Overview in PDF format - Tallinna Tehnikaülikool

2. MODELS FOR MONITORING TECHNOLOGICAL PROCESSES2.1. Process Modelling in Machining UnitTechnologically, machining as a process has made a significant progress. However,lack of suitable cutting models is a main bottleneck in developing of machine toolsusing adaptive control. It is rather difficult to predict vibration, tool wear andbreakage, thermal deformation of the machine tools, and similar processes based onevents using off-line theoretical models (Altintas, 1999). Drawbacks of the prior artmodels are unsuitability for complex cutting path with changing diameter (e.g.stepping shafts) and complex cutting contour.In turning operations, tool vibrations influence both product quality andproductivity and may also have a negative influence on the working environment(Hakansson, 1999). The effect of lathe vibration on the roughness of machinedsurface is considered in this work. Dynamical phenomena concerned with thevibration are caused from external factors on the strained system of the lathe.During the machining of material all disturbances finally lead to relativedisplacements of the cutter and the blank. It allows to link parameters of the surfaceroughness with the relative vibrodisplacements of the cutter and the blank(Gaponkin, 1995).Integration of different sensors and measuring instruments into a uniform systemis a problem of the modern technological machinery (Russo, 1999). Data collectedby contact and non-contact measuring means should be managed in one system.Solutions capable of data mining can be achieved, in theory, by SOM (Self-Oriented Maps) programme, for example in MatLab software environment(Vesanto, 2002). By stimulating cooperation between different manufacturers andresearch laboratories, synergies can be achieved to realise the abovementioned maincriteria successfully. However, achievement of this goal could become veryexpensive; therefore networking and cooperation are remarkable resources of acompany on the way to become more efficient (Riives, 2002). To solve thisproblem the idea about real time web based advisory system was initiated. The userinteracts with the control module, the role of which is to work out the searchstrategy on the basis of initial data given by user, via user’s interface (Papstel,1999). As a result, a knowledge supply chain is created.Tool wear as one of the important factors in machining has been investigated byusing different methods. Indirect methods involve noise recognition with amicrophone, cutting temperature monitoring with thermocouples, vibrationsmeasurement (speed or acceleration) with a modal analyser, surface roughnessmeasurement with a profilograph, work piece dimension changes measurement withgauges. Direct methods include tool wear monitoring with CCD cameras. Everymethod suffers from some drawback; however, combined monitoring could be asolution. In Fig. 2.1 a monitoring system covering all the aforementioned devicesfor tool condition monitoring is proposed.16