Pith. sign in

REVIEW 4 major objections 6 minor 1 cited by

Advanced Langevin thermostats: Properties, extensions to rheology, and a lean momentum-conserving approach

T0 review · 4 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This paper shows that a recycled-random-number Langevin integrator gives exact harmonic thermal averages at two-to-ten-fold larger time steps, while a bin-based momentum-conserving Langevin scheme preserves long-wavelength modes.

desk verdict Useful thermostat variants with transparent derivations and public code; the 'best-possible' branding is unproved and MCL's hydrodynamic claim rests on a single 8-atom 1D chain. read the letter →

arxiv 2506.15579 v1 pith:5DFYH3YH submitted 2025-06-18 cond-mat.stat-mech cond-mat.mtrl-scicond-mat.soft

Shubham Agarwal , Sergey V. Sukhomlinov , Marc Honecker , Martin H. Müser This is my paper · ORCID
classification cond-mat.stat-mechcond-mat.mtrl-scicond-mat.soft MSC 82C3182C8065C30
keywords Langevinthermostatmoleculardynamicsthermalaveragesharmonicoscillatormomentum-conservingGJschemestochasticintegratorshydrodynamicmodes
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper argues that thermostats which rescale every velocity by a common factor—global kinetic-energy controls—cannot be true canonical samplers, because degenerate, weakly coupled vibrational modes stay phase-locked forever, corrupting quantities such as elastic response. Its counter-proposal is a family of Langevin-equation solvers whose update coefficients are fixed by matching exact limits: a memory-free “best-possible conventional Langevin” (BPCL) scheme, and the GJ scheme, which recycles a Gaussian random number from the previous step and thereby reproduces harmonic-oscillator thermal averages exactly. The paper also introduces a lean momentum-conserving Langevin (MCL) thermostat, which damps only motion relative to a small local center of mass, leaving long-wavelength, hydrodynamic modes underdamped. If these claims are right, standard molecular-dynamics practice can shift from global rescaling toward these Langevin solvers, gaining time steps two to ten times larger while keeping mean potential-energy errors within 0.0025 $k_B T$ per degree of freedom.

What carries the argument

The machinery is the four-coefficient leapfrog update $v_{n+1/2}=c_{vv}v_{n-1/2}+c_{vf}f_n+c_{vg}g_n$, $x_{n+1}=x_n+c_{xv}v_{n+1/2}$, with coefficients fixed by exact limits. BPCL chooses $c_{vv}=\exp(-\Delta t/\tau)$ and enforces exact constant-force drift, exact equipartition, and exact kinetic energy for a force-free particle. The GJ variant changes the noise term from one fresh Gaussian per step to a fresh plus a recycled Gaussian, which closes every harmonic covariance identity and makes the oscillator distribution exact; BPCL is the best possible version without that memory. MCL applies BPCL to coordinates measured relative to each bin's center of mass and advances the center of mass symplectically, so only relative motion is damped and noisy. The job of this machinery is to push the leading error in the harmonic mean potential energy down to $O(\Delta t^2)$ and, for GJ, to remove it entirely.

What would settle it

Scan the full four-parameter space of the conventional update (or solve its covariance equations symbolically) for a harmonic oscillator and check whether any memory-free coefficient set yields a mean potential-energy error below $\omega_0^2\Delta t^2/4$; a counterexample refutes BPCL's optimality. For MCL, run the scheme in a three-dimensional Lennard-Jones liquid with random bin origins and measure the damping of the longest-wavelength mode; if it follows the $q$-dependence of laboratory-frame damping rather than the momentum-conserving prediction, the central MCL claim fails.

