Pith. sign in

REVIEW 3 major objections 5 minor 47 references

Slow dynamic nonlinearity in unconsolidated glass bead packs

T0 review · 3 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

Pith's one-line read Unconsolidated glass bead packs exhibit slow dynamic nonlinearity after impulsive, harmonic, and quasi-static conditioning, with quasi-static pumping demonstrated for the first time.

desk verdict Careful bead-pack study convincingly shows slow dynamics after three pump types, including a new quasi-static protocol; the quantitative strain-dependence claims are shakier, and the stress-test's arithmetic inconsistency does not hold up. read the letter →

arxiv 1908.08935 v4 pith:EVDFZU2B submitted 2019-08-23 cond-mat.soft cond-mat.otherphysics.class-phphysics.geo-ph

classification cond-mat.softcond-mat.otherphysics.class-phphysics.geo-ph
keywords slowdynamicsglassbeadpackscodawaveinterferometrynonlinearelasticitylogarithmicrecoverystrainthresholdgranularmedia
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

This paper argues that a simple laboratory material—loose glass beads packed under a dead-weight load—can serve as a clean test bed for slow dynamics, the poorly understood loss of stiffness followed by logarithmic recovery seen in rocks and concrete. Using ultrasound and coda wave interferometry to detect tiny wave-speed changes, the authors show that the bead pack recovers logarithmically after three kinds of low-frequency conditioning: a dropped ball, a 60 Hz shaker, and the quasi-static addition or removal of a 1 kg mass. The quasi-static conditioning result, they report, has not appeared in the literature before. The paper also finds that the recovery rate scales linearly with the root-mean-square pump strain above about one microstrain and levels off below it, suggesting that any threshold strain for slow dynamics in this system lies below $6\times10^{-7}$, if a sharp threshold exists at all. If these claims hold, bead packs give theorists a simpler, better-characterized microstructure in which to test mechanisms of slow dynamics.

What carries the argument

The load-bearing measurement is coda wave interferometry (CWI) applied to multiply scattered ultrasonic waves. A broadband pulse is sent through the bead pack every 0.01 s, 100 received signals are averaged every few seconds, and each averaged signal is cross-correlated with a reference signal in short signal-time windows. The time delay at maximum correlation is plotted against signal-time and fit to a line $\mathcal{T}_{ni} = \sigma_n t_i$; the slope, called stretch, equals the relative change in wave speed and therefore the relative change in modulus. Because the ultrasonic coda has explored many bead contacts, tiny stiffness changes show up as a measurable stretch. The three conditioning methods supply the perturbation, and the stretch recovery curves provide the slow-dynamics signature.

What would settle it

Record the pump onset with an independent trigger and refit the recoveries; if the slopes change beyond stated precision, the log-time fits are artifacts of the chosen $T_0$. Mount a strain gauge in or on the bead pack to measure conditioning strain directly rather than through Eqs. (6) and (7); if the leveling-off strain is not near $10^{-6}$, the threshold claim fails.

Watch

Extended reading notes

Core claim

The central claim is that unconsolidated glass bead packs exhibit reproducible slow dynamic nonlinearity when probed by diffuse ultrasound, with a stiffness drop during low-frequency mechanical conditioning and a recovery that is linear in the logarithm of time. After impulsive, harmonic, and quasi-static pumping, the measured coda-wave stretch $\sigma$ follows $\sigma = m \ln(T - T_0) + \text{const.}$, with slopes of order $10^{-6}$ per log cycle. The quasi-static pump—adding or removing a 1 kg weight—produces slow logarithmic relaxation after both loading and unloading, showing that the system relaxes toward equilibrium regardless of the sign of the imposed strain. In addition, the recovery slope $m$ is linear in the rms pump strain for strains at and above roughly $10^{-6}$, and it flattens at lower strains, which the paper reads as evidence that the threshold for slow dynamics is much lower than previously estimated for bead packs, or absent.

Load-bearing premise

The paper's quantitative conclusions would collapse if direct strain measurement showed that the conditioning strains estimated from the simplified force-to-strain relations are systematically wrong, since the linear-above-one-microstrain and no-sharp-threshold claims are built on those estimates.

