pith. sign in

arxiv: 2605.28446 · v1 · pith:WHW4OC65new · submitted 2026-05-27 · 💻 cs.CE

Modelling the effect of fiber distribution on the transverse mechanical characteristics of unidirectionally reinforced continuous-fiber composite

Sergejs Tarasovs , Janis Modniks , Andrea Bercini Martins , Christina Scheffler , Janis Andersons This is my paper

Pith reviewed 2026-06-29 09:31 UTC · model grok-4.3

classification 💻 cs.CE
keywords fiber distributiontransverse mechanical propertiesunidirectional compositesrepresentative volume elementnearest neighbor distancefinite element homogenizationglass/epoxy
0
0 comments X

The pith

Clustered fiber distributions raise transverse stiffness but lower tensile strength in unidirectional composites at fixed volume fraction.

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

The paper investigates how the spatial arrangement of fibers influences the mechanical response perpendicular to the fiber direction in continuous-fiber composites. Models with controlled arrangements from clustered to evenly distributed show that clustering increases stiffness while reducing strength. A single measure, the mean nearest neighbor distance, captures the effect on both properties through linear relations. This indicates that fiber placement during manufacturing can be used to adjust transverse performance without altering fiber content.

Core claim

At constant fiber volume fraction, clustered fiber distributions yield significantly higher transverse stiffness but lower transverse tensile strength compared to the equilibrium distributions. For glass/epoxy composites, transverse stiffness varies by up to 20% depending on the degree of fiber clustering. A single scalar descriptor, the mean nearest neighbor distance, was shown to efficiently characterize sufficiently random fiber distributions: effective stiffness decreases, whereas transverse tensile strength increases linearly with mean nearest neighbor distance.

What carries the argument

Swelling & Random Migration algorithm generating representative volume elements with fiber arrangements from clustered to equilibrium, combined with finite element homogenization under periodic boundary conditions.

If this is right

  • Transverse stiffness in glass/epoxy composites can change by as much as 20 percent solely due to the degree of fiber clustering.
  • Effective transverse stiffness decreases linearly as mean nearest neighbor distance increases.
  • Transverse tensile strength increases linearly as mean nearest neighbor distance increases.
  • A single scalar metric suffices to characterize random fiber distributions for property prediction.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Processing methods that control fiber migration could be tuned to target specific transverse stiffness or strength values.
  • The linear relations observed might be checked for validity in other fiber-matrix combinations or at different volume fractions.
  • Design models for composites could incorporate mean nearest neighbor distance as an input variable for transverse property estimation.

Load-bearing premise

The fiber arrangements produced by the Swelling & Random Migration algorithm are statistically equivalent to experimental microstructures when checked with nearest neighbor distance, Ripley's K-function, pair distribution function, and local fiber volume fraction.

What would settle it

Direct experimental tests on glass/epoxy specimens with controlled fiber clustering at fixed volume fraction that find no measurable difference in transverse stiffness or strength.

read the original abstract

This study investigates the influence of fiber spatial distribution on the transverse mechanical properties of unidirectionally reinforced continuous-fiber composites. A Swelling & Random Migration algorithm was employed to generate representative volume elements with controlled fiber arrangements, ranging from clustered to equilibrium configurations. Finite element homogenization with periodic boundary conditions was used to estimate effective elastic properties. To characterize fiber randomness and assess statistical equivalence with experimental microstructures, several descriptors are employed, including nearest neighbor distance, Ripley's K-function, pair distribution function, and local fiber volume fraction. Results reveal that, at constant fiber volume fraction, clustered fiber distributions yield significantly higher transverse stiffness but lower transverse tensile strength compared to the equilibrium distributions. For glass/epoxy composites, transverse stiffness varies by up to 20% depending on the degree of fiber clustering. A single scalar descriptor, the mean nearest neighbor distance, was shown to efficiently characterize sufficiently random fiber distributions: effective stiffness decreases, whereas transverse tensile strength increases linearly with mean nearest neighbor distance. The findings highlight the critical role of microstructural characteristics in tailoring composite performance and provide a robust framework for predictive modeling of fiber reinforced materials.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript uses a Swelling & Random Migration algorithm to generate RVEs of unidirectional fiber composites at fixed volume fraction with fiber arrangements ranging from clustered to equilibrium. Finite-element homogenization with periodic boundary conditions computes transverse elastic stiffness and tensile strength; statistical descriptors (nearest-neighbor distance, Ripley’s K-function, pair-distribution function, local fiber volume fraction) are used to argue that the generated RVEs are representative of experimental microstructures. The central claims are that clustered distributions produce up to 20 % higher transverse stiffness but lower strength than equilibrium distributions, and that mean nearest-neighbor distance alone linearly correlates with both properties (stiffness decreases, strength increases).

