This record has versions v1, v2. This is version v1.

Recommended by

Indexed by

Machine learning for metallurgy: a neural network potential for Al-Cu-Mg

Daniel Marchand1*, W.A. Curtin1*

1 Institute of Mechanical Engineering, École Polytechnique Fédérale de Lausanne, CH-1015, Vaud, Switzerland

* Corresponding authors emails:,
DOI10.24435/materialscloud:z9-24 [version v1]

Publication date: Jul 14, 2021

How to cite this record

Daniel Marchand, W.A. Curtin, Machine learning for metallurgy: a neural network potential for Al-Cu-Mg, Materials Cloud Archive 2021.106 (2021), doi: 10.24435/materialscloud:z9-24.


High-strength metal alloys achieve their performance via careful control of precipitates and solutes. The nucleation, growth, and kinetics of precipitation, and the resulting mechanical properties, are inherently atomic-scale phenomena, particularly during early-stage nucleation and growth. Atomistic modeling using interatomic potentials is a desirable tool for understanding the detailed phenomena involved in precipitation and strengthening, which requires length and time scales far larger than those accessible by first-principles methods. Current interatomic potentials for alloys are not, however, sufficiently accurate for such studies. Here, a family of neural-network potentials (NNPs) for the Al-Cu-Mg system is presented as the first example of a machine-learning potential that can achieve near-first-principles accuracy for many different metallurgically-important aspects of this alloy. High fidelity predictions of intermetallic compounds, elastic constants, dilute solid-solution energetics, precipitate/matrix interfaces, generalized stacking fault energies, and surfaces for slip in matrix and precipitates, antisite defect energies, and other quantities, are shown. The NNP shows some significant transferability to defects and properties outside the structures used to develop the NNP but also shows some errors, highlighting that the use of any interatomic potential requires careful validation in application to specific metallurgical problems of interest.

Materials Cloud sections using this data

No Explore or Discover sections associated with this archive record.


File name Size Description
2.3 MiB All neural network potentials used in this work.
5.8 MiB Structures after removal via furthest point sampling
9.2 MiB All structures prior to removal via furthest point sampling
4.6 MiB All properties as computed by all neural network potentials and DFT in json format for plotting
1.5 KiB README file


Files and data are licensed under the terms of the following license: Creative Commons Attribution 4.0 International.
Metadata, except for email addresses, are licensed under the Creative Commons Attribution Share-Alike 4.0 International license.

External references

Journal reference
D. Marchand and W. A. Curtin [in preparation]


Aluminum Machine Learning Metallurgy MARVEL/DD2

Version history:

2022.39 (version v2) Mar 14, 2022 DOI10.24435/materialscloud:ea-y9
2021.106 (version v1) [This version] Jul 14, 2021 DOI10.24435/materialscloud:z9-24