DEVELOPING A CHIP CONTROL AND MANAGEMENT SYSTEM

Abstract

With the rapid development of high-degree automated production, unacceptable chips lead to some additional disadvantages, such as the reduced reliability of machining processes and decreased productivity of high-cost machining equipment. As a result, the significance of chip control is increasingly emphasised. Therefore, the aim of this project is to develop a chip control and management system to predict chip breaking, the trend of chip breaking or/and the extent of chip breaking, and to guide users for the selection of machining condition to achieve broken chips. As a portion of the system, a three-dimensional chip breaking model is developed, based on a set of introduced equivalent parameters. The mechanism of the effect of the machining parameters on chip breaking has been analysed with the help of these equivalent parameters. Meanwhile, a criterion to assess chip breaking and its trend is proposed, using the ratio of the strain in the chip to the chip fracture strain. In the verification of the model, it can be concluded that the predictive data from the model and the experimental results are in reasonable agreement. Moreover, the predictive data from the model should not only provide the boundary deciding whether or not chips are broken, but also display the trend of chip breaking according to the magnitude of the ratio. In order to extend the applicable range of the model, more effective parameters dealing with cutting tools themselves were studied, such as side cutting edge angles, through chip grooves and wavy cutting edges. For the sake of reliability, experiments were performed to verify the new model and reasonable agreement was achieved. Especially for the side cutting edge angle, the effect of it is different when machining with tool inserts with or without grooves. In the thesis, a new chip classification code is presented according to five chip shapes and ten chip sizes, based on the experimental results. Furthermore, a parameter called Chip Packing Density Index (CPDI) is proposed to assess chip breaking and its extent, which also reflects the ease of chip disposal. The greater the chip breaking ability, the greater the chip disposal ability. CPDI shows more definite information, compared to the ratio of the strain in a chip to the chip fracture strain s/sr. To simplify the experimental results, Fractional Factorial Analysis (FF A) is applied to test the significance of various machining parameters. The achieved outcome is statistically reliable without the need for replicate experiments. Because it is proved that the influence of the interactions being more than two-factors on CPDI is small enough, these interactions are proposed to be an estimate of errors in FFA. Only main factors and two-factor interactions appear significant. Combining the chip breaking model and the condensed experimental results by FFA, the knowledge about chip breaking has been extended further. However, this is still not enough in view of practical machining processes. Therefore, the knowledge is broadened by available chip breaking data provided by tool insert manufacturers. Finally, a chip control management system is established, based on the model and the experiment as well as the available chip breaking data. Because all the available knowledge about chip breaking partially overlaps each other, its combination will naturally widen the applicable range of the system for practical machining processes. The independent representation and reasoning of knowledge base provide the flexibility for the system to be further developed, if new knowledge about chip breaking is available. The structure of representation and reasoning for this system is introduced and analysed, using "If-Then-Do", or assertion-hypothesis-action. To understand this chip breaking management system, a typical case is studied. The main objective of the system focuses on the prediction of chip breaking, but the amount of information included in the three resources of knowledge is different. The available data contain nothing but information concerning whether or not chips are broken. In addition to chip breaking, the model can predict the trend of chip breaking in broken chip zones. Furthermore, the experimental results can be used to predict the extent of chip breaking. In the knowledge island, the agenda can be shifted from one to another, normally following the sequence - the experimental results, the model and then the available data, to get more information. On the other hand, the system can automatically search the knowledge base both in forward chaining and in backward chaining. If broken chips are achieved, the investigation will be discontinued automatically. Otherwise, the system will prompt machining parameters to be determined by users. Once users determine a parameter to be revised, they can select either the manual or the automatic way to adjust the value of the parameter until the broken chips are achieved. The chip control and management system combines the knowledge of the developed three-dimensional chip breaking model, the condensed experimental results and the available chip breaking data together. This is because the three aspects of the knowledge are partially overlapped so that the range in which the system works effectively are extended. The system can be applied to predict chip breaking, the trend of chip breaking and/or the extent of chip breaking in a given machining condition. On the other hand, the design of the system provides a great potential ability for the system due to the independent knowledge representation and reasoning strategy, to cater for the rapid development of the chip control technology, especially the continuous introduction of new chip former configuration in market. The independent characteristics of the system makes the system easily improved if new findings are achieved.