This chapter is dedicated to the reliability and maintenance of assets that are characterized by a degradation process. The item state is related to a degradation mechanism that represents the unit-to-unit variability and time-varying dynamics of systems. The maintenance scheduling has to be updated considering the degradation history of each item. The research method relies on the updating process of the reliability of a specific asset. Given a degradation process and costs for preventive/corrective maintenance actions, an optimal inspection time is obtained. At this time, the degradation level is measured and a prediction of the degradation is conducted to obtain the next inspection time. A decision criterion is established to decide whether the maintenance action should take place at the current time or postpone. Consequently, there is an optimal number of inspections that allows to extend the useful life of an asset before performing the preventive maintenance action. A numerical case study involving a non-stationary Wiener-based degradation process is proposed as an illustration of the methodology. The results showed that the expected cost per unit of time considering the adaptive maintenance strategy is lower than the expected cost per unit of time obtained for other maintenance policies.

No abstract

No abstract

The important portion of machining costs associated with cutting inserts and scraps induces the search for better effectiveness in turning. This paper presents the results of an exploratory study on the influence of tool flank wear on roughness indicators (arithmetic average roughness, root mean square roughness and maximum height of the roughness profile). The objective is to determine which of these indicators is best correlated with the cutting tool flank wear. In order to do this, specimens of AISI 1045 are machined until the end of life of a cutting insert. Significant, strong and positive correlations are found between all three roughness indicators and the tool flank wear. The most significant correlation is found with the arithmetic average roughness and the root mean square roughness of the profile. The choice of the end-of-life criteria for cutting inserts in industrial contexts is also discussed.