MODEL PENGGANTIAN SISTEM DENGAN MEMPERTIMBANGKAN GARANSI UNTUK HORIZON WAKTU TERBATAS

ABSTRACT <br /> Most systems (industrial equipments) are sold with warranty and, the warranty period offered tends to be longer. As a result, warranty can affect maintenance costs bourned by the users. Research that links preventive replacement with warranty are very few and all of those...

Full description

Saved in:
Bibliographic Details
Main Author: WAYAN SULETRA, I
Format: Theses
Language:Indonesia
Online Access:https://digilib.itb.ac.id/gdl/view/1740
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Institut Teknologi Bandung
Language: Indonesia
Description
Summary:ABSTRACT <br /> Most systems (industrial equipments) are sold with warranty and, the warranty period offered tends to be longer. As a result, warranty can affect maintenance costs bourned by the users. Research that links preventive replacement with warranty are very few and all of those works are for an infinite time horizon case. <br /> This study proposes new preventive replacement models based on age replacement model for finite time horizon for systems sold with warranty. Upon the arrival of a new technology, the users replace the current technology with a new one. Hence, the time horizon of the system starts at <br /> time t=0 and terminates at the arrival of a new technology. For this reason, the time between arrival of a new technology is considered as a time horizon. The objective system is single item and repairable with warranty periode w. Further, the time between failure is assumed to be Weibull <br /> distributed with parameters a and R. A minimal repair action is taken whenever the system fails. We consider two cases. In the first case, the time between arrival of the new technology is assumed to be deterministic, and in the second case, it is considered as a random variable. During the time horizon, preventive replacement by an identical new system is performed when the age of the objective system reaches (w+T), where T is a decision variable to be optimized. T* optimum is obtained by minimizing the expected total cost bourned by the user for a finite time horizon. By using the analytical approach, it has been shown that there is finite and unique solution if failure is charaterized by Weibull distribution and 1FR (increasing failure rate) property. Further, numerical examples are presented to illustrate model solutions. The results show that models considering warranty (our models) have a shorter optimum replacement <br /> i nterval than those of without considering warranty.