Abstract by Hayden Oliver

Personal Infomation

Presenter's Name

Hayden Oliver

Degree Level



Chandramouli Nyshadham
Brayden Bekker
Carlos Leon
Gus L.W. Hart

Abstract Infomation


Physics and Astronomy

Faculty Advisor

Gus Hart


Accelerating superalloy discovery using moment tensor potentials*


The high-temperature mechanical strength of Ni-based superalloys lends them well to high-temperature applications. Their extraordinary mechanical properties rely on the formation of the L12 - γ' phase. We report the discovery of a stable L12 configuration in the candidate superalloy system Co-Al-W analagous to Ni-based alloys. In the past, density functional theory (DFT) has been used to investigate superalloy systems but its expensive computations limit how fast we can discover new superalloys. Moment Tensor Potentials (MTP) are a polynomial based interatomic potential trained via machine learning that accurately evaluates the crystal structure energy many orders of magnitude faster than DFT.

*Authors acknowledge funding from ONR (MURI N00014-13-1-0635)