MIT engineers have achieved a remarkable breakthrough in materials science by harnessing the power of machine learning to create an incredibly strong 3D printable aluminum alloy. This innovative alloy not only outperforms traditional aluminum but is also capable of withstanding high temperatures, making it five times stronger than what we typically find in aluminum manufacturing.
The creation of this new alloy stems from a clever combination of aluminum and other elements, precisely identified through simulations and machine learning techniques. This approach significantly narrowed down the vast array of possible material combinations to just 40, leading the team to discover the perfect mixture.
After successfully printing the alloy, the researchers conducted tests that confirmed their expectations: this new aluminum alloy boasts strength comparable to the best materials currently available. The possibilities for its application are truly exciting. The team imagines this advanced aluminum being used in a variety of fields, including the aerospace industry, where it could revolutionize products like fan blades in jet engines. Traditionally, these blades are made from titanium, which is over 50% heavier and significantly more expensive than aluminum.
As Assistant Professor Mohadeseh Taheri-Mousavi from Carnegie Mellon University, who led the research at MIT, points out, using lighter and stronger materials could lead to substantial energy savings in transportation. The potential extends beyond aviation; John Hart, the former head of MIT's Department of Mechanical Engineering, envisions applications in advanced vacuum pumps, luxury automobiles, and cooling systems for data centers, thanks to the unique designs and material efficiency that 3D printing allows.
The journey to this innovation began in 2020 during a materials science class taught by Professor Greg Olson at MIT, where Taheri-Mousavi was inspired to explore how computational simulations could help design high-performance alloys. Although the class project did not yield a winning design, it sparked a curiosity in Taheri-Mousavi about the potential of machine learning to enhance the process.
She reflects on the complexity of material properties, noting, “At some point, there are a lot of things that contribute nonlinearly to a material’s properties, and you are lost.” However, with machine learning tools, researchers can focus their efforts more effectively, identifying key elements that influence material characteristics.
In this recent study, Taheri-Mousavi and her team continued the exploration initiated in Olson's class. Their machine-learning approach unveiled a formula that incorporated aluminum along with five additional elements, yielding an alloy with strength surpassing what they had previously discovered through over one million simulations without these advanced tools.
To bring this innovative alloy to life, the team opted for 3D printing, a method well-suited for capturing the unique microstructure and rapid cooling rates that contribute to the alloy's impressive properties. This new material maintains its stability at remarkably high temperatures, reaching up to 400 degrees Celsius.
Taheri-Mousavi shares her aspirations for the future, expressing a heartfelt vision: “My dream is that one day, passengers looking out their airplane window will see fan blades of engines made from our aluminum alloys.” This breakthrough not only showcases the ingenuity of MIT's researchers but also opens up new avenues for the future of lightweight, high-strength materials that can enhance various industries.
