Specialists from Ogarev Mordovia State University have developed a self-compacting, non-shrink, fine-grained concrete that surpasses traditional heavy concretes by two to four times in terms of its useful characteristics. This concrete was created using artificial intelligence, specifically machine learning methods for designing and optimizing the composition. To achieve this, a database was collected and analyzed, including over a thousand experimental compositions of cement materials.
The developed concrete is a dry construction mix consisting of Portland cement, quartz sand, and a complex organic-mineral additive, including metallurgical production waste and sedimentary rocks. The mixture has the ability to self-compact and compensate for shrinkage deformations when water is added.
Thanks to its low permeability, high strength, and corrosion resistance, the service life of this concrete under normal operating conditions can range from 150 to 200 years. Its production cost is 10-15% lower than that of its counterparts, due to the reduced cement content of the composition and the use of local raw materials and industrial waste.
In addition, the machine learning methods used in the development allow for monitoring and analysis of the influence of various parameters on the characteristics of the concrete, with the possibility of adjusting the composition and improving its properties based on new data. Industries and industrial partners have shown interest in the new concrete and intelligent design systems.
Read materials on the topic:
Coffee Grounds Concrete Created in Rostov
Wallpaper, Tire Rubber, Bags, and Pipes: High-Molecular Polymers Complex Launched in Tula Region