Significance. If the generated RVEs are mechanically equivalent to real microstructures, the reported 20 % stiffness variation and the linear NND–property relations would be useful for microstructure-sensitive design of transverse composite performance. The multi-descriptor validation approach is a methodological strength.

major comments (2)
  1. [Abstract] Abstract and results section: the quantitative claims (20 % stiffness variation, linear NND–stiffness and NND–strength trends) rest on the premise that matching the listed second-order spatial statistics guarantees mechanical equivalence. The descriptors do not constrain higher-order local packing geometry that governs stress concentrations and therefore transverse strength; without additional validation (e.g., direct comparison of simulated stress fields or strength to experimental data on the same material system), the magnitude and linearity of the reported effects remain generation-specific.
  2. [Methods] Methods / results: the manuscript does not report mesh-convergence studies, representative volume-element size sensitivity, or statistical error bars on the extracted effective properties. These omissions make it impossible to assess whether the stated 20 % stiffness difference exceeds numerical uncertainty.
minor comments (2)
  1. [Abstract] Specify the exact fiber volume fraction and constituent properties (glass/epoxy moduli, strengths) used for the 20 % claim so that the results can be reproduced or compared with other studies.
  2. [Results] Clarify whether the linear NND–property relations are obtained by regression across all generated RVEs or only within the “sufficiently random” subset; state the R² values.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments on our manuscript. We address each major point below and indicate the revisions we will make.

read point-by-point responses

  1. Referee: [Abstract] Abstract and results section: the quantitative claims (20 % stiffness variation, linear NND–stiffness and NND–strength trends) rest on the premise that matching the listed second-order spatial statistics guarantees mechanical equivalence. The descriptors do not constrain higher-order local packing geometry that governs stress concentrations and therefore transverse strength; without additional validation (e.g., direct comparison of simulated stress fields or strength to experimental data on the same material system), the magnitude and linearity of the reported effects remain generation-specific.

    Authors: We agree that second-order descriptors (NND, Ripley’s K, pair-distribution function, local fiber volume fraction) do not fully constrain higher-order packing features that control local stress concentrations and strength. Our study focuses on the mechanical response of synthetically generated RVEs spanning controlled clustering levels rather than claiming direct equivalence to any specific experimental microstructure. The linear NND–property correlations are observed within the simulated ensemble; we will revise the abstract and discussion to explicitly qualify these trends as generation-specific and to note the potential influence of higher-order statistics not captured by the chosen descriptors. revision: partial

  2. Referee: [Methods] Methods / results: the manuscript does not report mesh-convergence studies, representative volume-element size sensitivity, or statistical error bars on the extracted effective properties. These omissions make it impossible to assess whether the stated 20 % stiffness difference exceeds numerical uncertainty.