Watch

Extended reading notes

Core claim

The central claim is that the GJ scheme is the most accurate Langevin-type thermostat for harmonic and near-harmonic systems, yielding correct thermal averages across a broad range of damping constants and time steps, and that the new MCL thermostat preserves long-wavelength vibrational modes and hydrodynamic interactions, avoiding the overdamping artifacts of thermostats acting in a fixed laboratory frame. The derivation proceeds by writing the integrator as a four-coefficient update and fixing those coefficients through asymptotic requirements: correct free-particle decay, correct drift under a constant force, equipartition, and exact kinetic energy for a force-free particle. Adding a second, recycled random number closes all covariance equations for a harmonic reference, which is why GJ reproduces the exact second and fourth moments of the oscillator while BPCL and simpler schemes carry a leading error of $\omega_0^2\Delta t^2/4$ in mean potential energy. Across single-sinusoidal potentials, one- and three-dimensional Lennard-Jones liquids, and liquid copper, the paper reports energy errors below $k_B T/400$ per degree of freedom at time steps two to ten times larger than conventional stability limits.

Load-bearing premise

The load-bearing premise is that no memory-free Langevin solver can beat the leading error $\omega_0^2\Delta t^2/4$ in the harmonic-oscillator mean potential energy; if a competitor did, the claim that BPCL is the best conventional Langevin thermostat collapses, and the MCL branch separately assumes that a one-dimensional eight-atom chain at very low temperature is a reliable guide to bulk three-dimensional bins.

Editorial extensions

If this is right

  • Molecular dynamics of stiff materials can run at time steps of 5–10 fs for copper at 1400 K with the GJ scheme, compared with the usual 1–2 fs, while keeping mean potential-energy errors below $0.0025 k_B T$ per atom.
  • Global velocity-rescaling thermostats should not be used for response functions built from weakly coupled quasi-harmonic variables, such as box-shape fluctuations used to extract elastic constants.
  • Applying the paper's additive correction $U_{\mathrm{BPCL}^*}=U_{\mathrm{BPCL}}+k_BT/2-\langle T_{\mathrm{kin}}\rangle$ doubles the usable time step of BPCL for a fixed accuracy target.
  • In near-harmonic systems, a GJ run that stays numerically stable can be trusted as accurate, because the range of time steps where the scheme is stable but inaccurate is extremely narrow or absent.
  • The MCL thermostat lets long-wavelength modes in large homogeneous systems keep oscillating, which makes it a candidate for equilibrium and nonequilibrium runs that need hydrodynamic interactions intact.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A decisive bulk test not yet performed is MCL in a three-dimensional Lennard-Jones liquid with random bin origins; if the long-wavelength damping follows the derived $q$-dependence, MCL becomes a cheap stand-in for dissipative particle dynamics.
  • Because MCL conserves linear but not angular momentum, it will add some viscosity; a natural next measurement is the excess viscosity as a function of bin size and damping, with bin masses tuned to offset it.
  • The asserted lower bound on conventional Langevin schemes is stated without proof; a direct search over the four propagation coefficients for a harmonic oscillator would settle whether the bound is true.
  • The same coefficient-fixing strategy could be applied to auxiliary variables such as cell-shape degrees of freedom, where a Brownian-driven Maxwell element may give smoother equilibration than direct Langevin damping.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. This paper develops and tests several Langevin-based thermostats. It introduces a generic form for conventional Langevin solvers, derives the lowest-order Langevin (LOL) and 'best-possible conventional Langevin' (BPCL) coefficients, discusses the Grønbech-Jensen (GJ) scheme and an advanced Brownian thermostat, proposes a Maxwell-element version called GJFM, and introduces a momentum-conserving Langevin (MCL) thermostat based on spatial bins. The numerical sections compare mean potential energy errors for a harmonic oscillator, a sinusoidal potential, a one-dimensional Lennard-Jones chain, bulk liquid Lennard-Jonesium, and liquid copper, and show a single one-dimensional chain demonstration for MCL. The central claims are that BPCL is the optimal conventional Langevin scheme, that GJ reproduces exact harmonic-reference averages and remains accurate for anharmonic systems, and that MCL preserves long-wavelength vibrational modes and hydrodynamic interactions.

