A UK startup has launched a unique Artificial Intelligence (AI) toolset that can train deep neural networks from sparse and noisy real-world experimental data. Several proven applications have been demonstrated in alloy design. The technology brings all the available data together and uses underlying correlations to accurately predict missing values and generate the most complete models possible. Applying this novel method to the available historical and simulated data, enables organisations to identify opportunities for reducing costs and downtime, time savings and overall performance improvements.
Despite the central importance of materials in enabling new technologies, historically the only way to develop new materials has been through experiment driven trial and improvement. This suggests that for example commercially available superalloys are the outcome of several years of empirical research and development. Even though these superalloys have good properties, they do not necessarily have the optimal balance of properties required for specific engineering applications.
The new AI tool incorporates uncertainty to allow alloys to be designed with the greatest probability of meeting a design specification containing many different material properties. It combines experimental data with computational thermodynamic predictions to rapidly, reliably, and robustly identify the alloy composition that is most likely to meet a multi-criterion specification. The tool was used to propose a new nickel-base superalloy alloy most likely to simultaneously fulfill 11 different physical criteria. The tool predicted that the new nickel-base alloy offered an ideal compromise between its properties for disc applications and seven of these properties were experimentally verified, demonstrating that it had better yield stress and oxidation resistance than commercially available alternatives. The figure depicts a relief plot showing how multiple property targets were satisfied simultaneously. The capability to quickly discover materials computationally using the AI tool will empower engineers to rapidly optimise bespoke materials for specific applications, bringing materials into the heart of the design process. The tool has also been used to design a nickel-base alloy for a combustor liner, and two Mo-based alloys for forging tools.
The company is delivering unprecedented solutions in materials including alloys, superalloys for aerospace and automotive sectors enabling organisations to design new formulations that meet their target criteria. The validation metrics, outliers, and confidence levels that the predictive model outputs guide where further testing is needed and allow the correct identification of the next best experiment that will yield the greatest insights.
The company is interested in commercial agreements with technical assistance, and technical cooperation including under European projects. The partners would typically have a need for a new material, chemical or any other scenarios where sparsity and inhomogeneity of data is a problem.