Reliability improvement of deformation tools with the Taguchi robust design

Abstract

Key Words: reliability improvement, design of experiments, Taguchi approach, renewal policy SUMMARY & CONCLUSIONSAs we are moving to more emerging global markets, one of the most important goals of a manufacturer is to improve the reliability of its products. While the reliability may be affected by many potentially factors, some factors are more important and that they have to be identified. The values of the significant factors that can improve reliability are also important to be recommended. Taguchi's robust design experiments provides an efficient way to achieve these goals and the concepts of the Taguchi's method in the context of the reliability improvement are emphasized at the beginning of the paper.The studies available in the literature demonstrate the importance of robust design experiments in improving products reliability. However, not so much work has been done on applying robust design experiments for reliability improvement of deformation tools used in automotive industry. This is mainly due to the difficulty to obtain the necessary data for experiments design. In this paper, the renewal policies are proposed to gather the data for experiments design and the minimization of the average maintenance cost rate was used as the criterion in formulating renewal policies.For this purpose, the reliability modeling of deformation tools used in automotive industry was performed by a goodness-of-fit test. A data acquisition system, composed by resistive strain gages-strain indicator-connector block-board acquisition-personal computer, was used for the identification of the tools failure. A case study illustrates the application of Taguchi's robust design paradigm.In conclusion, using robust designed experiments provides a proactive means by identifying important factors that affect product's reliability. These influential factors can then be set at levels that generate reliability improvement. The results of the case study demonstrate that the Taguchi approach of robust design is a method which can be successfully used in improving the reliability of deformation tools used in automotive industry.Abbreviations ARP -Age Replacement Policy BRP -Block Replacement Policy F(t) -Cumulative distribution function(cdf) R(t) -Reliability function m -Mean time-to-failure z(t) -Failure rate η -The signal-to-noise ratio

Experimental investigations of significant parameters of strain measurement employing Taguchi method

“…Design of experiments is successfully applied to improve the performance of the products to specification, by reducing sensitivity to uncertain or randomly varying factors. Among a number of designed experiments methods, Taguchi's robust strategy is one of the most effective in practice and have been applied in many studies [13]. The main focuses in this method are on the product parameter selection and optimization, to ensure that the key characteristics of the product can be easily set to the target value and the variance can be minimized.…”

“…It was shown that controlling and adjusting the noise factor is very costly. Thus, the objective of the Taguchi approach is to find the right settings of control factors to minimize the noise factors [13,14]. Taguchi analyzed the variation of experiments using different definitions of signal-to-noise ratio (S/N).…”

“…While, the reliability of a system can be affected by many parameters, it is valuable to identify some critical parameters, which are significantly influence the outcomes. Taguchi's robust design experiments facilitate these aims [12,13].…”

“…Through the strain measurement procedure, different tools and software are employed for the data collection, such as Data Acquisition system (DAQ) and Lab View software. It has been shown by Baban et al [13] that low accuracy, data lose and delay in presenting data can be…”

“…caused by the wiring length during the experiments as well as the physical dimension expand [13], however, the amount of errors has not been studied quantitatively. Therefore, the influence of length of wires can be considered and studied as an effective parameter through the measuring process.…”