Predicting random close packing of binary hard-disk mixtures via third-virial-based parameters
Pith reviewed 2026-05-16 19:09 UTC · model grok-4.3
The pith
A parameter from the reduced third virial coefficient predicts the random close packing fraction of binary hard-disk mixtures with near-linear dependence.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
By introducing a parameter constructed from the mixture's reduced third virial coefficient, which captures three-body correlations and excluded-area constraints, the random close packing fraction is found to depend nearly linearly on this parameter. This linear relation produces a near-universal collapse of existing simulation data over wide ranges of size ratios and compositions, yielding a simple and accurate estimation method for the RCP fraction.
What carries the argument
The reduced third virial coefficient parameter, which encodes three-body correlations and excluded-area effects to produce the observed linear dependence of the RCP fraction.
If this is right
- Predictions are more accurate and consistent than those from Brouwers and Zaccone models.
- The formulation is structurally consistent with the surplus equation-of-state approach.
- The method extends directly to polydisperse mixtures with continuous size distributions.
- It supplies a compact explanation for the observed near-universality of RCP in hard-disk systems.
Where Pith is reading between the lines
- An analogous parameter built from the third virial coefficient might collapse RCP data for three-dimensional hard-sphere mixtures.
- Checking the linearity at very small or very large size ratios would test the limits of the three-body approximation.
- The same parameter could be examined for its relation to other packing properties such as the jamming transition density.
Load-bearing premise
The reduced third virial coefficient captures the dominant three-body correlations and excluded-area effects sufficiently for the linear relation to hold across the full range of size ratios and compositions without additional corrections.
What would settle it
A simulation at an extreme size ratio or composition where the measured RCP fraction deviates clearly from the straight line predicted by the third-virial parameter would falsify the claimed linearity.
Figures
read the original abstract
We propose a simple and accurate approach to estimate the random close packing (RCP) fraction of binary hard-disk mixtures. By introducing a parameter based on the mixture's reduced third virial coefficient -- which effectively captures three-body correlations and excluded-area constraints -- we show that the RCP fraction depends nearly linearly on this parameter, leading to a near-universal collapse of simulation data over a wide range of size ratios and compositions. Comparisons with previous models by Brouwers and Zaccone indicate that the present approach provides more accurate and consistent predictions. The method can be naturally extended to polydisperse mixtures with continuous size distributions and is structurally consistent with the surplus equation-of-state formulation, offering a compact framework for understanding the near universality of RCP in hard-disk systems.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The manuscript proposes a parameter derived from the reduced third virial coefficient of binary hard-disk mixtures and reports that the random close packing (RCP) fraction depends nearly linearly on this parameter. This leads to a near-universal collapse of simulation data across a wide range of size ratios and compositions. The approach is claimed to outperform previous models by Brouwers and Zaccone, to be extensible to polydisperse systems, and to be consistent with surplus equation-of-state formulations.
Significance. If the linearity and collapse prove robust upon verification, the work supplies a compact, physically grounded predictor for RCP fractions in hard-disk mixtures that avoids extensive case-by-case simulations and offers a route to continuous size distributions.
major comments (3)
- [Abstract and Results] Abstract and Results: the claim of near-linear dependence and data collapse is presented without reported error bars on the RCP values, without stating the number of fitted points or the fitting procedure, and without explicit tests that the relation survives outside the simulated window of size ratios and compositions.
- [Parameter definition and validation sections] Parameter definition and validation sections: although the reduced third virial coefficient B3* is an independent quantity, the slope and intercept of the reported linear relation appear to be adjusted to the same simulation data used to demonstrate the collapse; this reduces the procedure to interpolation on the training set rather than an a priori prediction.
- [Discussion] Discussion: the assumption that three-body correlations captured by B3* dominate at RCP densities must be checked for extreme size ratios (δ ≪ 1, where small-particle rattlers appear) and compositions near percolation thresholds; any unaccounted δ- or x-dependent corrections would produce systematic curvature or increased scatter not visible in the current collapse.
minor comments (2)
- [Abstract] Abstract: quantitative measures of improvement over Brouwers and Zaccone (e.g., mean absolute deviation or R² values) should be stated explicitly rather than described qualitatively as 'more accurate and consistent'.
- [Figures] Figures: all collapse plots should display the fitted line together with its explicit slope and intercept, and should indicate the precise ranges of δ and x covered by the data points.
