This paper presented a new approach to decision making support of defects assessment in metal matrix composites (MMC). It is a continuation of the authors’ papers in terms of a uniform method of casting defects assessment. The idea of this paper was to design an open-access application (follow-up system called Open Atlas of Casting Defects (OACD)) in the area of industry and science. This a new solution makes it possible to quickly identify defect types considering the new classification of casting defects. This classification complements a classical approach by adding a casting defect group called structure defects, which is especially important for metal matrix composites. In the paper, an application structure, and the possibility of its use in casting defects assessment were introduced.

MATEC Web of Conferences, open access proceedings in Materials science, Engineering and Chemistry

MATEC Web of Conferences, open access proceedings in Materials science, Engineering and Chemistry

The paper presents an example of Instance-Based Learning using a supervised classification method of predicting selected ductile cast iron castings defects. The test used the algorithm of k-nearest neighbours, which was implemented in the authors’ computer application. To ensure its proper work it is necessary to have historical data of casting parameter values registered during casting processes in a foundry (mould sand, pouring process, chemical composition) as well as the percentage share of defective castings (unrepairable casting defects). The result of an algorithm is a report with five most possible scenarios in terms of occurrence of a cast iron casting defects and their quantity and occurrence percentage in the casts series. During the algorithm testing, weights were adjusted for independent variables involved in the dependent variables learning process. The algorithms used to process numerous data sets should be characterized by high efficiency, which should be a priority when designing applications to be implemented in industry. As it turns out in the presented mathematical instance-based learning, the best quality of fit occurs for specific values of accepted weights (set #5) for number k = 5 nearest neighbours and taking into account the search criterion according to “product index”.

The paper presents methodology of production control using statistical methods in foundry conditions, using the automatic DISAMATIC molding line. The authors were inspired by many years of experience in implementing IT tools for foundries. The authors noticed that there is a lack of basic IT tools dedicated to specific casting processes, that would greatly facilitate their oversight and thus improve the quality of manufactured products. More and more systems are installed in the ERP or CAx area, but they integrate processes only partially, mainly in the area of technology design and business management from finance and control. Monitoring of foundry processes can generate a large amount of process-related data. This is particularly noticeable in automated processes. An example is the modern DISAMATIC molding line, which integrates several casting processes, such as mold preparation, assembly, pouring or shake out. The authors proposed a methodology that supports the control of the above-mentioned foundry processes using statistical methods. Such an approach can be successfully used, for example, during periodic external audits. The mentioned methodology in the innovative DISAM-ProdC computer tool was implemented.