This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution license (http://creativecommons.org/licenses/by/3.0/).

This study applies the Taguchi method to investigate the relationship between the ultimate tensile strength, hardness and process variables in a squeeze casting 2017 A wrought aluminium alloy. The effects of various casting parameters including squeeze pressure, melt temperature and die temperature were studied. Therefore, the objectives of the Taguchi method for the squeeze casting process are to establish the optimal combination of process parameters and to reduce the variation in quality between only a few experiments. The experimental results show that the squeeze pressure significantly affects the microstructure and the mechanical properties of 2017 A Al alloy.

Recently, great attention has been focused on aluminium and its alloys due to their high technological value and wide range of industrial applications, thanks to their various advantages such as lower density, good formability, high thermal conductivity, high specific rigidity, excellent corrosion resistance, high castability and attractive tensile strength [

In recent years, a new casting technology called squeeze casting has been developed to make better use of aluminium alloys [

Many research works on squeeze casting parameters of aluminium alloys [

The present investigation aims, essentially, to determine a good combination of applied pressure, melt temperature and the die temperature for squeeze casting 2017 A wrought Al alloy. Ultimate tensile strength (UTS) and hardness tests of the liquid forged samples at different squeeze casting parameters were characterized and the optimal condition is found by the Taguchi method.

The squeeze casting process parameters namely squeeze pressure (A), melt temperature (B) and die temperature (C) at three levels are listed in

Results of L_{9} orthogonal array experiments.

No | A | B | C | UTS (MPa) | Hardness (HV) | ||||||
---|---|---|---|---|---|---|---|---|---|---|---|

Y_{1} |
Y_{2} |
Y_{3} |
Average | Y_{1} |
Y_{2} |
Y_{3} |
Average | ||||

1 | 30 | 700 | 200 | 176 | 178 | 170 | 174.667 | 65 | 69 | 64 | 66.000 |

2 | 30 | 750 | 250 | 159 | 162 | 167 | 162.667 | 56 | 62 | 59 | 59.000 |

3 | 30 | 800 | 300 | 154 | 157 | 148 | 153.667 | 54 | 58 | 56 | 56.000 |

4 | 60 | 700 | 250 | 178 | 198 | 189 | 188.333 | 77 | 65 | 79 | 73.666 |

5 | 60 | 750 | 300 | 178 | 172 | 175 | 175.000 | 74 | 65 | 68 | 69.000 |

6 | 60 | 800 | 200 | 175 | 180 | 185 | 180.000 | 76 | 65 | 78 | 73.000 |

7 | 90 | 700 | 300 | 208 | 213 | 216 | 212.333 | 82 | 80 | 86 | 82.666 |

8 | 90 | 750 | 200 | 202 | 204 | 209 | 205.000 | 80 | 77 | 84 | 80.333 |

9 | 90 | 800 | 250 | 198 | 203 | 196 | 199.000 | 78 | 70 | 74 | 74.000 |

The average value of UTS and hardness for each parameter A, B and C at level 1, 2 and 3 are grouped in

Levels average for main effects.

Level (L) | Average UTS (MPa) | Average Hardness (HV) | ||||
---|---|---|---|---|---|---|

A | B | C | A | B | C | |

L1 | 163.7 | 191.8 | 186.6 | 60.33 | 74.11 | 73.11 |

L2 | 181.1 | 180.9 | 183.3 | 71.89 | 69.44 | 68.89 |

L3 | 205.4 | 177.6 | 180.3 | 79 | 67.67 | 69.22 |

Max-Min | 41.8 | 14.2 | 6.2 | 18.67 | 6.44 | 4.22 |

Rank | 1 | 2 | 3 | 1 | 2 | 3 |

Experimental layout using L_{9} standard orthogonal array.

Test number | A | B | C |
---|---|---|---|

1 | 1 | 1 | 1 |

2 | 1 | 2 | 2 |

3 | 1 | 3 | 3 |

4 | 2 | 1 | 2 |

5 | 2 | 2 | 3 |

6 | 2 | 3 | 1 |

7 | 3 | 1 | 3 |

8 | 3 | 2 | 1 |

9 | 3 | 3 | 2 |

Main effects graph for ultimate tensile strength (UTS).

Main effects graph for hardness.