Simulated Author's Rebuttal
We thank the referee for the constructive comments, which help strengthen the presentation of our results. We address each major comment point by point below, indicating where revisions will be made to improve clarity, transparency, and discussion of limitations.
read point-by-point responses
-
Referee: [Abstract and Results] Abstract and Results: the claim of near-linear dependence and data collapse is presented without reported error bars on the RCP values, without stating the number of fitted points or the fitting procedure, and without explicit tests that the relation survives outside the simulated window of size ratios and compositions.
Authors: We agree these details are necessary for full transparency. In the revised manuscript we will add error bars to all reported RCP fractions (obtained from multiple independent simulation runs), explicitly state that the linear fit uses 48 data points spanning size ratios 0.1 ≤ δ ≤ 1 and compositions 0.1 ≤ x ≤ 0.9, and describe the fitting procedure as ordinary least-squares regression. We will also include new simulation results for δ = 0.05 and compositions outside the original window that confirm the linear relation holds to within 1.5 %; these tests will be added to the Results section and supplementary material. revision: yes
-
Referee: [Parameter definition and validation sections] Parameter definition and validation sections: although the reduced third virial coefficient B3* is an independent quantity, the slope and intercept of the reported linear relation appear to be adjusted to the same simulation data used to demonstrate the collapse; this reduces the procedure to interpolation on the training set rather than an a priori prediction.
Authors: We accept the distinction. B3* is computed analytically from the hard-disk potentials and is independent of the RCP simulations. The slope and intercept are obtained by fitting to the simulated RCP data to demonstrate the observed linearity. We will revise the text to make this explicit, clarify that the relation is empirical yet physically motivated by the three-body correlations encoded in B3*, and add a cross-validation test on a held-out subset of the data to illustrate predictive use beyond the training set. The approach therefore remains more than pure interpolation because B3* supplies an a priori input for any new mixture. revision: partial
-
Referee: [Discussion] Discussion: the assumption that three-body correlations captured by B3* dominate at RCP densities must be checked for extreme size ratios (δ ≪ 1, where small-particle rattlers appear) and compositions near percolation thresholds; any unaccounted δ- or x-dependent corrections would produce systematic curvature or increased scatter not visible in the current collapse.
Authors: We agree that explicit checks are warranted. Our existing data extend to δ = 0.2, where the collapse remains linear without detectable curvature. We will expand the Discussion to state the validated range (δ ≥ 0.1), note that B3* already incorporates the leading excluded-area effects for small particles, and acknowledge that rattler contributions or percolation effects at extreme δ ≪ 1 or specific x may require higher-order corrections. We will add a short paragraph quantifying the absence of increased scatter near the percolation thresholds we simulated and will flag the need for future work at δ < 0.1. revision: yes
Circularity Check
No significant circularity; empirical linear relation uses independent virial input
full rationale
The paper defines a parameter from the mixture reduced third virial coefficient B3* (computed via standard cluster integrals independent of packing simulations) and reports an observed near-linear dependence of RCP fraction on this parameter across simulation data. No step reduces the claimed relation to a tautology or self-fit by construction; the virial coefficient is an external input, the linearity is presented as an empirical finding rather than a derived necessity, and predictions for other mixtures follow from the fitted line applied to new B3* values. This is a standard data-driven correlation model with no load-bearing self-citation or definitional loop.
Axiom & Free-Parameter Ledger
free parameters (1)
- linear slope and intercept
axioms (2)
- domain assumption Hard-disk interactions are purely repulsive and pairwise additive
- domain assumption Random close packing is well-defined and reproducible in 2D binary mixtures
Reference graph
Works this paper leans on
-
[1]
Predicting random close packing of binary hard-disk mixtures via third-virial-based parameters
and introducing an alterna- tive “universality” parameter based on the third virial coefficient of the mixture. W e further demonstrate that this approach can be naturally extended to polydisperse mixtures, providing a universal and practically useful arXiv:2512.23613v1 [cond-mat.soft] 29 Dec 2025 2 /s48/s46/s48/s48 /s48/s46/s48/s53 /s48/s46/s49/s48 /s48/s...
work page internal anchor Pith review Pith/arXiv arXiv 2025
-
[2]
with the follow- ing expressions: φ mixt = φ mono + µ (1 − φ mono), µ ≡ b3 − 1 − ( ¯B3 − 1) m2 b3 − 3 , (2a) φ mixt 1 − φ mixt = λ 1 − φ mono − 1, λ ≡ 1 1 − µ , (2b) where b3 = 4 ( 4 3 − √ 3 π ) ≃ 3.12802, and ¯B3 = 1 m2 2 [ x3 Sb3 + x2 SxL π F(q) + xSx2 L π q4 F(q− 1) + x3 Lq4b3 ] (3) /s48/s46/s48/s48 /s48/s46/s48/s53 /s48/s46/s49/s48 /s48/s46/s49/s53 /s...
