Krasikov AntonNaumenko VladyslavBilyi VladyslavVasiliev OleksandrZgalat-Lozynskyi Ostap2024-12-052024-11https://archive.ipms.kyiv.ua/handle/123456789/99Sharing/Access Information Non-public. For cooperation and access, contact a.krasikov@ipms.kyiv.ua Data & File Overview Robocasting is a multi-stage 3D printing method that allows to print ceramic products of complex shapes. Despite the fact that the method itself is quite simple, at each stage it is necessary to take into account the influence of a large number of parameters. The file named "cleaned_robocasting_df" contains a set of experimental data that correlates the printing parameters with the characteristics of the printed products. Methodological Information The data was collected using an Ender 5 printer with a Stoneflower 2.0 ceramic equipment. and a separate control panel. In this printer, the paste (60 wt. % MoSi2 powder and 40 wt. % plasticizer) is extruded through nozzles with a diameter of 3mm to 0.4mm through a screw mechanism into which the ceramic paste is fed. As a programming language g-code was used. During each print, we obtained 16 samples. Also, an additional strip was printed at the beginning of the print to stabilize the extrusion.This dataset is designed for training machine learning models to optimize the robocasting method for MoSi₂-based ceramic products. Robocasting is a multi-stage 3D printing process that enables the fabrication of ceramic products with complex geometries. Despite its simplicity, each stage requires careful consideration of numerous parameters to ensure product quality. The dataset contains experimental data correlating printing parameters with the characteristics of the printed products. The data was collected using an Ender 5 printer equipped with Stoneflower 2.0 ceramic printing equipment and a custom control panel. The printing process utilized a ceramic paste comprising 60 wt.% MoSi₂ powder and 40 wt.% plasticizer, extruded through nozzles with diameters ranging from 3 mm to 0.4 mm. Variables include ambient temperature, humidity, desired and actual layer dimensions, nozzle speed, and extrusion multiplier. Each print produced 16 samples and an additional strip to stabilize extrusion. Number of the agreement under which the financing is provided: No. M/19-2024 of 16.05.2024 The dataset file, titled "cleaned_robocasting_df", is not publicly available for download. For access and collaboration, please contact a.krasikov@ipms.kyiv.ua.en-USRobocastingMachine learningMoSi₂ ceramics3D printingDataset