Elsevier

CIRP Annals

Volume 64, Issue 1, 2015, Pages 431-434
CIRP Annals

Production quality performance in manufacturing systems processing deteriorating products

https://doi.org/10.1016/j.cirp.2015.04.122Get rights and content

Abstract

In several manufacturing contexts including food industry, semiconductor manufacturing, and polymers forming, the product quality deteriorates during production by prolonged exposure to the air caused by excessive lead times. Buffers increase the system throughput while also increasing the production lead time, consequently affecting the product quality. This paper proposes a theory and methodology to predict the lead time distribution in multi-stage manufacturing systems with unreliable machines. The method allows to optimally set inventory levels to achieve target production quality performance in these systems. The industrial benefits are demonstrated in a real manufacturing system producing micro-catheters for medical applications.

Section snippets

Introduction, motivation and objectives

Production quality has been proposed recently as an emerging paradigm to achieve desired service levels of conforming products in advanced manufacturing systems, by simultaneously considering quality and productivity requirements [1]. With respect to this background, the importance of an integrated analysis of production logistics, product quality and equipment maintenance to achieve balanced manufacturing system solutions has been pointed out. This problem is particularly relevant in

System description

The considered system is formed by K manufacturing stages and K − 1 buffers of finite capacity, configured in serial layout (Fig. 1). Stages are denoted as Mk, with k = 1, …, K, and buffers are denoted as Bk, with k = 1, …, K − 1. The capacity of buffer Bk is Nk, that is an integer number. Finite capacity buffers can either model physical conveyors or the implementation of token-based production control rules, such as kanban, regulating the material flow release at each stage [10].

Single stage model.

Lead time distribution evaluation

In this section, an efficient and exact analytical method to compute the distribution of the lead time in the critical portion of the system is described. The rationale of the approach is explained in the following. Firstly, the probability that, at the moment when a randomly selected part enters buffer Be, the system is in a given state is derived. Secondly, the probability that, given that state of the system as initial condition, the last part in buffer Be crosses the critical portion of the

Numerical results and system behavior

In the first experiment, the distribution of the lead time in multi-stage manufacturing systems is investigated. A production line with five stages is considered with parameters reported in Table 1. The stages correspond to machines with one operational state and one failure state, with failure probability p = 1/MTTF and repair probability r = 1/MTTR (MTTF is Mean Time to Failure and MTTR is Mean Time to Repair). The machines produce one part per time unit if operational. The results are reported

Real case study

The proposed approach has been applied to the production of micro-catheters as high value medical products for the aging society in the medical technology sector at ENKI s.r.l. [13]. The manufacturing process is composed of three main phases: (i) material compound preparation and control, the (ii) micro-extrusion of the micro-tubes and (iii) final micro-catheter assembly. Defects are mainly geometrical, generated within the micro-extrusion process. The above defects lead to extremely high

Conclusions

This paper proposes a modeling framework and a methodology to predict the lead time distribution in multi-stage manufacturing systems. The lead time distribution is used to compute the system production quality performance, when products deteriorate. The theoretical work developed in this paper opens the way for applications in several different fields. For example, supply chain management approaches can benefit from the availability of a solution to the problem of deriving the lead time

Acknowledgements

The authors would like to thank Eng. Moreno Camanzi and Mr. Mario Di Cecio from ENKI s.r.l. for the support in this research.

References (13)

There are more references available in the full text version of this article.

Cited by (0)

View full text
Copyright © 2015 CIRP. Published by Elsevier Ltd. All rights reserved.