Bayesian calibration of interatomic potentials for binary alloys
Source abstract
Abstract
Developing reliable interatomic potential models with quantified predictive accuracy is crucial for atomistic simulations. Commonly used potentials, such as those constructed through the embedded atom method (EAM), are derived from semi-empirical considerations and contain unknown parameters that must be fitted based on training data. In the present work, we investigate Bayesian calibration as a means of fitting EAM potentials for binary alloys. The Bayesian setting naturally assimilates probabilistic assertions about uncertain quantities. In this way, uncertainties about model parameters and model errors can be updated by conditioning on the training data and then carried through to prediction. We apply these techniques to investigate an EAM potential for a family of gold-copper systems in which the training data correspond to density-functional theory values for lattice parameters, mixing enthalpies, and various elastic constants. Through the use of predictive distributions, we demonstrate the limitations of the potential and highlight the importance of statistical formulations for model error.
Source: openaire · Quality: candidate unverified · Source record ↗
Research context
OpenAlex topics: Machine Learning in Materials Science · X-ray Diffraction in Crystallography · Microstructure and mechanical properties
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (80)
Show source-provided references
1. The art and science of an analytic potential10.1002/(SICI)1521-3951(200001)217:1<23::AID-PSSB23>3.0.CO;2-N ↗
2. Linear scaling electronic structure methods10.1103/RevModPhys.71.1085 ↗
3. Bayesian ensemble approach to error estimation of interatomic potentials10.1103/PhysRevLett.93.165501 ↗
4. Bayesian uncertainty quantification and propagation in molecular dynamics simulations: a high performance computing framework10.1063/1.4757266 ↗
5. Bayesian calibration of force-fields from experimental data: TIP4P water10.1063/1.5030950 ↗
6. Uncertainty quantification for classical effective potentials: an extension to potfit10.1088/1361-651X/ab0d75 ↗
7. Uncertainty quantification for molecular dynamics
8. Bayesian calibration of force fields for molecular simulations
9. Ensembles are required to handle aleatoric and parametric uncertainty in molecular dynamics simulation10.1021/acs.jctc.1c00526 ↗
10. Statistical prediction and molecular dynamics simulation10.1529/biophysj.108.131623 ↗
11. Statistical approaches to forcefield calibration and prediction uncertainty in molecular simulation10.1063/1.3545069 ↗
12. A critical review of statistical calibration/prediction models handling data inconsistency and model inadequacy10.1002/aic.15781 ↗
13. Uncertainty quantification in MD simulations. Part I: Forward propagation10.1137/110853169 ↗
14. Uncertainty quantification in MD simulations. Part II: Bayesian inference of force-field parameters10.1137/110853170 ↗
15. Uncertainty quantification and reduction of molecular dynamics models
16. Uncertainty quantification in multiscale simulation of materials: A prospective10.1146/annurev-matsci-071312-121708 ↗
17. Rapid production of accurate embedded-atom method potentials for metal alloys
18. Application of the embedded-atom method to covalent materials: a semiempirical potential for silicon10.1103/PhysRevLett.59.2666 ↗
19. Modified embedded-atom potentials for cubic materials and impurities10.1103/PhysRevB.46.2727 ↗
20. Spectral neighbor analysis method for automated generation of quantum-accurate interatomic potentials10.1016/j.jcp.2014.12.018 ↗
21. Machine learning for interatomic potential models10.1063/1.5126336 ↗
22. Becker · 2011
23. Considerations for choosing and using force fields and interatomic potentials in materials science and engineering10.1016/j.cossms.2013.10.001 ↗
24. Evaluating variability with atomistic simulations: the effect of potential and calculation methodology on the modeling of lattice and elastic constants10.1088/1361-651X/aabc05 ↗
25. The potential of atomistic simulations and the knowledgebase of interatomic models10.1007/s11837-011-0102-6 ↗
26. Knowledgebase of interatomic models (KIM) application programming interface (API)
27. The embedded atom method
