 2024–2031.

17.

Sun

Chen

. Numerical optimization of gating system parameters for a magnesium alloy casting with multiple performance characteristics. J Mater Process Tech2008; 199: 256–264.

18.

Sun

Chen

X. Arunachalam

. Gating system design for a magnesium alloy casting. J Mater Sci Technol2008; 24(1): 93–95.

19.

Lee

Song

. Estimation of submerged-arc welding design parameters using Taguchi method and fuzzy logic. Proc IMechE, Part B: J Engineering Manufacture2013; 227(4): 532–542.

20.

Dabade

Bhedasgaonkar

. Casting defect analysis using design of experiments (DoE) and computer aided casting simulation technique. Procedia CIRP (Forty Sixth CIRP Conference on Manufacturing Systems 2013)2013; 7: 616–621.

21.

Gantar

Pepelnjak

Kuzman

. Optimization of sheet metal forming processes by the use of numerical simulations. J Mater Process Tech2002; 131: 54–59.

22.

Vijian

Arunachalam

. Optimization of squeeze cast parameters of LM6 aluminium alloy for surface roughness using Taguchi method. J Mater Process Tech2006; 180: 161–166.

23.

Hou

Cheng

. The technology of the prediction of shrinkage and porosity during low pressure casting process. Foundry Technol2004; 25(10): 769–771.

24.

Syrcos

. Die casting process optimization using Taguchi methods. J Mater Process Tech2003; 135: 68–74.

Optimization of low-pressure die casting process parameters for reduction of shrinkage porosity in ZL205A alloy casting using Taguchi method

Abstract

Keywords

Introduction

Experimental procedure

Taguchi method for analysis of casting defects

Casting simulation for analysis of casting defects

Results and discussion

Main effects

ANOVA

Confidence interval

Confirmation experiment

Conclusion

Footnotes

Declaration of conflicting interests

Funding

References