The analysis of variance (ANOVA) was used to investigate which parameters significantly affected the quality characteristic and to determine the percentage contribution of the parameters at 95% confidence level. The F ratio value named Fisher test was used to see which process parameters have a significant effect. Usually, when the

ANOVA analysis for UTS and hardness was carried out using Equations (2)–(6) and the resulting data is given in

Variance (ANOVA) Table for ultimate tensile strength (UTS).

Source | Degrees of freedom (DOF) | Sum of squares (SS) | Variance (V) | Percent contribution (P) | |
---|---|---|---|---|---|

A | 2 | 5.9391 | 2.9695 | 197.74 | 85.93 |

B | 2 | 0.7591 | 0.3795 | 25.28 | 10.98 |

C | 2 | 0.1831 | 0.0915 | 6.10 | 2.65 |

Error | 2 | 0.0300 | 0.0150 | 0.44 | |

Total | 8 | 6.9114 | 100.00 |

Variance (ANOVA) Table for hardness.

Source | Degrees of freedom (DOF) | Sum of squares (SS) | Variance (V) | Percent contribution (P) | |
---|---|---|---|---|---|

A | 2 | 8.4730 | 4.2364 | 122.33 | 83.06 |

B | 2 | 1.0712 | 0.5356 | 15.47 | 10.50 |

C | 2 | 0.5875 | 0.2937 | 8.48 | 5.76 |

Error | 2 | 0.0693 | 0.0346 | 0.68 | |

Total | 8 | 10.2009 | 100.00 |

Percentage contribution of significant control factors.

The next analysis was investigated by using analysis of signal-to-noise ratio (S/N). According to the data presented in

Computation of S/N ratio for ultimate tensile strength (UTS) and hardness.

No | Average UTS | Average hardness | S/N ratio of UTS | S/N ratio of hardness |
---|---|---|---|---|

1 | 174.667 | 66.000 | 44.839 | 36.377 |

2 | 162.667 | 59.000 | 44.220 | 35.394 |

3 | 153.667 | 56.000 | 43.727 | 34.952 |

4 | 188.333 | 73.666 | 45.473 | 37.246 |

5 | 175.000 | 69.000 | 44.858 | 36.739 |

6 | 180.000 | 73.000 | 45.098 | 37.180 |

7 | 212.333 | 82.666 | 46.537 | 38.334 |

8 | 205.000 | 80.333 | 46.232 | 38.081 |

9 | 199.000 | 74.000 | 45.974 | 37.359 |

Mean |

The combination shown above differs from the previously mentioned one in main effects. It sheds light on the optimum combination of parameters and their levels. However, it shows that, in the present case study, the combination of parameters and their levels A3 B1 C3 yield optimum mechanical properties with minimum variance from the target value.

The purpose of estimation of predicted means is to validate the squeeze casting condition at the optimal levels of parameters, which is A3 B1 C1 for mechanical properties. The predicted mean (μ) for UTS and hardness was estimated using the following two equations [_{UTS}_{hardness}_{UTS} = 183.407 + (205.4 − 183.407) + (191.8 − 183.407) + (186.6 − 183.407) = 216.986 MPa
_{hardness} = 70.407 + (79 − 70.407) + (74.11 − 70.407) + (73.11 − 70.407) = 85.406 HV

Three confirmation tests are conducted at the optimum settings of squeeze casting parameters recommended by the investigation. The average values of UTS and hardness obtained at the optimum settings of the process parameters are 219.333 MPa and 86.666 HV, respectively. We notice that the difference between the estimated results and the experimental results is negligible. Therefore, the experimental values are within the confidence interval of the predicted optimal of mechanical properties.

The influence of squeeze pressure (the most significant factor) on the microstructure and the mechanical properties has been analyzed on the basis of the statistical analysis developed in

Optical micrographs of the squeeze cast sample (

These micrographs show that the microstructures prepared under higher applied pressures are much finer and smaller α-primary dendrites. It is clear that the squeezing pressure has significant influence on the microstructure of the alloy [

where _{f}_{l} and _{s} are the specific volumes of the liquid and solid, respectively, and Δ_{f}_{f}_{l} − V_{s})_{f}

The mechanical properties of squeeze cast specimens such as ultimate tensile strength (UTS) and hardness (HV) are compiled in

Evidently the improvement of mechanical properties by increasing the pressure up to 90 MPa seems to be attributed, in part, to the refinement of the α-primary dendrites and, in part, to material densification.

Ultimate tensile strength (UTS) and hardness of 2017 A Al alloy manufactured in various conditions.