Editorial extensions

If this is right

  • A simpler material is now available for testing theories of slow dynamics, with controllable grain size, surface treatment, humidity, and saturation.
  • Quasi-static conditioning extends slow dynamics to a pump mode relevant to tectonic loading, since adding and removing a load both produce logarithmic recovery.
  • The linear dependence of recovery slope on rms strain above about one microstrain and the leveling off below it indicate that the threshold strain in bead packs, if any, is below $6\times10^{-7}$.
  • Ultrasonic probing via CWI can detect slow dynamics even when resonant frequencies do not shift, suggesting that reported threshold strains in sandstones should be re-examined at lower strains.
  • The paper notes that extrapolated full-recovery times differ enormously by pump type—about 50 hours for the dropped ball versus effectively infinite for the quasi-static case—and leaves this discrepancy as an open puzzle.

Reading between the lines

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

  • One testable extension is to vary humidity or apply hydrophobic bead coatings and watch whether the log-time recovery rate shifts; that would discriminate moisture-assisted contact healing from purely mechanical creep.
  • A longer-duration, temperature-controlled quasi-static recovery measurement could test whether the astronomically long extrapolated recovery reflects an unusually flat activation-energy distribution or a genuinely different relaxation process.
  • If the absence of a sharp strain threshold survives direct strain measurement, then small ambient vibrations may continuously excite slow dynamics in granular media, meaning even quiescent measurements of modulus in such materials are already affected by the measurement process.
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

3 major / 5 minor

Summary. The manuscript presents an experimental study of slow dynamic nonlinearity in unconsolidated glass bead packs, using ultrasonic pulses and coda wave interferometry (CWI) to measure changes in wave speed (stretch). The authors apply three low-frequency conditioning methods: impulsive (dropped ball), harmonic (60 Hz shaker), and quasi-static (adding/subtracting a 1 kg mass). They report a drop in stretch followed by logarithmic recovery after all three pump types, with the quasi-static method identified as novel. They also report a preliminary dependence of the recovery slope on the estimated rms pump strain, finding an apparent linear regime above about one microstrain and a leveling off below it, and use this to suggest that any threshold strain for slow dynamics in this system is below 6e-7 if it exists at all. Control experiments with brass side walls and without rubber shims indicate that the observed effects originate in the bead pack rather than in the support apparatus.

Significance. If the central qualitative claim holds, the paper provides a clean, low-noise venue for studying slow dynamics, with the advantage of a simpler and better-characterized microstructure than rocks or cement. The demonstration of slow dynamics after quasi-static conditioning is a new experimental result. The CWI-based stretch measurements achieve precision better than about 1e-6, and the control experiments for the foam and rubber shims strengthen the attribution of the effect to the bead pack. These are genuine strengths and make the system potentially useful for future parametric studies. The quantitative strain-threshold conclusions in Sec. V are more fragile: they depend on estimates of the conditioning strain and on slopes fitted without reported uncertainties, so they should be treated as preliminary unless the calibration is independently validated.

