Man-Machine System of the Machining Area and the assessment of its reliability

Resumen

This paper contains the design of a man-machine system model for the machine tools based on a serial-parallel configuration and their component-subsystem relationships. Causal Analyses and a classification system using the Cluster Analysis were used to analyze the non-conforming pieces that occurred in 3 machine shops from the metal mechanical sector, and to design the man-machine system model. The man-machine system model proposes to provide adequate conditions to apply some statistical reliability assessment models with a serial-parallel systems approach, and on the assumptions of the constant and additive failure rate of the Exponential and Weibull functions respectively. As an input of the first statistical model, the times to failure from non-conforming pieces and goodness of fit test have been used to assess the reliability of component-subsystem relationships. For the second statistical model the reliabilities from component-subsystem relationships were used to assess the reliability of subsystems as serial systems configuration. Inputs for the third statistical model have used reliabilities from subsystems to assess the reliability of the man-machine system as a parallel system configuration. The man-machine system and statistical models were implemented and validated in a company from the metal-mechanical sector with a low-volume and high-mix production rate whose operations were carried out by both conventional and CNC machines. Based on the man-machine system analysis was possible to know the formal structure of machine tools that generate non-conforming pieces in the scrap area as well as to identify the subcauses of failure that affect the system (components). Components identified were: maintenance/set up, institutional conditions/attitude, admission profile, information/communication, calibration, and supplier. Also, the most common places where failures occurred (subsystems) were identified. Subsystems identified were: machine, tool, process/operation, measurement/instrumentation, programming/design, and material. Through the selected statistical models was possible to get reliability indices for each component-subsystem relationship, for each subsystem as well as for the whole man-machine system, identifying areas for improvement. The company where the methodology was implemented and validated presented a high volume of scrap in the machine tool and it was possible to identify that the subsystem: process/operation and the components:  maintenance/set up and institutional conditions/attitude had the lowest reliability, and they were considered as opportunities areas for improvement. For 7 hours the man-machine reliability is 0.1409 (the probability to find a product without failures for which it must be scrapped is 14.09%). In the machine shop, the issues that caused these low levels of reliability are attributed to the insufficient preventive and predictive maintenance program. Improvement actions are considered in another research. This kind of research investigates a little-studied problem, as it is about the reliability of the man-machine system in machining tools and investigates from an innovative perspective and preparing the ground for new studies in the metal mechanical sector.