Significance. The paper has genuine strengths: the asymptotic derivations in Section II are compact and mostly checkable, the code and data are made publicly available, and the paper usefully contrasts global kinetic-energy controls with local Langevin thermostats. If the BPCL optimality claim and the GJ harmonic-exactness claim were fully proven, and if MCL were validated in bulk systems, the work would provide practitioners with a practical comparison of thermostat algorithms and with new thermostat options. However, as it stands, the two most distinctive claims—the 'best-possible' property of BPCL and the hydrodynamic-preservation claim of MCL—rest on an unproved lower bound and on a single one-dimensional 8-atom simulation, respectively. The significance of the paper is therefore conditional on closing those gaps.

major comments (4)
  1. [Section II.B.1, before Eq. (19)] The statement that no conventional Langevin thermostat can have a systematically smaller relative leading error in the mean potential energy of a harmonic oscillator than omega_0^2 dt^2/4 is asserted without proof. Equation (14) expresses <x^2> in terms of the four propagation coefficients, but no optimization over the allowed c_vv, c_vf, c_vg, c_xv parameter space is performed. Since this lower bound is what justifies the name 'best-possible conventional Langevin' and is quoted in the conclusions, it needs a rigorous derivation, or the claim should be weakened to a statement about the specific scheme proposed.
  2. [Section II.C.2, Eqs. (41)-(45) and the passage after Eq. (45)] The paper acknowledges that the coefficient choices make Eq. (41) a 'self-fulfilling prophecy,' and then states that exactness can be confirmed by evaluating all possible covariances, but that evaluation is not shown. The abstract and conclusions nonetheless claim that GJ is exact for harmonic references. As written this is a circular validation; a complete closed-form verification of the second-moment relations, or a reference to a full proof, is required before the exactness claim can be considered established.
  3. [Section II.E and Fig. 10] The MCL algorithm is introduced for spatial bins with random origins and atoms crossing bin boundaries, but the only simulation presented is a one-dimensional 8-atom Lennard-Jones chain at k_B T = 0.001 epsilon_LJ with damping set to twice the largest period. The conclusions claim that MCL 'preserves long-wavelength vibrational modes and hydrodynamic interactions' in general, but the bulk three-dimensional algorithm is never simulated. The paper itself notes that 'the accuracy of advanced thermostats can be decremented' in this setting and that the GJ variant was abandoned for this reason; therefore the central MCL claim is currently unverified.
  4. [Figures 5-9] The numerical comparisons of mean potential energy are presented without statistical error bars or confidence intervals, even though Section II.H mentions multiple replicas for on-the-fly error estimation. Without such information, quantitative claims such as the factor-of-80 reduction in leading-order error in Fig. 5 and the factor-of-9 time-step advantage cannot be distinguished from statistical noise.
minor comments (6)
  1. [Section II.C.2, Eqs. (44)-(45)] The relation c_vf = c_xv/m appears twice, once as Eq. (44) and again in the line after Eq. (45); this duplication obscures the derivation of c_xv and should be cleaned up.
  2. [Figure 10 caption] The caption contains the misspelling 'momentem-conserving'; it should read 'momentum-conserving.'
  3. [Section III.A] The symbol T is used both for the oscillation period and for temperature; although the text warns the reader about this convention, using a different symbol for one of the two quantities would avoid confusion.
  4. [Section III.C, Eq. (65)] The potential is written as U(x) = -U_0 cos(qx), but the text immediately afterward uses kappa = q^2 |V_0|; the notation for the potential amplitude should be harmonized.
  5. [Section II.E, pseudo-code] The pseudo-code line 'g = g - mean(g)' is ambiguous because it is not specified whether the mean is subtracted per Cartesian component; this should be stated explicitly.
  6. [Introduction] The parenthetical remark about sex-based differences in voice pitch is not connected to the technical content of the paper and should be removed or placed in a clearly separate discussion.

