Russian scientists have found a way to predict the strength of concrete in advance. Specialists from Don State Technical University (DSTU) have integrated a neural network: it analyzes the material, identifies potential defects, and predicts its strength with 89% accuracy. The development will help to check concrete quality faster, select suitable compositions, and reduce construction costs.
This refers to variatropic concrete – a material that is specially compacted in different ways during manufacturing, for example, by vibration, rotation in a centrifuge, or a combination thereof. Because of this, concrete within a single product can be uneven: in some places it is denser and stronger, in others – lighter. Scientists have created algorithms that take these differences into account and help to assess the strength of the finished structure in advance. This same development can also be used to check other building materials and products.
Scientists tested the technology in a series of experiments: they refined the equipment, manufactured samples, and selected the most successful production modes. Based on this data, new types of concrete are already being created, including those with microsilica and various fillers.
Specialists also compared three methods of mixture compaction: vibration, rotation in a centrifuge, and a combination of these methods. Vibrocentrifugation yielded the best result. Hollow concrete-filled steel columns manufactured in this way withstood loads 13% better than solid counterparts.
According to scientists' calculations, the new technology allows either to make structures 15% stronger, or to maintain the previous strength but use 15% less material. Assembly and installation can be 24% faster, and construction labor costs can decrease by up to 28%. At the same time, defects should become fewer, and the reliability and safety of buildings should increase by 12%.
In the future, DSTU scientists want to study how nano-additives change the structure and properties of concrete. This data will help to more accurately calculate the strength of reinforced concrete products. In perspective, the technology will allow for building more reliably, faster, and with less material consumption.