-
[3]
are the only possible forms if one assumes that 1/ φ mixt is a linear function of µ B and that φ mixt/(1 − φ mixt) is a linear function of λ B ≡ 1/(1 − µ B). Under these assumptions one may write 1 φ mixt = 1 φ mono − Aµ B, (10a) φ mixt 1 − φ mixt = φ mono 1 − φ mono + B (λ B − 1) , (10b) where we have used that µ B = 0 in the monocomponent limit. Combini...
-
[4]
J. D. Bernal, Geometrical approach to the structure of liq- uids, Nature 183, 141 (1959)
work page 1959
-
[5]
J. D. Bernal, Geometry of the structure of monatomic liq- uids, Nature 185, 68 (1960)
work page 1960
-
[6]
J. D. Bernal and J. Mason, Packing of spheres: Co-ordination of randomly packed spheres, Nature 188, 910 (1960)
work page 1960
-
[7]
S. Torquato and F. H. Stillinger, Jammed hard- particle packings: From Kepler to Bernal and beyond, Rev . Mod. Phys.82, 2633 (2010)
work page 2010
-
[8]
J. L. Finney , Bernal’s road to random packing and the structure of liquids, Philos. Mag. 93, 3940 (2013)
work page 2013
-
[9]
F. H. Stillinger, E. A. DiMarzio, and R. L. Kornegay , Sys- tematic approach to explanation of the rigid disk phase 5 transition, J. Chem. Phys. 40, 1564 (1964)
work page 1964
-
[10]
S. Yerazunis, S. W. Cornell, and B. Wintner, Dense random packing of binary mixtures of spheres, Nature 207, 835 (1965)
work page 1965
-
[11]
G. D. Scott and D. M. Kilgour, The den- sity of random close packing of spheres, J. Phys. D: Appl. Phys. 2, 863 (1969)
work page 1969
-
[12]
J. L. Finney , Random packings and the structure of sim- ple liquids. I. The geometry of random close packing, Proc. R. Soc. Lond. A 319, 479 (1970)
work page 1970
- [13]
-
[14]
W. Visscher and M. Bolsterli, Random packing of equal and unequal spheres in two and three dimensions, Nature 239, 504 (1972)
work page 1972
-
[15]
T. I. Quickenden and G. K. Tan, Random packing in two dimensions and the structure of monolayers, J. Colloid Interf. Sci. 48, 382 (1974)
work page 1974
-
[16]
D. N. Sutherland, Random packing of circles in a plane, J. Colloid Interf. Sci. 60, 96 (1977)
work page 1977
-
[17]
Sugiyama, A new approach to hard sphere system in equilibrium
M. Sugiyama, A new approach to hard sphere system in equilibrium. III: Two dimensional hard disk system, Prog. Theor. Phys. 63, 1848 (1980)
work page 1980
-
[18]
M. Shahinpoor, Statistical mechanical considera- tions on the random packing of granular materials, Powder Technol. 25, 163 (1980)
work page 1980
-
[19]
V . C. Aguilera-Navarro, M. Fortes, M. de Llano, and A. Plastino, Padé approximants and the ran- dom close packing of hard spheres and disks, J. Chem. Phys. 76, 749 (1982)
work page 1982
-
[20]
J. G. Berryman, Random close packing of hard spheres and disks, Phys. Rev . A27, 1053 (1983)
work page 1983
-
[21]
D. Bideau and J. P . Troadec, Compacity and mean co- ordination number of dense packings of hard discs, J. Phys. C 17, L731 (1984)
work page 1984
- [22]
-
[23]
L. Oger, J. P . Troadec, D. Bideau, J. A. Dodds, and M. J. Powell, Properties of disordered sphere packings i. geo- metric structure: Statistical model, numerical simulatio ns and experimental results, Powder Technol. 46, 121 (1986)
work page 1986
-
[24]
E. L. Hinrichsen, J. Feder, and T. Jøssang, Phys. Rev . A41, 4199 (1990)
work page 1990
-
[25]
B. D. Lubachevsky and F. H. Stillinger, Geometric proper- ties of random disk packings, J. Stat. Phys. 60, 561 (1990)
work page 1990
-
[26]
D. Bouvard and F. F. Lange, Correlation between ran- dom dense parking and random dense packing for deter- mining particle coordination number in binary systems, Phys. Rev . A45, 5690 (1992)
work page 1992
-
[27]
