DOI: 10.7763/IPEDR. 2014. V75. 29 Estimation the System Reliability using Weibull Distribution D. R. Dolas 1 + , M.D. Jaybhaye 2 , S. D. Deshmukh 3 1 PhD Student , MGM’S JNEC Aurangabad , Aurangabad-431003 India 2 Associate Professor in Production Engineering Dept, COEP, Pune-411005, India. 3 Principal MGM’S JNEC Aurangabad - Aurangabad, 431003, India Abstract. An increasing complexity of systems often leads to an increase of failure mechanisms. A statistical analysis of the lifetime of systems with several failure mechanisms consisting of several sub- components considered The Weibull distribution is commonly used as a lifetime distribution in reliability applications. The two-parameter Weibull distribution can represent a decreasing, constant or increasing failure rate. This paper presented the estimation of system reliability using two parameter Weibull distributions. The parameters are estimated using Weibull probability plot & statistical analysis and the results being presented in charts. Computation is made using ‘Windchill quality solution 10.1 Tryout’ software. Keywords: Reliability, Weibull distribution, WPP Method. 1. Introduction Weibull distribution is named after Walodi Weibull (1887 – 1979). It is very flexible and can through an appropriate choice of parameters and model many types of failure rate behaviors. This distribution can be found with two or three parameters; scale, shape and location parameters. There are a number of methods for estimating the values of these parameters; some are graphical and others are analytical. Graphical methods include Weibull probability plotting and hazard plot. These methods are not very accurate but they are relatively fast. The analytical methods include maximum likelihood method, least square method and method of moments. These methods are considered as more accurate and reliable compared to the graphical method. In this article an attempt is made to estimate the system reliability using two parameter Weibull distributions & Computation is made using ‘Windchill quality solution 10.1 Tryout’ Software. The Maximum likelihood and median rank regression methods are used by researchers to estimate parameters of Weibull distribution [1]. Parameter estimation method for the machine tool reliability analysis to overcome the problem of unavailability of a well-defined failure data collection mechanism was given by Lad et al.[2]. It uses the knowledge and experience of maintenance personnel to obtain the parameters of lifetime distribution of the repairable as well as non-repairable components subassemblies. The Weibull distribution is the standard function used by the wind energy community to model the wind speed frequency distribution and compare the methods [3]. Estimation the reliability function using the Maximum likelihood Method & Weighted Least Square Method & simulation procedure using Monte Carlo Method are used and several experiments are implemented to find the best estimators which have smallest mean square error [4]. Monte Carlo simulations are used to investigate the correlation between system complexity and component lifetime distributions. [5].Moments, maximum likelihood and least squares are compared & used the means square error & total deviation, as measurement for the comparison between these methods [6]. Weibull technique is used for the reliability analysis & the Pareto analysis is carried out. The spare parts optimization was also carried out for a few vital components of this wind farm and the results are presented [7]. Gear box assembly analysis and the failure data in various operating conditions was taken from the logbooks of the vehicles. For modeling purposes the Weibull distribution has been chosen & result will be useful to the maintenance engineer to find the unreliability of the gear box assemblies and also for future strategic decision making [8]. Paritosh Bhattacharya presented the analytical methods for estimation of parameters [9]. + Corresponding author. Tel.: + 91 9422655271; fax: +0240- 2482893. E-mail address: dhananjay_dolas@rediffmail.com 144