Circularity Check

1 steps flagged · score 4.0 of 10

GJ harmonic exactness is a self-acknowledged self-fulfilling construction; other central claims (BPCL, MCL, GJFM, anharmonic benchmarks) are independent.

  1. self definitional [Sect. II C 2, Eqs. (41)-(45); conclusions Sect. IV]
    "Ideally, the integration scheme satisfies equipartition ⟨x2⟩=kBT/k (41a) ⟨v2⟩=kBT/m. (41b) ... Up to this point, we have only shown that the cαβ are optimal provided that Eq. (41) is a self-fulfilling prophecy."

    Eq. (41) postulates the exact harmonic equipartition values; Eqs. (43) and (45) solve for c_vg and c_vf by inserting that postulate into the variance equation. The subsequent demonstration that GJ reproduces those moments (Fig. 4) and the conclusion that GJ 'yield[s] correct thermal averages' for harmonic systems therefore re-state the construction. The paper itself calls this a 'self-fulfilling prophecy', confirming that the harmonic test is not an independent outcome. The circularity is real but narrow: the coefficients were not fitted to simulation data, and the anharmonic benchmarks plus the novel MCL/BPCL/GJFM results are not derived from the equipartition assumption.

full rationale

The only circular element I found is the GJ harmonic-oscillator validation: the coefficients are constructed by imposing the very harmonic equipartition moments that the paper later says GJ 'demonstrated.' The paper explicitly labels this a 'self-fulfilling prophecy,' so the reduction is visible in the text rather than inferred. This does not taint the paper's other claimed contributions: BPCL is optimized against free-particle and constant-force limits rather than against the harmonic energy used for validation; the MCL claim is supported by an analytic harmonic-chain dispersion argument plus a 1D simulation; GJFM is a separate Maxwell-element construction; and the anharmonic benchmarks (sinusoidal potential, LJ chain and bulk liquid, liquid copper) were not used to fit any GJ coefficient. The BPCL 'best-possible' label relies on an unproved optimality assertion before Eq. (19), which is an unsupported premise rather than a circular step, and the MCL bulk 3D claim is an extrapolation from an 8-atom 1D test, which is an evidence gap rather than circularity. Overall score 4 reflects the partial self-definition in the GJ harmonic validation while recognizing that the central new results retain independent content.

Assumptions & free parameters 2 free parameters · 5 assumptions · 0 invented entities

The central derivations rest on the fluctuation-dissipation relation and on harmonic oscillators as the accuracy reference. The main uncounted assumption is the unproved optimality of BPCL. The proposed methods introduce user-chosen parameters (Maxwell stiffness, bin size) that will need case-specific tuning, though the paper does not hide this.

free parameters (2)
  • Maxwell element spring constant k_Mxw = k/8 for LJ chain; k/2 in Fig. 2
    User-chosen parameter of the proposed GJFM thermostat; the paper says the typical parameterization varies by use case, so no general rule is given, and the benchmark values are set by hand.
  • MCL bin size = about two atomic diameters (roughly 8 atoms in 3D)
    User-specified linear dimension of the binning boxes; the method's accuracy and hydrodynamic behavior will depend on this choice, and the paper gives only a rough guideline.
assumptions (5)
  • domain assumption The integration scheme should produce the correct drift under a constant external force (Sect. II.B, Eqs. 22-24).
    Used to fix c_xv c_vf for BPCL and GJ.
  • domain assumption Exact first and second moments for a harmonic reference are the right design target (Sect. II.C, Eq. 41).
    The paper's notion of thermostat accuracy is anchored to harmonic references; non-harmonic systems are then judged by proximity.
  • ad hoc to paper There is no conventional Langevin scheme with smaller leading position error than BPCL (Sect. II.B, around Eq. 19).
    Unproved optimality assertion that justifies the name 'best-possible'.
  • standard math The linear chain continuum dispersion relation captures the relevant long-wavelength physics (Sect. II.G, Eqs. 60-63).
    Standard continuum limit with q^2 a^2 replaced by 4 sin^2(qa/2) for discreetness.
  • domain assumption The Markovian noise has zero mean and the fluctuation-dissipation relation (Eq. 3) holds.
    Standard Langevin bath assumption, invoked throughout the derivations.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Advanced Langevin thermostats: Properties, extensions to rheology, and a lean momentum-conserving approach." pith.science (2026). https://pith.science/paper/5DFYH3YH