    Authors: We agree that mesh-convergence, RVE-size sensitivity, and statistical error bars must be reported to substantiate the 20 % stiffness variation. These studies were conducted during the work but omitted from the manuscript. We will add dedicated subsections in Methods and Results documenting the convergence criteria, RVE-size independence tests, and error bars (standard deviation across multiple realizations) on all reported effective properties. revision: yes

Circularity Check

0 steps flagged

No significant circularity; results from direct FE simulation of generated RVEs

full rationale

The paper generates RVEs via the Swelling & Random Migration algorithm, then computes transverse stiffness and strength via finite-element homogenization under periodic BCs. Reported effects (up to 20% stiffness variation, linear NND trends) are direct numerical outputs, not reductions of predictions to fitted inputs or self-citations by construction. Statistical equivalence to experiments is assessed via standard descriptors (NND, Ripley's K, pair distribution, local Vf) without feeding mechanical results back into the generation or claiming uniqueness theorems. The derivation chain is therefore self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The modeling rests on standard assumptions from composite micromechanics and statistical geometry; the Swelling & Random Migration algorithm is the primary methodological choice whose parameters are not detailed in the abstract.

axioms (2)
  • domain assumption Finite element homogenization with periodic boundary conditions yields accurate effective transverse elastic properties for the generated RVEs.
    Invoked implicitly as the core computational method for property estimation.
  • domain assumption Statistical descriptors (nearest neighbor distance, Ripley's K-function, pair distribution function, local fiber volume fraction) sufficiently characterize fiber randomness and equivalence to experimental microstructures.
    Used to validate the generated distributions against real composites.

pith-pipeline@v0.9.1-grok · 5743 in / 1590 out tokens · 39584 ms · 2026-06-29T09:31:57.960777+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

38 extracted references · 35 canonical work pages · 1 internal anchor

  1. [1]

    Pyrz, Quantitative description of the microstructure of composites

    R. Pyrz, Quantitative description of the microstructure of composites. Part I: Morphology of unidirectional composite systems, Compos. Sci. Technol. 50(2) (1994) 197 -208. https://doi.org/10.1016/0266-3538(94)90141-4

    work page doi:10.1016/0266-3538(94)90141-4 1994

  2. [2]

    Dutra, R.T.L

    T.A. Dutra, R.T.L. Ferreira, H.B. Resende, L.M. Oliveira, B.J. Blinzler, L.E. Asp, Identification of Representative Equivalent Volumes on the Microstructure of 3D-Printed Fiber-Reinforced Thermoplastics Based on Statistical Characterization, Polymers 14(5) (2022) 972. https://doi.org/10.3390/polym14050972

    work page doi:10.3390/polym14050972 2022

  3. [3]

    Gusev, P .J

    A.A. Gusev, P .J. Hine, I.M. Ward, Fiber packing and elastic properties of a transversely random unidirectional glass/epoxy composite, Compos. Sci. Technol. 60(4) (2000) 535 -541. https://doi.org/10.1016/S0266-3538(99)00152-9

    work page doi:10.1016/s0266-3538(99)00152-9 2000

  4. [4]

    Vaughan, C.T

    T.J. Vaughan, C.T. McCarthy, A combined experimental–numerical approach for generating statistically equivalent fibre distributions for high strength laminated composite materials, Compos. Sci. Technol. 70(2) (2010) 291-297. https://doi.org/10.1016/j.compscitech.2009.10.020

    work page doi:10.1016/j.compscitech.2009.10.020 2010

  5. [5]

    Bednarcyk, J

    B.A. Bednarcyk, J. Aboudi, S.M. Arnold, Analysis of fiber clustering in composite materials using high - fidelity multiscale micromechanics, Int. J. Solids Struct. 69 -70 (2015) 311 -327. https://doi.org/10.1016/j.ijsolstr.2015.05.019

    work page doi:10.1016/j.ijsolstr.2015.05.019 2015

  6. [6]

    Schey, T

    M.J. Schey, T. Beke, L. Appel, S. Zabler, S. Shah, J. Hu, F. Liu, M. Maiaru, S. Stapleton, Identification and Quantification of 3D Fiber Clusters in Fiber-Reinforced Composite Materials, JOM 73(7) (2021) 2129-2142. https://doi.org/10.1007/s11837-021-04703-0

    work page doi:10.1007/s11837-021-04703-0 2021

  7. [7]