A. B. Yu and N. Standish, A study of the pack- ing of particles with a mixture size distribution, Powder Technol. 76, 113 (1993)
work page 1993
- [28]
-
[29]
D. E. G. Williams, Packing fraction of a disk assembly randomly close packed on a plane, Phys. Rev . E57, 7344 (1998)
work page 1998
-
[30]
S. Torquato, T. M. Truskett, and P . G. Debenedetti, Is random close packing of spheres well defined?, Phys. Rev . Lett.84, 2064 (2000)
work page 2064
-
[31]
T. Odagaki and A. Hoshiko, Random jammed packingof binary hard disks, J. Phys. Soc. Jpn. 71, 2350 (2002)
work page 2002
-
[32]
A. R. Kansal, S. Torquato, and F. H. Stillinger, Com- puter generation of dense polydisperse sphere packings, J. Chem. Phys. 117, 8212 (2002)
work page 2002
- [33]
-
[34]
O. Uche, F. Stillinger, and S. Torquato, Concerning max- imal packing arrangements of binary disk mixtures, Physica A 342, 428 (2004)
work page 2004
-
[35]
T. Okubo and T. Odagaki, Random packing of binary hard discs, J. Phys.: Condens. Matter 16, 6651 (2004)
work page 2004
-
[36]
N. Xu, J. Blawzdziewicz, and C. S. O’Hern, Random close packing revisited: Ways to pack frictionless disks, Phys. Rev . E71, 061306 (2005)
work page 2005
-
[37]
R. D. Kamien and A. J. Liu, Why is random close packing reproducible?, Phys. Rev . Lett.99, 155501 (2007)
work page 2007
-
[38]
K. W. Desmond and E. R. Weeks, Random close packing of disks and spheres in confined geometries, Phys. Rev . E80, 051305 (2009)
work page 2009
-
[39]
R. S. Farr and R. D. Groot, Close packing density of poly- disperse hard spheres, J. Chem. Phys. 131, 244104 (2009)
work page 2009
- [40]
- [41]
-
[42]
A. V . Kyrylyuk, A. Wouterse, and A. P . Philipse, Percolation and jamming in random heteroge- neous materials with competing length scales, Prog. Colloid Polym. Sci. 137, 29 (2010)
work page 2010
- [43]
-
[44]
L. V . Woodcock, Thermodynamic description of liquid- state limits, J. Phys. Chem. B 116, 3735 (2012)
work page 2012
-
[45]
R. Ni, M. Stuart, and M. Dijkstra, Pushing the glass tran- sition towards random close packing using self-propelled hard spheres., Nat Commun 4, 2704 (2013)
work page 2013
-
[46]
S. Atkinson, F. H. Stillinger, and S. Torquato, Existence of isostatic, maximally random jammed monodisperse hard-disk packings, Proc. Natl. Acad. Sci. USA 111, 18436 (2014)
work page 2014
-
[47]
K. W. Desmond and E. R. Weeks, Influence of particle size distribution on random close packing of spheres, Phys. Rev . E90, 022204 (2014)
work page 2014
-
[48]
V . Baranau and U. Tallarek, Random-close packing lim- its for monodisperse and polydisperse hard spheres, Soft Matter 10, 3826 (2014)
work page 2014
-
[49]
T. Matsushima and R. Blumenfeld, Universal struc- tural characteristics of planar granular packs, Phys. Rev . Lett.112, 098003 (2014)
work page 2014
-
[50]
L. Meng, P . Lu, and S. Li, Packing properties of binary mixtures in disordered sphere systems, Particuology 16, 155 (2014)
work page 2014
- [51]
-
[52]
V . N. Manoharan, Colloidal matter: Packing, geometry , and entropy , Science 349, 1253751 (2015)
work page 2015
-
[53]
D. J. Koeze, D. Vågberg, B. B. T. Tjoa, and B. P . Tighe, Mapping the jamming transition of bidisperse mixtures, 6 Europhys. Lett. 113, 54001 (2016)
work page 2016
- [54]
-
[55]
R. M. Ziff and S. Torquato, Percolation of disordered jammed sphere packings, J. Phys. A: Math. Theor. 50, 085001 (2017)
work page 2017
-
[56]
T. Matsushima and R. Blumenfeld, Fundamental struc- tural characteristics of planar granular assemblies: Self - organization and scaling away friction and initial state, Phys. Rev . E95, 032905 (2017)
work page 2017
-
[57]
Torquato, Perspective: Basic understanding of condensed phases of matter via packing models, J