@misc{pith2026250615579,
  author       = {Pith},
  title        = {Pith review of: Advanced Langevin thermostats: Properties, extensions to rheology, and a lean momentum-conserving approach},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5DFYH3YH}},
  note         = {Machine review of arXiv:2506.15579}
}
read the original abstract

The Langevin equation accounts for unresolved bath degrees of freedom driving the system toward the bath temperature. Because of this, numerical solutions of the Langevin equation have a long history. Here, we recapitulate, combine, and extend existing Langevin-equation based thermostats, scrutinize their properties and demonstrate their superiority over global kinetic-energy controls. Our work includes compact, asymptotic-analysis based derivations of stochastic thermostats, including the highly accurate Gr{\o}nbech-Jensen scheme. Proposed extensions include a precise, colored and a lean, momentum-conserving thermostat.

Figures

Figures reproduced from arXiv: 2506.15579 by the authors.

Figure 1
Figure 1. FIG. 1. Coupling of a central harmonic degree of freedom to [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 3
Figure 3. FIG. 3. Exact solution (lines) of the critically damped [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗
Figure 4
Figure 4. FIG. 4. Dimensionless second (top) and fourth (bottom) [PITH_FULL_IMAGE:figures/full_fig_p011_4.png] view at source ↗
Figures from the paper (5 more)
Figure 5
Figure 5. Figure 5: FIG. 5. Error in the mean potential energy ∆ [PITH_FULL_IMAGE:figures/full_fig_p012_5.png]
Figure 6
Figure 6. Figure 6: FIG. 6. Mean potential energy [PITH_FULL_IMAGE:figures/full_fig_p014_6.png]
Figure 8
Figure 8. Figure 8: FIG. 8. Mean potential energy [PITH_FULL_IMAGE:figures/full_fig_p014_8.png]
Figure 10
Figure 10. Figure 10: It shows the (relative) mean-square displace [PITH_FULL_IMAGE:figures/full_fig_p015_10.png]
Figure 9
Figure 9. Figure 9: reveals that time steps of 5–10 fs are readily possible when simulating copper using GJL, while most studies use time steps from 1–2 fs. An interesting feature of GJL is that the method either produces highly accu￾rate results, i.e., with errors in the potential energy…

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Finite-temperature properties of the Frenkel-Kontorova model: Relation to tribological systems and fluid rheology

    cond-mat.soft 2025-07 conditional novelty 6.0 of 10

    In the thermal Frenkel-Kontorova chain, shear thinning and viscosity are set by energy drops at mechanical instabilities, not by thermally activated barrier crossing, so Eyring fits can be coincidental.

Reference graph

Works this paper leans on

47 extracted references · 42 canonical work pages · cited by 1 Pith paper

  1. [1]

    The damping force is integrated as in a (non-symplectic) Euler scheme, which means that cvv = 1−∆t/τ.(16) When inserting this into Eq

    Lowest-order Langevin thermostat In the lowest-order Langevin thermostat (LOL), the coefficientsc vf andc xv are used as in the leapfrog Verlet algorithm, i.e.,c vf = ∆t/mandc xv = ∆t. The damping force is integrated as in a (non-symplectic) Euler scheme, which means that cvv = 1−∆t/τ.(16) When inserting this into Eq. 11, one can immediately recognize tha...