    X. Pang, F. Huang, F. Zhu, S. Zhang, Y . Wang, X. Chen, Progressive failure characteristics of unidirectional FRP with fiber clustering, Compos. Struct. 280 (2022) 114880. https://doi.org/10.1016/j.compstruct.2021.114880 23

    work page doi:10.1016/j.compstruct.2021.114880 2022

  8. [8]

    Husseini, E.J

    J.F. Husseini, E.J. Pineda, S.E. Stapleton, Generation of artificial 2 -D fiber reinforced composite microstructures with statistically equivalent features, Compos. Part A Appl. Sci. Manuf. 164 (2023) 107260. https://doi.org/10.1016/j.compositesa.2022.107260

    work page doi:10.1016/j.compositesa.2022.107260 2023

  9. [9]

    Petersen, Y

    H.N. Petersen, Y . Kusano, P . Brøndsted, K. Almdal, Preliminary characterization of glass fiber sizing, Proceedings of the Risø international symposium on materials science, Risø National Laboratory, 2013, pp. 333-340

    2013

  10. [10]

    Krauklis, A.I

    A.E. Krauklis, A.I. Gagani, A.T. Echtermeyer, Long -Term Hydrolytic Degradation of the Sizing -Rich Composite Interphase, Coatings 9(4) (2019) 263. https://doi.org/10.3390/coatings9040263

    work page doi:10.3390/coatings9040263 2019

  11. [11]

    Cech, Plasma -polymerized organosilicones as engineered interlayers in glass fiber/polyester composites, Compos

    V. Cech, Plasma -polymerized organosilicones as engineered interlayers in glass fiber/polyester composites, Compos. Interfaces 14(4) (2007) 321-334. https://doi.org/10.1163/156855407780452850

    work page doi:10.1163/156855407780452850 2007

  12. [12]

    Romanov, S.V

    V. Romanov, S.V. Lomov, Y . Swolfs, S. Orlova, L. Gorbatikh, I. Verpoest, Statistical analysis of real and simulated fibre arrangements in unidirectional composites, Compos. Sci. Technol. 87 (2013) 126 -134. https://doi.org/10.1016/j.compscitech.2013.07.030

    work page doi:10.1016/j.compscitech.2013.07.030 2013

  13. [13]

    K.C. Liu, A. Ghoshal, Validity of random microstructures simulation in fiber -reinforced composite materials, Compos. Part B Eng. 57 (2014) 56-70. https://doi.org/10.1016/j.compositesb.2013.08.006

    work page doi:10.1016/j.compositesb.2013.08.006 2014

  14. [14]

    W. Ge, L. Wang, Y . Sun, X. Liu, An efficient method to generate random distribution of fibers in continuous fiber reinforced composites, Polym. Compos. 40(12) (2019) 4763 -4770. https://doi.org/10.1002/pc.25344

    work page doi:10.1002/pc.25344 2019

  15. [15]

    Ghosh, Z

    S. Ghosh, Z. Nowak, K. Lee, Quantitative characterization and modeling of composite microstructures by Voronoi cells, Acta Mater. 45(6) (1997) 2215-2234. https://doi.org/10.1016/S1359-6454(96)00365-5

    work page doi:10.1016/s1359-6454(96)00365-5 1997

  16. [16]

    Elnekhaily, R

    S.A. Elnekhaily, R. Talreja, Damage initiation in unidirectional fiber composites with different degrees of nonuniform fiber distribution, Compos. Sci. Technol. 155 (2018) 22 -32. https://doi.org/10.1016/j.compscitech.2017.11.017

    work page doi:10.1016/j.compscitech.2017.11.017 2018

  17. [17]

    Nagaraj, M

    M.H. Nagaraj, M. Schey, T. Beke, S.E. Stapleton, M. Maiaru, Influence of microstructural variabilities on the mechanical properties of fiber -reinforced composites accounting for manufacturing effects, Compos. Struct. 373 (2025) 119631. https://doi.org/10.1016/j.compstruct.2025.119631

    work page doi:10.1016/j.compstruct.2025.119631 2025

  18. [18]

