An official website of the United States government
The Cluster Expansion (CE) Method encounters significant computational challenges in multicomponent systems due to the computational expense of generating training data through density functional theory (DFT) calculations. This work aims to refine the cluster and structure selection processes to mitigate these challenges. We introduce a novel method that significantly reduces the computational load associated with the calculation of fitting parameters. This method employs a Graph Neural Network (GNN) model, leveraging the M3GNet network, which is trained using a select subset of DFT calculations at each ionic step. The trained surrogate model excels in predicting the volume and energy of the final structure for a relaxation run. By employing this model, we sample thousands of structures and fit a CE model to the energies of these GNN-relaxed structures. This approach, utilizing a large training dataset, effectively reduces the risk of overfitting, yielding a CE model with a root-mean-square error (RMSE) below 10 meV/atom. We validate our method’s effectiveness in two test cases: the (Ti, Cr, Zr, Mo, Hf, Ta)B2 diboride system and the Refractory High-Entropy Alloy (HEA) AlTiZrNbHfTa system. Our findings demonstrate the significant advantages of integrating a GNN model, specifically the M3GNet network, with CE methods for the efficient predictive analysis of chemical ordering in High Entropy Materials. The accelerating capabilities of the hybrid ML-CE approach to investigate the evolution of Short Range Ordering (SRO) in a large number of stoichiometric systems. Finally, we show how it is possible to correlate the strength of chemical ordering to easily accessible alloy parameters.
more »
« less
Bayesian active machine learning for cluster expansion construction
The Cluster expansion (CE) is a powerful method for representing the energetics of alloys from a fit to first principles energies. However, many common fitting methods are computationally demanding and do not provide the guarantee that the system’s ground states are preserved. This paper demonstrates the use of an efficient implementation of a Bayesian algorithm for cluster expansion construction that ensures all the input structural energies are fitted perfectly while reducing computational cost. The method incorporates an active learning scheme that searches for new optimal structures to include in the fit. As performance tests, we calculate the phase diagram of the Fe-Ir system and study the short range order in an equimolar MoNbTaVW system. The new method has been integrated into the Alloy Theoretic Automated Toolkit (ATAT).
more »
« less
- Award ID(s):
- 2001411
- PAR ID:
- 10541989
- Publisher / Repository:
- Computational material science
- Date Published:
- Journal Name:
- Computational Materials Science
- Volume:
- 231
- Issue:
- C
- ISSN:
- 0927-0256
- Page Range / eLocation ID:
- 112571
- Subject(s) / Keyword(s):
- cluster expansion Bayesian algorithm active learning
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Abundance spectra of (CO 2 ) N clusters up to N ≈ 500 acquired under a wide range of adiabatic expansion conditions are analyzed within the evaporative ensemble framework. The analysis reveals that the cluster stability functions display a strikingly universal pattern for all expansion conditions. These patterns reflect the inherent properties of individual clusters. From this analysis the size-dependent cluster binding energies are determined, shell and subshell closing sizes are identified, and cuboctahedral packing ordering for sizes above N ≈ 130 is confirmed. It is demonstrated that a few percent variation in the dissociation energies translates into significant abundance variations, especially for the larger clusters.more » « less
-
Efficient and accurate computational methods for dealing with interacting electron problems on a lattice are of broad interest to the condensed matter community. For interacting Hubbard models, we introduce a cluster slave-particle approach that provides significant computational savings with high accuracy for total energies, site occupancies, and interaction energies. Compared to exact benchmarks using density matrix renormalization group for d-p Hubbard models, our approach delivers accurate results using two to three orders of magnitude lower computational cost. Our method is based on a slave-particle decomposition with an improved description of particle hoppings, and a density matrix expansion method where the interacting lattice slave-particle problem is turned into a set of overlapping real-space clusters which are solved self-consistently with appropriate physical matching constraints at shared lattice sites between clusters.more » « less
-
Using the many-body expansion to predict crystal lattice energies (CLEs), a pleasantly parallel process, allows for flexibility in the choice of theoretical methods. Benchmark-level two-body contributions to CLEs of 23 molecular crystals have been computed using interaction energies of dimers with minimum inter-monomer separations (i.e., closest contact distances) up to 30 Å. In a search for ways to reduce the computational expense of calculating accurate CLEs, we have computed these two-body contributions with 15 different quantum chemical levels of theory and compared these energies to those computed with coupled-cluster in the complete basis set (CBS) limit. Interaction energies of the more distant dimers are easier to compute accurately and several of the methods tested are suitable as replacements for coupled-cluster through perturbative triples for all but the closest dimers. For our dataset, sub-kJ mol−1 accuracy can be obtained when calculating two-body interaction energies of dimers with separations shorter than 4 Å with coupled-cluster with single, double, and perturbative triple excitations/CBS and dimers with separations longer than 4 Å with MP2.5/aug-cc-pVDZ, among other schemes, reducing the number of dimers to be computed with coupled-cluster by as much as 98%.more » « less
-
The large-Z asymptotic expansion of atomic exchange energies has been useful in determining exact conditions for corrections to the local density approximation in density functional theory. We find that the necessary correction is fit well with a leading ZlnZ term, and find its coefficient numerically. The gradient expansion approximation also displays such a term, but with a substantially smaller coefficient. Analytic results in the limit of vanishing interaction with hydrogenic orbitals (a Bohr atom) are given, leading to the conjecture that the true coefficients for all atoms are precisely 2.7 times larger than their gradient expansion counterpart. Combined with the hydrogen atom result, this yields an analytic expression for the exchange-energy correction which is accurate to ∼5% for all Z.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.1016/j.commatsci.2023.112571