    work page

  2. [2]

    fudge factor

    Best-possible conventional Langevin thermostat The LOL produces acceptable thermal averages at small time steps with errors in⟨x 2⟩for the harmonic oscillator of order ∆t 2. However, it does not lead to ac- curate deterministic trajectories and higher-order errors 4 grow quickly with ∆twhen the damping is large. The question arises what the optimum choice...

    work page

  3. [3]

    Advanced Brownian thermostat A highly accurate solver for Brownian dynamics, pro- posed by Grønbech-Jensen and Farago (GJF) can be con- structed from the integration scheme [31] x++ =x+c xff+c xg(g+ +g−),(31) 5 where we use similar notation as above, e.g.,x ++ stands forx n+1,fforf n andg − forg n−1/2. For a harmonic oscillator, the integration scheme red...

    work page

  4. [4]

    usual rules

    Grønbech-Jensen thermostat In our notation, the GJ algorithm [9, 31] can be written as v+ =c vvv− +cvff+c vg(g− +g +) (37a) x++ =x+c xvv+,(37b) As in the advanced Brownian integrator, random num- bers are added to the velocity in two successive rather than in a single step. Thus, only those previous results can be recycled that do not depend on thermal ra...

    work page

  5. [5]

    M. P. Allen and D. J. Tildesley,Computer Simulations of Liquids(Oxford University Press, Oxford, 2017)

    work page 2017

  6. [6]

    Frenkel and B

    D. Frenkel and B. Smit,Understanding Molecular Sim- ulation: From Algorithms to Applications, 2nd ed. (Aca- demic Press, San Diego, 2001)

    work page 2001

  7. [7]

    H. J. C. Berendsen, J. P. M. Postma, W. F. van Gun- steren, A. DiNola, and J. R. Haak, Molecular dynamics with coupling to an external bath, The Journal of Chem- ical Physics81, 3684–3690 (1984)

    work page 1984

  8. [8]

    Nos´ e, A unified formulation of the constant tempera- ture molecular dynamics methods, The Journal of Chem- ical Physics81, 511–519 (1984)

    S. Nos´ e, A unified formulation of the constant tempera- ture molecular dynamics methods, The Journal of Chem- ical Physics81, 511–519 (1984)

    work page 1984

Show all 47 references
  1. [9]

    W. G. Hoover, Canonical dynamics: Equilibrium phase- space distributions, Physical Review A31, 1695–1697 (1985)

    1985

  2. [10]

    D. J. Evans and B. L. Holian, The Nose–Hoover ther- mostat, The Journal of Chemical Physics83, 4069–4074 (1985)

    1985

  3. [11]

    G. J. Martyna, M. L. Klein, and M. Tuckerman, Nos´ e–Hoover chains: The canonical ensemble via con- tinuous dynamics, The Journal of Chemical Physics97, 2635–2643 (1992)

    1992

  4. [12]

    Schneider and E

    T. Schneider and E. Stoll, Molecular-dynamics study of a three-dimensional one-component model for distortive phase transitions, Physical Review B17, 1302–1322 (1978)

    1978

  5. [13]

    Grønbech-Jensen, Complete set of stochastic Verlet- type thermostats for correct Langevin simulations, Molecular Physics118, e1662506 (2019)

    N. Grønbech-Jensen, Complete set of stochastic Verlet- type thermostats for correct Langevin simulations, Molecular Physics118, e1662506 (2019)

    2019

  6. [14]

    Espa˜ nol and P

    P. Espa˜ nol and P. Warren, Statistical mechanics of dissi- pative particle dynamics, Europhysics Letters (EPL)30, 191–196 (1995)

    1995

  7. [15]

    C. P. Lowe, An alternative approach to dissipative par- ticle dynamics, Europhysics Letters (EPL)47, 145–151 (1999)

    1999

  8. [16]

    Soddemann, B

    T. Soddemann, B. D¨ unweg, and K. Kremer, Dissipa- tive particle dynamics: A useful thermostat for equi- librium and nonequilibrium molecular dynamics simula- tions, Physical Review E68, 10.1103/physreve.68.046702 (2003)