major comments (3)
  1. [Sec. IV, Eq. (6), Eq. (7); Table 1; Fig. 8] The quantitative claims in Sec. V—especially the linear-above-one-microstrain behavior and the 'below 6e-7' threshold statement—rest on the absolute values of the estimated pump strain. The strain values in Table 1 are actually consistent with the coefficient in Eq. (6): using epsilon/F = 2.19e-7 N^-1 gives 2.8e-5 for F = 128 N, 6.15e-7 for F = 2.8 N, and 2.15e-6 for F = 9.81 N, so there is no internal factor-of-ten inconsistency. However, Eq. (6) assumes a uniform uniaxial strain over the whole pack with a single wave speed c = 825 m/s, and Eq. (7) models the 87 kg load plus pack as a single-degree-of-freedom spring-mass system with a resonance at 20 Hz. These are reasonable order-of-magnitude estimates, but they are not independently validated. Because the x-axis of Fig. 8b and the numerical value of the putative threshold depend directly on this calibration, the paper should either provide an independent check (e.g., a direct displacement measurement of the load or pack boundary, or a comparison with static deflection under the added mass) or explicitly frame the strain axis and the threshold as order-of-magnitude estimates with a stated uncertainty. As written, the threshold conclusion in Sec. V is stronger than the calibration supports.
  2. [Sec. IV.A, footnote 5; Table 1; Fig. 8b] The zero time T0 for each recovery is adjusted post hoc to make the log-time plots linear, as acknowledged in footnote 5. The recovery slopes used in Table 1 and Fig. 8b are obtained by fitting from 15 s to 3 min, but no error bars or sensitivity analysis are reported. The slope values are the central quantitative output of the paper, and the apparent linear-versus-plateau structure in Fig. 8b is the basis for the threshold discussion. The authors should report at least the variation of the fitted slopes as T0 is varied within the plausible three-second interval, and ideally provide confidence intervals from the fits. Without this, it is difficult to assess whether the apparent leveling off at low strain in Fig. 8b is significant or within the fitting uncertainty.
  3. [Sec. V, Fig. 8] The low-strain plateau in Fig. 8b is inferred from very few data points, and footnote 7 states that measurements below the lowest strain were contaminated by drift. If the plateau is defined by only one or two points, the claim that the slope 'levels off' below about one microstrain should be worded more cautiously. Additional data points at intermediate strains, or a clear statement of how many independent measurements define each point, would make the threshold discussion more convincing. As it stands, the distinction between a genuine plateau and a scatter-limited flattening is not established.
minor comments (5)
  1. [Abstract] The phrase 'a creep-like phenomena' should be 'a creep-like phenomenon' (singular) or 'creep-like phenomena' without the indefinite article.
  2. [Sec. II, Eq. (2) and surrounding text] Some of the superscript and subscript characters are garbled in the transcript (e.g., the density unit 'g/cm1' should be 'g/cm^3', and several strain exponents such as '10#$' appear as raw symbols). Please ensure the final typeset version uses correct superscripts; this is a readability issue that should not affect the physics.
  3. [Table 1] The table lists slopes for the harmonic conditioning with 'shaker off' and 'shaker on'. It would help to state explicitly in the caption whether the 'shaker on' slope refers to the conditioning phase or to the recovery after the shaker is turned off, and to define the sign convention used for both.
  4. [Fig. 8] Panel (b) would benefit from error bars or at least a statement of the estimated uncertainty in each slope value. As currently plotted, the visual impression of a linear regime and a plateau may overstate the precision of the measurements.
  5. [References] Reference [24] appears to be a thesis citation without a title or institution; if it is publicly available, adding the full reference would improve completeness.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the recovery slopes and Hertzian theory-data comparisons are measured outputs and independent checks, not reproductions of the paper's inputs.

full rationale

Walking the paper's derivation chain: the central observations (log-time recovery after impulsive, harmonic, and quasi-static conditioning) are directly measured outputs, namely CWI stretch versus laboratory time. The recovery slopes m are fit to those measured data and then used to extrapolate full-recovery times; that is an extrapolation of a fit, not a prediction forced by the inputs. The Hertzian comparisons are genuine theory-vs-data checks: Eq. (1)-(2) predict a 249 kHz cutoff from bead geometry and material constants, compared with the observed ~200 kHz spectrum, and the quasi-static stretch prediction of 0.011/6 = 1.8e-3 from the Hertzian speed-force scaling is compared with the measured stretch and explicitly found to disagree by a factor of 6. The cited coda-wave-interferometry technique [33] and impulsive-pump precedent [11] are self-authored but published, externally validated methods; they are not invoked to forbid alternatives or to supply an unverified uniqueness theorem, and no load-bearing claim reduces to those citations. The one notable concern is the internal consistency of the strain calibration in Eq. (6) relative to Table 1 and Fig. 8, which is an experimental-calibration/correctness issue rather than a circularity: the strains are inputs to the x-axis of Fig. 8b, not outputs of the recovery-slope fits. No step in the paper's derivation chain is equivalent by construction to its own inputs, and no prediction is a renamed fit. Therefore the circularity score is 0.

Assumptions & free parameters 1 free parameters · 5 assumptions · 0 invented entities

The central observations do not require any invented entities. The only hand-tuned quantity is T0. The strain scale used to interpret the threshold behavior rests on simplified mechanical models.

free parameters (1)
  • T0 (zero time for recovery) = not stated; chosen per fit to maximize linearity
    Adjusted so log plots are fully linear (footnote 5); affects fitted slopes m and extrapolated recovery times.
assumptions (5)
  • domain assumption Stretch sigma from coda wave interferometry equals relative wave-speed change, hence modulus change.
    Standard result cited to refs [33,46]; central to interpreting all figures.
  • domain assumption The ultrasonic probe does not itself condition the bead pack.
    Verified roughly by quadrupling pulse amplitude (Sec. IV, footnote 4); assumed for the pump-probe interpretation.
  • domain assumption Observed slow dynamics originates in the bead pack, not foam walls or rubber shims.
    Supported by replacement tests (brass cylinder, shim removal) in Sec. II.
  • domain assumption Equation (6) converts applied force to conditioning strain via pack mass, thickness, and wave speed.
    Used for all strain values in Table 1 and Fig. 8b; ignores spatial variation and dynamics.
  • standard math Diffusion fit parameters (D, alpha) and the hcp cutoff formula are standard derived results.
    Used in Sec. II for characterization; do not affect slow dynamic claims.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Slow dynamic nonlinearity in unconsolidated glass bead packs." pith.science (2026). https://pith.science/paper/EVDFZU2B

