The aim of the project is to develop a technology for casting large-size, cored thin-walled six-bladed segments for the low-pressure turbine steering apparatus of the GP7000 jet engine. Currently, this type of component is made by only one company in the world: the Alcoa Howmet.
In order to meet the requirements for the production of this type of advanced components, it is necessary to develop new technological solutions and improve individual stages of the casting process, this being effected through using new materials developed within the project, such as new casting waxes, ceramic mixtures and cores.
The INFOSTER sp. z o.o. together with the WSK “PZL-Rzeszów” S.A. are working on Task WP10: “Process data management (Intelligent systems)”, which is defined as The system of assisting design and technological test implementation. Knowledge base.
The research work is expected to yield the Expert System for Designing and Assisting Technological Trials and Process Monitoring, founded on a Knowledge Base. Research will be carried out on the development of individual modules of the Expert System for Design and Assisting Technological Trials and Process Monitoring, these including:
The system can be divided into three logical layers. The first one relates to the acquisition of process data coming from production-line, from laboratory trials or from bench tests off-line. The second one relates to planning experiments, their supervision, data analysis and inferencing. The third one deals with data archiving.
How the system operates will be verified during technological trials at the launch of a new product.
An essential element of the system is the identification of production unit, in the form of a single blade, the successive individual processes involving the creation of the wax model, positioning of the mould during the wax smelting process, mould annealing and the casting process.
The data analysis module is meant to ensure processing of the data until the expected form is reached, to select key information items, e.g. the maximum value of a feature extracted from the results of its distribution measured over a period of time. The module will also provide visualisations, produce statistical models, enable selection of coefficient values in physical-chemical models and will assist construction of models based on regression and correlation analysis.