    2003 doi

  9. [17]

    Ceriotti, G

    M. Ceriotti, G. Bussi, and M. Parrinello, Langevin equa- tion with colored noise for constant-temperature molec- ular dynamics simulations, Physical Review Letters102, 10.1103/physrevlett.102.020601 (2009)

    2009 doi

  10. [18]

    Parrinello and A

    M. Parrinello and A. Rahman, Polymorphic transitions in single crystals: A new molecular dynamics method, Journal of Applied Physics52, 7182–7190 (1981)

    1981

  11. [19]

    M. H. M¨ user, Dissipative split-charge formalism: Ohm’s law, Nyquist noise, and non-contact friction, The Journal of Chemical Physics161, 10.1063/5.0242185 (2024)

    2024 doi

  12. [20]

    Bussi, D

    G. Bussi, D. Donadio, and M. Parrinello, Canonical sam- pling through velocity rescaling, The Journal of Chemical Physics126, 10.1063/1.2408420 (2007)

    2007 doi

  13. [21]

    Parrinello and A

    M. Parrinello and A. Rahman, Strain fluctuations and elastic constants, The Journal of Chemical Physics76, 2662–2666 (1982)

    1982

  14. [22]

    Vattulainen, M

    I. Vattulainen, M. Karttunen, G. Besold, and J. M. Pol- son, Integration schemes for dissipative particle dynam- ics simulations: From softly interacting systems towards hybrid models, The Journal of Chemical Physics116, 3967–3979 (2002)

    2002

  15. [23]

    Ceriotti, G

    M. Ceriotti, G. Bussi, and M. Parrinello, Colored-noise thermostats ` a la carte, Journal of Chemical Theory and Computation6, 1170–1180 (2010)

    2010

  16. [24]

    Kubo, The fluctuation-dissipation theorem, Reports on Progress in Physics29, 255–284 (1966)

    R. Kubo, The fluctuation-dissipation theorem, Reports on Progress in Physics29, 255–284 (1966)

    1966

  17. [25]

    Car and M

    R. Car and M. Parrinello, Unified approach for molecular dynamics and density-functional theory, Physical Review Letters55, 2471–2474 (1985)

    1985

  18. [26]

    R. D. Skeel and J. A. Izaguirre, An impulse integrator for Langevin dynamics, Molecular Physics100, 3885 (2002)

    2002

  19. [27]

    Ricci and G

    A. Ricci and G. Ciccotti, Algorithms for Brownian dy- namics, Molecular Physics101, 1927–1931 (2003)

    2003

  20. [28]

    Melchionna, Design of quasisymplectic propagators for Langevin dynamics, The Journal of Chemical Physics 127, 10.1063/1.2753496 (2007)

    S. Melchionna, Design of quasisymplectic propagators for Langevin dynamics, The Journal of Chemical Physics 127, 10.1063/1.2753496 (2007)

    2007 doi

  21. [29]

    Vanden-Eijnden and G

    E. Vanden-Eijnden and G. Ciccotti, Second-order in- tegrators for Langevin equations with holonomic con- straints, Chemical Physics Letters429, 310 (2006)

    2006

  22. [30]

    N. Goga, A. Rzepiela, A. De Vries, S. Marrink, and H. Berendsen, Efficient algorithms for Langevin and DPD dynamics, Journal of Chemical Theory and Computation 8, 3637 (2012)

    2012

  23. [31]

    J. A. Izaguirre, D. P. Catarello, J. M. Wozniak, and R. D. Skeel, Langevin stabilization of molecular dynamics, The Journal of Chemical Physics114, 2090 (2001)

    2001

  24. [32]

    Ruiz-Franco, L

    J. Ruiz-Franco, L. Rovigatti, and E. Zaccarelli, On the effect of the thermostat in non-equilibrium molecular dy- namics simulations, The European Physical Journal E 41, 1 (2018)

    2018

  25. [33]

