Innovations in Finite-Temperature Density Functionals

Karasiev, Valentin V.; Sjostrom, Travis; Chakraborty, Debajit; Dufty, James W.; Runge, Keith; Harris, Frank E.; Trickey, S. B.

doi:10.1007/978-3-319-04912-0_3

Cited by 15 publications

(25 citation statements)

References 94 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A further challenge for finite-temperature KS-DFT is the computational cost, because a large number of KS states must be accounted for under WDM conditions [77]. Despite this limitation, and the above-cited difficulties with developing suitable XC approximations, KS-DFT is currently the predominant first-principles method for simulations of large systems in this thermodynamic regime; alternative approaches such as orbital-free DFT (OF-DFT) [78][79][80] or path-integral Monte-Carlo (PIMC) [35,[81][82][83][84] tend either to not be sufficiently accurate (OF-DFT [85,86]) or too expensive (PIMC).…”

Section: Introductionmentioning

confidence: 99%

First-principles derivation and properties of density-functional average-atom models

Callow¹,

Kraisler²,

Hansen³

et al. 2021

Preprint

View full text Add to dashboard Cite

Finite-temperature Kohn-Sham density-functional theory (KS-DFT) is a widely-used method in warm dense matter (WDM) simulations and diagnostics; however, full KS-DFT-molecular dynamics models scale unfavorably with temperature and there remains uncertainty regarding the performance of exchange-correlation (XC) functionals under WDM conditions. Average-atom (AA) models, which significantly reduce the computational cost of KS-DFT calculations, have therefore become an integral part of WDM modeling. In this paper, we present a derivation of a first-principles AA model from the fully-interacting many-body Hamiltonian, carefully analysing the assumptions made and terms neglected in this reduction. We explore the impact of different choices within this model -such as boundary conditions and XC functionals -on common properties in WDM, for example equation-of-state data. Furthermore, drawing upon insights from ground-state KS-DFT, we speculate on likely sources of error in KS-AA models and possible strategies for mitigating against such errors.

show abstract

Section: Introductionmentioning

confidence: 99%

First-principles derivation and properties of density-functional average-atom models

Callow¹,

Kraisler²,

Hansen³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…For functional development work, this artifact-free uniformity of treatment is important. For materials simulation research generally, the Profess@Q-Espresso package makes up-to-date OF-DFT functionals and optimization techniques available to those who presently run KS-AIMD calculations, again on a uniform, artifact-free basis [28]. Third, use of packages such as Profess and Quantum Espresso which have published source code avoids wasteful duplication of software development effort.…”

Section: Introductionmentioning

confidence: 99%

Finite-temperature orbital-free DFT molecular dynamics: Coupling Profess and Quantum Espresso

Karasiev

Sjostrom

Trickey

2014

Computer Physics Communications

Self Cite

View full text Add to dashboard Cite

Implementation of orbital-free free-energy functionals in the Profess code and the coupling of Profess with the Quantum Espresso code are described. The combination enables orbital-free DFT to drive ab initio molecular dynamics simulations on the same footing (algorithms, thermostats, convergence parameters, etc.) as for Kohn-Sham (KS) DFT. All the non-interacting free-energy functionals implemented are single-point: the local density approximation (LDA; also known as finite-T Thomas-Fermi, ftTF), the second-order gradient approximation (SGA or finite-T gradient-corrected TF), and our recently introduced finite-T generalized gradient approximations (ftGGA). Elimination of the KS orbital bottleneck via orbital-free methodology enables high-T simulations on ordinary computers, whereas those simulations would be costly or even prohibitively time-consuming for KS molecular dynamics (MD) on very high-performance computer systems. Example MD simulations on H over a temperature range 2, 000 K ≤ T ≤ 4, 000, 000 K are reported, with timings on small clusters (16-128 cores) and even laptops. With respect to KS-driven calculations, the orbital-free calculations are between a few times through a few hundreds of times faster.

show abstract

“…15.Mb,31.15.EWarm dense matter (WDM) is a highly energetic phase of matter with characteristics of both solids and plasmas [1]. The high temperatures and pressures necessary for creation of WDM are present in the centers of giant planets and on the path to ignition of inertial confinement fusion capsules [2,3].…”

mentioning

confidence: 99%

“…The computational bottleneck in these calculations is the solution of the KS equations, a step that becomes increasingly expensive as temperatures and fractional occupations rise. In fact, the cost exhibits nearly exponential scaling with temperature due to the KS cycle including many states at WDM temperatures [15].…”

mentioning

confidence: 99%