Problem Definition
The occurrence of variation in bore-diameter of B.G collar components is a recurrent multi-response optimization issue. In a boring operation, the B.G collar-component dimension of a roller-bearing superseded tolerance.
Objective
It is hoped to reduce to minimal the bore-diameter variation of the B.G collar component. The collar component has specifications of 145
0.79. This variable problem will be tacked through a nominal goal.Factors and Levels Identification
Response
The problem constitutes multi-response and two-response optimization dimension -variation machining-time.
Analysis of data
The approach for obtaining multi-response optimization follows:
- Transforming experimental data into signals to noise-ratio in line with Taguchi-quadratic-loss-function method;
- Normalize S/N -ratio for objectives;
- Collect multi-response S/N (MRSN) for each experiment through appropriate response weightages.
Results and Predicting the Response
Following average-effect-response-values, main effects are derivable. An increment in MRSN-ratio infers a more accepted quality. For this work, Factors A, C are assigned level-1 and factor-B is assigned 3 for optimizing process parameters and response prediction is at 95.0% confidence intervals for both responses.
Validation experiment
To achieve validation, Fuzzy-logic has been produced from Fuzzy-set-theory with the use of MATLAB-7.1. Optimum-values show nominal-the-better constituting appropriate factor settings. Optimal-values is then obtained using Taguchi-design-and-Fuzzy model.
Recommendations and conclusions
A well-defined approach is hereby suggested for achieving optimization of multi-response-problems using Taguchi method. The development of loss-function here is appropriate for the improvement of multi-response characteristics in boring operations and is applicable to all sorts of multiple-response models which deal with multiple-characteristics.