@misc{pith2026190808935,
  author       = {Pith},
  title        = {Pith review of: Slow dynamic nonlinearity in unconsolidated glass bead packs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EVDFZU2B}},
  note         = {Machine review of arXiv:1908.08935}
}
read the original abstract

Slow dynamic nonlinearity describes a poorly understood, creep-like phenomena that occurs in brittle composite materials such as rocks and cement. It is characterized by a drop in stiffness induced by a mechanical conditioning, followed by a log(time) recovery. A consensus theoretical understanding of the behavior has not been developed. Here we introduce an alternative experimental venue with which to inform theory. Unconsolidated glass bead packs are studied rather than rocks or cement because the structure and internal contacts of bead packs are less complex and better understood. Slow dynamics has been observed in such systems previously. However, the measurements to date tend to be irregular. Particular care is used here in the experimental design to overcome the difficulties inherent in bead pack studies. This includes the design of the bead pack support, the use of low frequency conditioning, and the use of ultrasonic waves as a probe with coda wave interferometry to assess changes. Slow dynamics is observed in our system after three different methods for low-frequency conditioning, one of which has not been reported in the literature previously.

Figures

Figures reproduced from arXiv: 1908.08935 by the authors.

Figure 1
Figure 1. A schematic, panel (a), and photo, panel (b), of the experimental setup. Panel (a) shows that the bead pack is sandwiched by 1.6mm steel plates, which, in turn, are sandwiched by thick-walled hollow cylinders of high strength foam. The same foam is used to confine the beads laterally. A broadband pulse is sent to the top steel plate where it spreads laterally and then propagates into the bead pack. A transducer at t… view at source ↗
Figure 2
Figure 2. Typical ultrasonic signal (2a) and its spectrum (blue curve in 2b) thro [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. A plot of energy (smoothed signal squared) versus signal-time (solid blue curve). A fit to a diffusion equation is plotted as well (dashed red curve): 𝐸 = K √M e#OM#PQ/RSM. A diffusivity of 𝐷 = 0.36𝑚=/𝑠 and an absorption of α = 4200 𝑠#A were used to create the curve [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Summary of the coda wave interferometry procedure. Panel (a) shows the reference signal, 𝜙 = 𝜓A, (blue curve) and the 202nd signal (red curve) in a measurement. The plot is shaded to signify the 13 windows (𝑊 = 200 𝜇𝑠) used to construct 𝑋=Ñ= qÖA,=,…A1. Panel (b) shows …
Figure 5
Figure 5. Figure 5: The slow dynamics results for impulsive pumping. Stretch i [PITH_FULL_IMAGE:figures/full_fig_p012_5.png]
Figure 6
Figure 6. Figure 6: The slow dynamic results for harmonic pumping. Stretch is plotted versus laboratory [PITH_FULL_IMAGE:figures/full_fig_p014_6.png]
Figure 7
Figure 7. Figure 7: The slow dynamic results for quasi-static pumping. Stretch is plotted versus the laboratory-time in panel (a). The shaded regions indicate the times in which the 1kg mass was placed on top of the large static load. Both conditioning (1kg added) and recovery (1kg subtra…
Figure 8
Figure 8. Figure 8: Slow dynamics recovery for different pump strain levels. Panel (a) shows stretch versus log(time) for [PITH_FULL_IMAGE:figures/full_fig_p017_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

47 extracted references · 47 canonical work pages

  1. [1]

    J. A. TenCate and T. J. Shankland, Geophys. Res. Lett. 23, 3019 (1996)

  2. [2]

    J. A. TenCate, E. Smith, and R. A. Guyer, Phys. Rev. Lett. 85, 1020 (2000)

  3. [3]

    J. A. TenCate, Pure Appl. Geophys. 168, 2211 (2011)

  4. [4]