28. On fitting a gold embedded atom method potential using the force matching method10.1063/1.2124667 ↗
29. Fitting empirical potentials: Challenges and methodologies10.1016/j.cossms.2013.09.001 ↗
30. Gao · 2016
31. Multicomponent high-entropy cantor alloys
32. A simple empirical N-body potential for transition metals10.1080/01418618408244210 ↗
33. Embedded-atom method: Derivation and application to impurities, surfaces, and other defects in metals10.1103/PhysRevB.29.6443 ↗
34. An embedded atom method potential of beryllium10.1088/0965-0393/21/8/085001 ↗
35. Corrigendum: An embedded atom method potential of beryllium (modelling simul. Mater. Sci. Eng. 21 085001)10.1088/0965-0393/23/6/069501 ↗
36. EAM Potentials for BCC, FCC and HCP metals with farther neighbor atoms10.1007/s00339-015-9149-5 ↗
37. Performance analysis and optimization of the RAMPAGE metal alloy potential generation software
38. Fast parallel algorithms for short-range molecular dynamics10.1006/jcph.1995.1039 ↗
39. Accurate interatomic potentials for Ni, Al and Ni3Al
40. Riegner · 2016
41. Rapid production of accurate embedded-atom method potentials for metal alloys
42. Bond-order bond energy model for alloys10.1016/j.actamat.2019.08.048 ↗
43. Generation and performance of special quasirandom structures for studying the elastic properties of random alloys: Application to Al-Ti10.1103/PhysRevB.81.094203 ↗
44. Ab initio molecular-dynamics simulation of the liquid-metal–amorphous-semiconductor transition in germanium10.1103/PhysRevB.49.14251 ↗
45. Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set10.1016/0927-0256(96)00008-0 ↗
46. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set10.1103/PhysRevB.54.11169 ↗
47. From ultrasoft pseudopotentials to the projector augmented-wave method10.1103/PhysRevB.59.1758 ↗
48. Bayesian calibration of computer models10.1111/1467-9868.00294 ↗
49. Combining field data and computer simulations for calibration and prediction10.1137/S1064827503426693 ↗
50. Computer model calibration using high-dimensional output10.1198/016214507000000888 ↗
51. A framework for validation of computer models10.1198/004017007000000092 ↗
52. On the statistical calibration of physical models10.1002/kin.20906 ↗
53. Bernardo · 2009
54. Gelman · 1995
55. Computational aspects of many-body potentials10.1557/mrs.2012.96 ↗
56. Performance and cost assessment of machine learning interatomic potentials10.1021/acs.jpca.9b08723 ↗
57. Bishop · 2006
58. Hyperparameters: optimize, or integrate out?
59. Bayes factors10.1080/01621459.1995.10476572 ↗
60. Bayesian interpolation10.1162/neco.1992.4.3.415 ↗
61. Posterior predictive assessment of model fitness via realized discrepancies
62. Liu · 2008
63. Design and analysis of computer experiments
64. Posterior consistency for Gaussian process approximations of Bayesian posterior distributions10.1090/mcom/3244 ↗
65. A comparison of three methods for selecting values of input variables in the analysis of output from a computer code
66. Misfit-energy-increasing dislocations in vapor-deposited CoFe/NiFe multilayers10.1103/PhysRevB.69.144113 ↗
67. Voter · 1993
68. Rasmussen · 2006
69. Learning about physical parameters: The importance of model discrepancy10.1088/0266-5611/30/11/114007 ↗
70. Embedded model error representation for Bayesian model calibration
71. Optimizing special quasirandom structure (SQS) models for accurate functional property prediction in disordered 2D alloys
72. Numerical challenges in the use of polynomial chaos representations for stochastic processes10.1137/S1064827503427741 ↗
73. The uncertainty quantification toolkit (UQTk)
74. Data analysis recipes: Using markov chain monte carlo10.3847/1538-4365/aab76e ↗
75. Global sensitivity analysis and estimation of model error, toward uncertainty quantification in Scramjet Computations10.2514/1.J056278 ↗
76. Interatomic potentials from first-principles calculations: the force-matching method10.1209/0295-5075/26/8/005 ↗
77. Generalized gradient approximation made simple10.1103/PhysRevLett.78.1396 ↗
78. Special points for Brillouin-zone integrations10.1103/PhysRevB.13.5188 ↗
79. VASPKIT: A user-friendly interface facilitating high-throughput computing and analysis using VASP code10.1016/j.cpc.2021.108033 ↗
80. Efficient stochastic generation of special quasirandom structures