    Schwerdtfeger and D

    P. Schwerdtfeger and D. J. Wales, 100 years of the Lennard-Jones potential, Journal of Chemical Theory and Computation20, 3379–3405 (2024)

    2024

  26. [34]

    H. Gao, T. P. W. Menzel, M. H. M¨ user, and D. Mukherji, Comparing simulated specific heat of liquid polymers and oligomers to experiments, Physical Review Materials5, 10.1103/physrevmaterials.5.065605 (2021)

    2021 doi

  27. [35]

    Grønbech-Jensen and O

    N. Grønbech-Jensen and O. Farago, A simple and effec- tive Verlet-type algorithm for simulating Langevin dy- namics, Molecular Physics111, 983–991 (2013)

    2013

  28. [36]

    Farago, Langevin thermostat for robust configura- tional and kinetic sampling, Physica A: Statistical Me- chanics and Its Applications534, 122210 (2019)

    O. Farago, Langevin thermostat for robust configura- tional and kinetic sampling, Physica A: Statistical Me- chanics and Its Applications534, 122210 (2019)

    2019

  29. [37]

    M. H. M¨ user, S. V. Sukhomlinov, and L. Pastewka, In- teratomic potentials: achievements and challenges, Ad- vances in Physics: X8, 10.1080/23746149.2022.2093129 (2022)

    2022

  30. [38]

    Bugnicourt, P

    R. Bugnicourt, P. Sainsot, N. Lesaffre, and A. Lubrecht, Transient frictionless contact of a rough rigid surface on a viscoelastic half-space, Tribology International113, 279–285 (2017)

    2017

  31. [39]

    Sukhomlinov and M

    S. Sukhomlinov and M. H. M¨ user, On the viscous dissipa- tion caused by randomly rough indenters in smooth slid- ing motion, Applied Surface Science Advances6, 100182 (2021). 18

    2021

  32. [40]

    P. K. Patra and B. Bhattacharya, Nonergodicity of the nose-hoover chain thermostat in computationally achievable time, Physical Review E90, 10.1103/phys- reve.90.043304 (2014)

    2014 doi

  33. [41]

    Mukherji, M

    D. Mukherji, M. M¨ uller, and M. H. M¨ user, Computing finite–temperature elastic constants with noise cancella- tion (2025), in preparation

    2025

  34. [42]

    D. A. Puts, Beauty and the beast: mechanisms of sexual selection in humans, Evolution and Human Behavior31, 157–175 (2010)

    2010

  35. [43]

    A. P. Thompson, H. M. Aktulga, R. Berger, D. S. Bolin- tineanu, W. M. Brown, P. S. Crozier, P. J. in ’t Veld, A. Kohlmeyer, S. G. Moore, T. D. Nguyen, R. Shan, M. J. Stevens, J. Tranchida, C. Trott, and S. J. Plimpton, LAMMPS - a flexible simulation tool for particle-based mat...

    2022

  36. [44]

    L. T. Kong, G. Bartels, C. Campa˜ n´ a, C. Denniston, and M. H. M¨ user, Implementation of Green’s function molecular dynamics: An extension to lammps, Computer Physics Communications180, 1004–1010 (2009)

    2009

  37. [45]

    Grabowski, L

    B. Grabowski, L. Ismer, T. Hickel, and J. Neugebauer, Ab initio up to the melting point: Anharmonicity and vacancies in aluminum, Physical Review B—Condensed Matter and Materials Physics79, 134106 (2009)

    2009

  38. [46]

    R. G. Della Valle and E. Venuti, Quasiharmonic lattice- dynamics and molecular-dynamics calculations for the Lennard-Jones solids, Physical Review B58, 206 (1998)

    1998

  39. [47]

    Jalkanen and M

    J. Jalkanen and M. H. M¨ user, Systematic analysis and modification of embedded-atom potentials: case study of copper, Modelling and Simulation in Materials Science and Engineering23, 074001 (2015)

    2015

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.