Experimental Approbation of Intellectual System for Machining Accuracy Control

Provision of machining accuracy is a relevant objective in technology of mechanical engineering and its solution allows to guarantee an operational accuracy of mechanisms and machines, their wear resistance, reliability and durability. In order to solve the given task a method has been proposed in the paper that permits to ensure the highest machining accuracy margin on the basis of multiple-factor optimization of parameters for technological process while using methods of artificial intelligence. To ensure the machining accuracy by point tools, an intellectual system has been developed and it is based on technologies of functional semantic networks. For revealing correctness of SEMANTIC intellectual system operation an experimental inspection of its working capacity has been carried out in the paper. The paper contains description of the methodology for In article the technique of experimental investigations and their purpose is to make a comparative analysis of actual errors in opening machining with the machining errors predicted by the SEMANTIC system on the basis of functional semantic network application. The investigations have made it possible to determine dependences of axial misalignment in the machined openings on tool advance and its rotation frequency while making openings by high-speed steel drills. The paper presents analysis results of limiting and probable values for components of a total machining error. It has been shown that while making assessment of machining accuracy it is necessary to consider probabilistic nature of occurrence of total error components that are setting upper boundary as it is realized in the described intellectual system. The semantic network permits to compare errors in arrangement of opening axes which are machined in accordance with admissible drill advance and rotation frequency and corresponding experimental values. The experimental investigations prove the possibility to use an approach for forecasting a total machining error which is based on application of functional semantic network technology.