    Pathan, V.L

    M.V. Pathan, V.L. Tagarielli, S. Patsias, P .M. Baiz -Villafranca, A new algorithm to generate representative volume elements of composites with cylindrical or spherical fillers, Compos. Part B Eng. 110 (2017) 267-278. https://doi.org/10.1016/j.compositesb.2016.10.078

    work page doi:10.1016/j.compositesb.2016.10.078 2017

  19. [19]

    Melro, P .P

    A.R. Melro, P .P . Camanho, S.T. Pinho, Influence of geometrical parameters on the elastic response of unidirectional composite materials, Compos. Struct. 94(11) (2012) 3223 -3231. https://doi.org/10.1016/j.compstruct.2012.05.004

    work page doi:10.1016/j.compstruct.2012.05.004 2012

  20. [20]

    Zhang, Y

    T. Zhang, Y . Yan, A comparison between random model and periodic model for fiber -reinforced composites based on a new method for generating fiber distributions, Polym. Compos. 38(1) (2017) 77 -

    2017

  21. [21]

    https://doi.org/10.1002/pc.23562

    work page doi:10.1002/pc.23562

  22. [22]

    Efficient generation of large-scale non-equilibrium distributions of particles

    S. Tarasovs, Efficient generation of large -scale non -equilibrium distributions of particles, arXiv (arXiv:2605.18254) (2026). https://doi.org/10.48550/arXiv.2605.18254

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2605.18254 2026

  23. [23]

    Vila-Chã, B.P

    J.L.P . Vila-Chã, B.P . Ferreira, F.M.A. Pires, An adaptive multi-temperature isokinetic method for the RVE generation of particle reinforced heterogeneous materials, Part I: Theoretical formulation and computational framework, Mech. Mater. 163 (2021) 1 04069. https://doi.org/10.1016/j.mechmat.2021.104069

    work page doi:10.1016/j.mechmat.2021.104069 2021

  24. [24]

    Rintoul, S

    M.D. Rintoul, S. Torquato, Reconstruction of the Structure of Dispersions, J. Colloid Interface Sci. 186(2) (1997) 467-476. https://doi.org/10.1006/jcis.1996.4675

    work page doi:10.1006/jcis.1996.4675 1997

  25. [25]

    Geuzaine, J

    C. Geuzaine, J. -F. Remacle, Gmsh: A 3 -D finite element mesh generator with built -in pre- and post- processing facilities, Int. J. Numer. Methods Eng. 79(11) (2009) 1309 -1331. https://doi.org/10.1002/nme.2579 24

    work page doi:10.1002/nme.2579 2009

  26. [26]

    Melro, P .P

    A.R. Melro, P .P . Camanho, S.T. Pinho, Generation of random distribution of fibres in long -fibre reinforced composites, Compos. Sci. Technol. 68(9) (2008) 2092 -2102. https://doi.org/10.1016/j.compscitech.2008.03.013

    work page doi:10.1016/j.compscitech.2008.03.013 2008

  27. [27]

    Ghayoor, S.V

    H. Ghayoor, S.V. Hoa, C.C. Marsden, A micromechanical study of stress concentrations in composites, Compos. Part B Eng. 132 (2018) 115-124. https://doi.org/10.1016/j.compositesb.2017.09.009

    work page doi:10.1016/j.compositesb.2017.09.009 2018

  28. [28]

    Trias, J

    D. Trias, J. Costa, A. Turon, J.E. Hurtado, Determination of the critical size of a statistical representative volume element (SRVE) for carbon reinforced polymers, Acta Mater. 54(13) (2006) 3471 -3484. https://doi.org/10.1016/j.actamat.2006.03.042

    work page doi:10.1016/j.actamat.2006.03.042 2006

  29. [29]

