Metal alloys are mixtures of metals (or sometimes one metal with other elements). Metals have good properties regarding industrial applications. Those include high strength, resistance to corrosion, hardness, electrical conductivity, etc. Nevertheless, people have managed to combine those metals forming alloys with even better properties. Alloys are currently utilized in many industries including the construction industry, technology industry, aerospace industry etc.
The metals and elements that are used to create the alloys define their properties. As a result, their behavior is highly dependent on the boundaries between the crystalline grains that comprise the final mixture. These boundary zones play a significant role when it comes to the total performance of the alloy and determine the aforementioned properties. However, scientists have not fully understood how these boundary zones behave. According to Christopher Schuh, co-author of the study and a Professor of materials science and engineering at MIT, the atoms in a single metal are arranged in a manner that their behavior can be readily predicted, however, in polycrystalline metals, crystals are "smashed together" and there are numerous "possible atomic arrangements" that may emerge. Thus, boundaries with unknown properties are created.
According to the study, recently published in Nature Communications, a new method that can provide insights into these boundary zones has evolved. The methodology utilizes both computer simulations and a machine learning procedure.
The team assessed more than 200 possibilities of combining a base and an alloying metal based on existing literature and they simulated via computer models their boundary zones. Subsequently, they utilized a machine learning technique to make predictions which were then validated. The results showed that the predictions were in accordance with the existing data.
The method is capable of predicting which combinations of metals can produce an alloy with enhanced properties. On the contrary, many combinations that do not generate alloys of good performance were detected. This provides great help to scientists who want to experiment on mixing two or more metals. The team suggests that the possible combinations of metals are numerous, therefore, ruling out a great proportion of them is highly beneficial. “So, you take the periodic table and you cross it with itself, and you would check every possible combination. For most of those combinations, basic data are not yet available...,” Prof. Schuh, added.
Once more data are available, the new method can improve the existing database. This will aid in the field of designing alloys with optimum characteristics.