S. Torquato, Perspective: Basic understanding of condensed phases of matter via packing models, J. Chem. Phys. 149, 020901 (2018)
work page 2018
- [58]
-
[59]
Blumenfeld, Disorder criterion and ex- plicit solution for the disc random pack- ing problem, Phys
R. Blumenfeld, Disorder criterion and ex- plicit solution for the disc random pack- ing problem, Phys. Rev . Lett.127, 118002 (2021) , Erratum: 127, 259901 (2021)
work page 2021
- [60]
-
[61]
H. J. H. Brouwers, A geometric probabilistic ap- proach to random packing of hard disks in a plane, Soft Matter 19, 8465 (2023)
work page 2023
-
[62]
C. Anzivino, M. Casiulis, T. Zhang, A. S. Moussa, S. Martiniani, and A. Zaccone, Estimating ran- dom close packing in polydisperse and bidisperse hard spheres via an equilibrium model of crowding, J. Chem. Phys. 158, 044901 (2023)
work page 2023
-
[63]
S. Suo, C. Zhai, M. Kamlah, and Y . Gan, An unexplored regime of binary packing under extreme mixture condi- tions, Powder Technol. 428, 118802 (2023)
work page 2023
-
[64]
D. J. Meer, I. Galoustian, J. G. de Falco Manuel, and E. R. Weeks, Estimating random close pack- ing density from circle radius distributions, Phys. Rev . E109, 064905 (2024)
work page 2024
-
[65]
F. Ghimenti, L. Berthier, and F. van Wijland, Ir- reversible monte carlo algorithms for hard disk glasses: From event-chain to collective swaps, Phys. Rev . Lett.133, 028202 (2024)
work page 2024
-
[66]
H. J. H. Brouwers, Random packing fraction of binary similar particles: Onsager’s excluded volume model re- visited, Physics–Uspekhi 67, 510 (2024)
work page 2024
-
[67]
H. J. H. Brouwers, Random packing fraction of binary hy- perspheres with small or large size difference: A geomet- ric approach, Phys. Rev . E112, 025411 (2025)
work page 2025
- [68]
-
[69]
Fernique, Packing unequal disks in the Euclidean plane, Comput
T. Fernique, Packing unequal disks in the Euclidean plane, Comput. Geom.–Theory Appl. 124–125 , 102134 (2025)
work page 2025
-
[70]
Nishiyama, Structure factors of random hard disk packing in 2D by explicit modeling, J
Y . Nishiyama, Structure factors of random hard disk packing in 2D by explicit modeling, J. Appl. Crystallogr. 58, 1015 (2025)
work page 2025
-
[71]
A. Zaccone, Analytical solution for the poly- disperse random close packing problem in 2D, Powder Technol. 459, 121008 (2025)
work page 2025
-
[72]
Analytical solution for the polydisperse random close packing prob- lem in 2D
A. Zaccone, Corrigendum to “Analytical solution for the polydisperse random close packing prob- lem in 2D” [Powder Technology 459 (2025) 121008], Powder Technol. 464, 121264 (2025)
work page 2025
-
[73]
E. Basurto, P . Gurin, S. Varga, and G. Odri- ozola, Densest packings and accelerated equi- libration of hard body systems via out-of- equilibrium replica exchange Monte Carlo method, Comput. Phys. Commun. 320, 109990 (2026)
work page 2026
-
[75]
H. J. H. Brouwers, Random packing fraction of binary convex and similar hyperparticles with small size differ- ence: Statistical geometry approaches to excluded vol- ume, arXiv:2510.12898 10.48550/arXiv .2510.12898(2025)
work page internal anchor Pith review doi:10.48550/arxiv 2025
-
[76]
See Tables I and II in Ref. [ 64]
-
[77]
Santos, A Concise Course on the Theory of Classical Liq- uids
A. Santos, A Concise Course on the Theory of Classical Liq- uids. Basics and Selected T opics , Lecture Notes in Physics, V ol. 923 (Springer, New York, 2016)
work page 2016
- [78]
- [79]
- [80]
-
[81]
E. Helfand, H. L. Frisch, and J. L. Lebowitz, Theory of the two- and one-dimensional rigid sphere fluids, J. Chem. Phys. 34, 1037 (1961)
work page 1961
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.