    Burachas et al.Lead tungstate PbWO4 crystals for high energy physics,Journal of Crystal Growth198- 199(1999) 881-884, https://doi.org/10.1016/S0022- 0248(98)01237-8

    J. Zeman, M. Šejnoha, Numerical evaluation of effective elastic properties of graphite fiber tow impregnated by polymer matrix, J. Mech. Phys. Solids 49(1) (2001) 69-90. https://doi.org/10.1016/S0022- 5096(00)00027-2

    work page doi:10.1016/s0022- 2001

  30. [30]

    Everett, J.H

    R.K. Everett, J.H. Chu, Modeling of Non-Uniform Composite Microstructures, J. Compos. Mater. 27(11) (1993) 1128-1144. https://doi.org/10.1177/002199839302701105

    work page doi:10.1177/002199839302701105 1993

  31. [31]

    Suquet P , Elements of Homogenization Theory for Inelastic Solid Mechanics, in: E

    M. Suquet P , Elements of Homogenization Theory for Inelastic Solid Mechanics, in: E. Sanchez - Palencia, A. Zaoui (Eds.), Homogenization Techniques for Composite Media, Springer -Verlag, Berlin, Germany, 1987, pp. 193-287

    1987

  32. [32]

    Li, On the nature of periodic traction boundary conditions in micromechanical FE analyses of unit cells, IMA J

    S. Li, On the nature of periodic traction boundary conditions in micromechanical FE analyses of unit cells, IMA J. Appl. Math. 77(4) (2012) 441-450. https://doi.org/10.1093/imamat/hxr024

    work page doi:10.1093/imamat/hxr024 2012

  33. [33]

    Z. Xia, C. Zhou, Q. Yong, X. Wang, On selection of repeated unit cell model and application of unified periodic boundary conditions in micro-mechanical analysis of composites, Int. J. Solids Struct. 43(2) (2006) 266-278. https://doi.org/10.1016/j.ijsolstr.2005.03.055

    work page doi:10.1016/j.ijsolstr.2005.03.055 2006

  34. [34]

    Jiang, K

    M. Jiang, K. Alzebdeh, I. Jasiuk, M. Ostoja-Starzewski, Scale and boundary conditions effects in elastic properties of random composites, Acta Mech. 148(1) (2001) 63-78. https://doi.org/10.1007/BF01183669

    work page doi:10.1007/bf01183669 2001

  35. [35]

    Hazanov, C

    S. Hazanov, C. Huet, Order relationships for boundary conditions effect in heterogeneous bodies smaller than the representative volume, J. Mech. Phys. Solids 42(12) (1994) 1995 -2011. https://doi.org/10.1016/0022-5096(94)90022-1

    work page doi:10.1016/0022-5096(94)90022-1 1994

  36. [36]

    Gusev, Controlled accuracy finite element estimates for the effective stiffness of composites with spherical inclusions, Int

    A.A. Gusev, Controlled accuracy finite element estimates for the effective stiffness of composites with spherical inclusions, Int. J. Solids Struct. 80 (2016) 227-236. https://doi.org/10.1016/j.ijsolstr.2015.11.006

    work page doi:10.1016/j.ijsolstr.2015.11.006 2016

  37. [37]

    Mechin, A

    P .-Y . Mechin, A. Borras, K. Cottard, V. Keryvin, A unified method to generate representative volume elements with tailored random fibre arrangements to estimate the shear and transverse behaviours of unidirectional continuous fibres composite plies, J . Compos. Mater. (2024) 00219983241300144. https://doi.org/10.1177/00219983241300144

    work page doi:10.1177/00219983241300144 2024

  38. [38]

    Sharma, S

    A. Sharma, S. Daggumati, Computational micromechanical modeling of transverse tensile damage behavior in unidirectional glass fiber -reinforced plastic composite plies: Ductile versus brittle fracture mechanics approach, Int. J. Damage Mech. 29(6) (201 9) 943 -964. https://doi.org/10.1177/1056789519894379

    work page doi:10.1177/1056789519894379