Research On Precision Grinding And Control Methods For Micro-holes

Machining parts with micro-holes play important role in many fields, but they also have seriouse negative effects by inner burr and sharp edge, even lead to accidents. Therefore, a huge of human and other resources have been invested in holes’ deburring and polishing, and some kinds of technologies have been developed. As quality requirements of product with micro-holes becomes higher and higher, existing technologies are hard to meet these requirements.Good result was achieved when abrasive suspension flow (ASF) was applied to micro-holes grinding. In the grinding process, a kind of grinding medium with low viscosity, good suspension property and liquidity was used. During the process of grinding, hydraulic oil and grinding medium were isolated, and the flow rate of oil which equals to that of grinding medium was measured, so that the flow rate change through micro-holes can be monitored and controlled online. Therefore, an excellent flow rate consistency of micro-holes parts can be obtained in single grinding process by using the ASF technology, which is hard to be copied by other technologies in principle.To implement micro-holes grinding by ASF, diesel engine needle valve bodies, a sort of representative part with micro-holes, were chosen as grinding objects. In the dissertation, online monitoring and data processing methods of flow rate, online real time grinding technique were studied and improved. In addition, expert system combined with intelligence algorithm was adopted to develop process model and optimize machining parameters. A grinding platform was built to perform various processing techonologies and the results were compared. Main research results are organized as follows:Necessity and fessibility of ASF grinding was analyzed, and advantage of grinding medium was studied. The machining parameters and evaluation targets of machining were analyzed which are basis of flow rate monitoring and control during grinding in the thesis.Hydraulic working model of the platform was simulated by power bond graphs, and feasibility of platform was analyzed. The platform adopts PID to stabilize system’s hydraulic pressure to ensure accuracy of flow rate. During the grinding process, flow rate is very small and constantly changing, and requirement of real-time calculation is very high, based on which the flow rate processing algorithm was improved. Different flow rates in different pressure were converted to equivalent flow rate in a certain pressure to guarantee the comparability flow rate.The shortcoming of present ASF technology was analyzed and a new reference process curve was constructed, and further the combined control method was introduced for online real-time control. At earlier steps in the grinding process, fuzzy control was adopted to grind parts with micro-holes, while at later steps, parts were grinded under fixed pressure to ensure the same flow rate between process and measurement on platform. During the period of fuzzy control, increasing of reference flow rate and difference of pressure were fuzzifiered by nonlinear quantization method, while error of flow rate increasing, error changing rate of flow rate increasing and increasing of output pressure were fuzzifiered by variable universe. For reducing the cost and improve effectiveness of computaion, stretch factor was calculated based on reference of flow increasing rate and difference of pressure.After a certain time, grinding ability of grinding medium becomes weaker, data and experience of experts need to be stored to service for later production precess, the grinding device needs overhauls and breakdown maintenance often. To solve all these problems, an expert system was built. The expert system in this dissertation primarily discusses machining modeling and parameters optimizing. During parameters optimizing period, expert system extracts existing knowledge to make the inference for machining process. If it is not successful, fuzzy neural network is employed to train and model and PSO is used for multi-object optimization to find the optimal parameters.Grinding platform for micro-holes parts production was realized and grinding results in different conditions were compared. The compare proved that theories of the dissertation are good for flow rate consistency, grinding efficiency and data stability.