Company representatives will attend the Spain Congress Foundry in Bilbao on September 26 to present the latest AI projects implemented in the company’s process area. Jose Antonio Molina, industrial director, and Iñigo Fresán, maintenance and programming engineer, will explain several improvement initiatives implemented in recent years in different phases of the production chain.
Spain Foundry Congress is the event of reference for foundry professionals at national and international level and brings together every year the most relevant professionals in the sector. This event brings together a powerful national industry, since Spain is the second largest automobile manufacturing country in Europe after Germany.
Defectability improvement project
One of the most relevant projects to be shared by TIF representatives will be the one carried out in the last two years to reduce the plant defectibility due to the variability of the conditions of the sand used in the molding process.
Specifically, at TIF more than 3,000 tons of sand are prepared every day in a closed circuit that is then used to produce the molds on which the molten iron is poured in the casting phase. It is therefore critical that this sand respects precise conditions of humidity, shear strength, etc., in order to control the quality of the parts.
In fact, the variability of conditions of the sand used in the molding process gave rise to batches with very different reject rates and was the main reason for defectibility in the plant. In response to this problem, TIF formed a multidisciplinary work team with people from the areas of maintenance, quality, production and manufacturing engineering.
This team analyzed the main variables and properties of the sand that caused the defective, generated a predictive AI model based on time series from the historical data of the last decade and designed a digital twin to predict how the mills in the sand area behave. In this way, a process that was previously managed semi-manually has been automated, avoiding human error and ensuring continuous and automated monitoring of the sand plant.
Thus, continuous control of critical parameters such as humidity and shear strength, which the production team can modify by adding water and other elements, has been achieved. Thanks to this project, the internal and external defective has been reduced by 60% and the work team has been trained in the potential and possible applications of AI. In fact, new projects are already planned for the implementation of this technology in new applications in the future, including one related to planification project and another for the improvement and optimization of the cupola, so that it takes into account the materials to be melted and the customer’s requirements.
