DEVELOPMENT OF A TECHNIQUE FOR GRAPHOANALYTICAL DESCRIPTION OF THE WORKPIECE OF A ROLLER-TYPE PART FOR CREATING A DIGITAL TWIN IN PRODUCTION PLANNING

Full abstract

Read the full abstract

The study objective is to develop a technique for a formalized description of a workpiece for the production based on a graphoanalytical representation of dimensional relationships between surfaces for the subsequent creation of a digital twin. The task to which the paper is devoted is to eliminate the problem of incompleteness of initial information and insufficient details of design decisions during production planning, related to the expert nature of decision-making and the high complexity of processing design data. Methods used and design and technological solutions: graph theory, system analysis of engineering processes, methods for formalizing the geometric structure of a part, sequential selection of elementary surfaces with their coding, construction of graphs of dimensional relationships with the introduction of dimensional characteristics, roughness parameters, shape deviations, relative position, database data and physical and mechanical properties of the surface layer. The novelty of the work: an original method of formalized description of a part as an object of production is proposed, which provides a visual representation of the relationships between surfaces in the form exceeding the information presented in the drawing; the graph model is adapted for use in an algorithmic environment and is focused on creating a digital twin of the workpiece. Results and conclusions: graph models are constructed for two variants of the "roller" workpiece (calibrated rolled products and hot stamping); it is shown that the model retains its structure when numerical size values change, but reacts to changes in sizing methods and design bases; the suitability of the method is found out for identifying technological heredity, substantiating the requirements for basic surfaces and automated selection of workpieces; the possibility of integrating the developed model into computer-aided process design systems is confirmed.