OPTIMIZATION OF GATING SYSTEM PARAMETERS OF CI CONVEYOR BRACKET USING TAGUCHI AND COMPUTER AIDED SIMULATION TECHNIQUES

In sand casting method, process of making casting depends upon various factors like mould, gating system, type of metal etc.For a sound casting, it is required to optimize the casting parameters for a given geometry of casting, to remove the defects and improve the quality in casted products.In this paper Taguchi technique along with design of experiments method (DOE) and Computer Aided Casting Simulation Technique are combined for optimization of gating system parameters of a CI Conveyor Bracket to improve the Product Yield (PY) and Hardness (HRD) of the casting.The gating system related parameters considered are, Runner Upper Diameter (mm), Runner Lower Diameter (mm), Runner Height (mm), Ingate Height (mm), Pouring Time (sec). In the First half part of the work, Taguchi based L18 Orthogonal Array was generated for experimental purpose and analysis of result was done using S/N ratio, Minitab 17 Software and Analysis of Variance (ANOVA) to analyze effect of gating system parameters on Product Yield and Hardness. In the second half part of work, AutoCast simulation software was used for validation of the result. Parameters are also validated by implementing it in foundry.


Introduction
As looking towards history traditional casting process is running since 6000 years long and is a basis of the entire mechanical industry. However, it still faces various problems such as difficulty in quality improvement, less production rate, to improve energy efficiency, to reduce material consumption and the question of environmental protection; one of the reasons behind it is complex casting process and lack of theoretical guidance (C.M. Choudhari et al., 2014). Sand casting is one of the most simplest type of casting process and highly used for centuries since it accounts for largest tonnage of production of different types of castings and because of its economical facts and flexibility in production (Manikanda Prasath K. et al., 2014). Casting process has many quality affecting casting process parameters some of which are controllable and some are noise factors. Therefore Taguchi Technique is the good choice for quality improvement in castings. Taguchi's contributions have made the practitioner's work simpler by advocating use of very less number of experimental designs and providing clear understanding of variation nature and economic consequences of quality engineering in the manufacturing sector (S. Guharaja et al., 2006). Dr. Taguchi has introduced several new statistical concepts in the field of quality improvement which have proven to be estimable. Many Japanese manufacturers have improved their product and process quality with a great success. Taguchi has calculate on W.E. Deming's observation that only 15% to the worker and more i.e. 85% of the poor quality is attribute to the manufacturing process, out of which again many researchers has reported that 90% of the defects are due to incorrect design of Gating and risering system.

Process parameters of Green Sand castings
The process parameters can be listed in five categories as follows: 1.
Shake-out-related parameters An Ishikawa diagram (cause and effect diagram) was constructed as shown in Fig. 1to identify the casting process parameters that may influence green sand casting defects (S. Guharaja et al. 2006 (1999)has discussed how computer system has brought revolution in manufacturing sector. A scene of Past-present-future of computer-Aided system is described in this paper and has shown that how this system helped industries to directly improve the profit of production by reducing lead time and cost of production. He has generated software named AutoCAST which has became a fine tool for casting industry. So instead of going to do shop floor experiments in making First Article Approval, the software found to be very helpful in generating step by step casting process and gives final result in which we can find out whether defects will be present or not and after final result we can apply it in shop floor. In this way the software was found to be very helpful in saving time, cost, material wastage, etc.
From all these literatures, Taguchi method is found to be the best method for optimization of parameters.

Research Methodology
This research work is done at a small scale Industry where the methodology of work done is shown briefly in figure 3.1.

Examine the casting process at Deepshikha Industry
To study the gating system at Deepshikha Industry

Experimental Process
Experiments were performed at small scale foundry producing grey cast iron components.

Conducting the experiments
After assigning the parameters and their interactions in the columns of selected orthogonal array, the factors are assigned for each trail as shown in experimental array table 3. As per L18 orthogonal array all the 18 experimental trials were performed sequentially.

Analysis of experimental result
The castings of Conveyor Bracket were made against the trial conditions. After getting the castings first they are taken to the process of sand removing throughout the surface of casting along with gating system. Then the weight of casting with gating system was measured and noted down with trial sequence. Then all the castings were sent to remove the gating system and for finishing process. Again the weight measurement was done for all finished castings and was noted down in the same sequence. The Product Yield (PY) of casting was then calculated by using following formula diameter of circular impression generated on casting was measured using microscope. Then from the BHN chart hardness number for respective diameter was found out.
The values of product yield and hardness for respective trials are recorded in table 4.

For Product Yield
The casting Product Yield is the "Larger the better" type of quality characteristics. Larger the better S/N ratios were computed for each of the 18 trials and the values are recorded in table 5. Similarly S/N ratio for all other trials is calculated and also S/N ratio is calculated by using MINITAB 17 software which is found to be the same.

M ain Ef fects Plot (data means ) f or SN ratios
Signal-to-noi se: Lar ger i s better

For Hardness
Similarly as S/N analysis of Product Yield done above, the casting Hardness is the "Larger the better" type of quality characteristics. Larger the better S/N ratios were computed for each of the 18 trials and the values are recorded in table 7.

M ain Ef fects Plot (data means ) f or SN ratios
Signal-to-noi se: Lar ger i s better

Validation by Autocast Software
AutoCAST-X software is a fine tool for casting industry in which simulation of casting product can be done; which in result gives the report of Casting Yield, optimized value of gating system and the quality affecting defects like shrinkage porosity, blow holes formation, cold shut, Hard zone etc.
A solid model in STL format is imported in AutoCAST software. All the values of process parameters are entered in initial stage. Then gating system is designed by giving the dimensional input as the optimum values we get by experimentation i.e. PT = 8 sec, RUD = 45 mm, RLD = 30 mm, RH = 90 mm, IH = 8 mm. By feeding these values the casting model with gating system gets generated as shown in figure 5. The next stage of software work is, it checks the gating system for its effectiveness in complete filling of mould cavity. And which is done in this case as can be seen in red colored view of a completely filled model in figure 6. After completion of simulation process, software generates simulation report. From this report we get that, Product Yield value is found to be 90.91% and which is very near to the experimental value of 90.32% we get. So we can say that our optimal value of all parameters is valid. Report of software is shown in figure 8.   With Taguchi optimization method and experimental analysis we can conclude that for the Product yield to be more than 90% and Hardness to be more than 175 BHN can be obtained. §

Result and Discussion
Computer aided casting simulation technique is found to be the most effective and accurate tool to design the gating system and to find out the quality and yield of casting in very less time, without doing actual trials in foundry.