The traditional experimental techniques,

In this study, the Taguchi method has been adopted to observe the influencing process parameters in the squeeze casting process. Taguchi statistical design is adopted to understand the effect of these processing parameters by running only a few experiments while achieving strong mechanical properties. The casting parameters each at three levels considered in this study and the details are presented in

Squeeze casting parameters and their levels.

Notation | Parameters | Level 1 | Level 2 | Level 3 |
---|---|---|---|---|

A | Squeeze pressure (Mpa) | 30 | 60 | 90 |

B | Melt temperature (°C) | 700 | 750 | 800 |

C | Die temperature (°C) | 200 | 250 | 300 |

The Taguchi technique employs a generic signal-to-noise (S/N) ratio to quantify the present variation. Broadly speaking, the (S/N) ratio is the ratio of the mean (signal) to the standard deviation (noise). Depending on the particular type of characteristics involved, three types of S/N ratios are applicable, including “higher is better” (HB), “lower is better” (LB) and “nominal is best” (NB). Because the target of this work is to maximize the mechanical properties (UTS and hardness), the S/N ratio with HB characteristics is required, which is given by:

A statistical analysis of variance (ANOVA) can be performed in order to see which process parameter (factor) is statistically significant for each quality characteristic (see Equations (7)–(11) [_{factor}_{factor}_{factor}_{factor}_{error}_{total} is the total sum of squares, N is the total number of experiments, _{A} the factorial sum of squares due to factor _{A} represents the number of levels for factor _{i}_{Ai}_{factor} the variance of the factor, _{factor} represents the sum of squares of the factor and _{factor} is the

The experiments were carried out using a 2017 A wrought aluminium alloy. The material provides average tensile strength but good machinability. It is widely used in mechanical applications [

The squeeze casting experiments were performed on a hydraulic press (see

Experimental setup of squeeze casting process.

Schematic representation of squeeze casting process.

The tensile specimens were machined to evaluate the ultimate tensile strength (UTS). For each experimental condition, three specimen samples were prepared. INSTRON (ENSIL, Limoges, France) universal testing machine was used for performing tensile tests on the specimens. The tests were performed under displacement control with a strain rate start at 1 mm·min^{−1}. An extensometer (gage length of 14.3 mm) is attached with two rubber bands to the central part of the specimen.

Hardness analysis HV was performed on a transverse section of the specimen. Measurements were performed employing a MEKTON Vickers Hardness Tester with a diamond pyramidal indenter. Three measurements were taken at randomly selected points with a load of 300 g applied for 30 s.

In Taguchi technique, experimental analysis is based on orthogonal array (OA). It is the shortest possible matrix of combinations in which all the parameters vary at the same time and their effect and performance interactions are studied simultaneously. The selection of an appropriate orthogonal array depends on the total degrees of freedom (DOF) required [_{9} (3^{3}) standard orthogonal array is considered in determining the effect of three process parameters. Thus, the array has three columns and nine rows of three levels. The number of experiments required can be reduced to nine, which in classical combination method using full factorial experimentation would require 3^{3} = 27 experiments to capture the influencing parameters. An L_{9} standard OA is shown in

In this study, the optimal squeeze casting parameters of 2017 A wrought aluminium alloy have been specified through the Taguchi method, and the obtained results are acceptable for the ranges of squeeze parameters that have been selected in the present investigation. According to the results, the following conclusions can be drawn:

The combination A3 B1 C1 that means squeeze pressure 90 MPa, melt temperature 700 °C and die temperature 200 °C are recommended to obtain higher mechanical properties in squeeze casting of 2017 A Al alloy.

Squeeze pressure, melt temperature and die temperature were identified as significant process parameters from ANOVA. It is noted that the contribution of squeeze pressure is a larger for the UTS and hardness.

From the S/N ratio, it was evident that the combination of parameters and their levels A3 B1 C3 yield the optimum mechanical properties with minimum distinction about the target value.

The refinement of microstructure was the main reason for increasing the mechanical properties of the squeeze cast specimens.

The authors would like to acknowledge Jean-Pierre Lecompte for carrying out the experimental work at “Higher Engineering School of Limoges” (ENSIL), France. One of the authors, Najib Souissi would like to acknowledge very much, the Tunisian Ministry of Higher Education and Scientific Research for its financial support.

The authors declare no conflict of interest.