Fraunhofer Institute for Algorithms and Scientific Computing SCAIBAUEN
The predicted compressive strength of aerated concrete in the composition space is color-coded in this graph. The points correspond to known compositions.
Background
A sustainable circular economy is essential to achieve the goal of climate neutrality. To this end, the development of new innovative building materials is crucial. However, this development has so far been a very laborious and lengthy process. To support and significantly accelerate this process, experts from the Fraunhofer institutes IBP and SCAI develop new data-driven virtual techniques for material design at the Fraunhofer Center for Machine Learning.
Goals of the Project
For demonstration purposes, the development and validation of a data-driven model for predicting the properties of mineral building materials with a focus on aerated concrete was pursued. To do this, experimental data were first collected and processed. This data collection was then used to develop a model for predicting the properties of aerated concrete based on mechanical material parameters and physical data. To further improve the prediction quality, this model was then extended to include chemical material parameters for the first time. In addition, this model served as the basis for the development and implementation of a demonstrator for the inverse design of aerated concrete. Here, the properties of the desired aerated concrete can be specified and the demonstrator then determines suitable mix designs.
Summary and Outlook
Overall, the project has shown that data-driven virtual material design has a very high potential to significantly accelerate the development and optimization process of sustainable building materials. For the first time, a model for the inverse design of aerated concrete could be developed. For this purpose, the proposed compositions were also validated experimentally. The concept can also be implemented for other mineral building materials.
The project was funded internally by Fraunhofer as part of the Research Center for Machine Learning.
Project duration: 10/2020 – 12/2021