    Johnson and A

    P. Johnson and A. Sutin, J. Acoust. Soc. Am. 117, 124 (2005)

  5. [5]

    R. A. Guyer and P. A. Johnson, Phys. Today 52, 30 (1999)

  6. [6]

    R. A. Guyer and P. A. Johnson, Nonlinear Mesoscopic Elasticity: The Complex Behaviour of Granular Media Including Rocks and Soil (2009)

  7. [7]

    Shokouhi, J

    P. Shokouhi, J. Rivière, R. A. Guyer, and P. A. Johnson, Appl. Phys. Lett. 111, (2017)

  8. [8]

    J. A. TenCate, J. Duran, and T. J. Shankland, in Nonlinear Acoust. Begin. 21st Century, edited by O. V. Rudenko and O. A. Sapozhnikov (Moscow State University, 2002), pp. 767–770

Show all 47 references
  1. [9]

    Brenguier, M

    F. Brenguier, M. Campillo, C. Hadziioannou, N. M. Shapiro, R. M. Nadeau, and E. Larose, Science (80-. ). 321, 1478 (2008)

  2. [10]

    Gassenmeier, C

    M. Gassenmeier, C. Sens-Schönfelder, T. Eulenfeld, M. Bartsch, P. Victor, F. Tilmann, and M. Korn, Geophys. J. Int. 204, 1490 (2016)

  3. [11]

    O. I. Lobkis and R. L. Weaver, J. Acoust. Soc. Am. 125, 1894 (2009)

  4. [12]

    Tremblay, E

    N. Tremblay, E. Larose, and V. Rossetto, J. Acoust. Soc. Am. 127, 1239 (2010)

  5. [13]

    Renaud, S

    G. Renaud, S. Calĺ, and M. Defontaine, Appl. Phys. Lett. 94, (2009)

  6. [14]

    Renaud, P

    G. Renaud, P. Y. Le Bas, and P. A. Johnson, J. Geophys. Res. Solid Earth 117, 1 (2012)

  7. [15]

    Rivière, G

    J. Rivière, G. Renaud, R. A. Guyer, and P. A. Johnson, J. Appl. Phys. 114, (2013)

  8. [16]

    Renaud, J

    G. Renaud, J. Rivière, P. Y. Le Bas, and P. A. Johnson, Geophys. Res. Lett. 40, 715 (2013)

  9. [17]

    Bureau, T

    L. Bureau, T. Baumberger, and C. Caroli, Eur. Phys. J. E 8, 331 (2002)

  10. [18]

    J. A. TenCate, E. Smith, L. W. Byers, and T. J. Shankland, AIP Conf. Proc. 524, 303 (2000)

  11. [19]

    X. Li, C. Sens-Schönfelder, and R. Snieder, Phys. Rev. B 97, 1 (2018)

  12. [20]

    Snieder, C

    R. Snieder, C. Sens-Schönfelder, and R. Wu, Geophys. J. Int. 208, 1 (2017)

  13. [21]

    A. V. Lebedev and L. A. Ostrovsky, Acoust. Phys. 60, 555 (2014)

  14. [22]

    Ostrovsky, A

    L. Ostrovsky, A. Lebedev, J. Riviere, P. Shokouhi, C. Wu, M. A. Stuber Geesey, and P. A. Johnson, J. Geophys. Res. Solid Earth 5003 (2019)

  15. [23]

    K. E.-A. Van Den Abeele, J. Geophys. Res. 107, 1 (2002)

  16. [26]

    J. A. Bittner and J. S. Popovics, Appl. Phys. Lett. 114, 021901 (2019)

  17. [27]

    Bocquet, E

    L. Bocquet, E. Charlaix, S. Ciliberto, and J. Crassous, Nature 396, 735 (1998)

  18. [28]

    O. O. Vakhnenko, V. O. Vakhnenko, T. J. Shankland, and J. A. TenCate, AIP Conf. Proc. 838, 120 (2006)

  19. [29]

    V. Y. Zaitsev, V. E. Gusev, V. Tournat, and P. Richard, Phys. Rev. Lett. 112, 1 (2014)

  20. [30]

    Zaitsev, V

    V. Zaitsev, V. Gusev, and B. Castagnede, Phys. Rev. Lett. 90, 4 (2003)

  21. [31]

    P. A. Johnson and X. Jia, Nature 437, 871 (2005)

  22. [32]

    Tournat and V

    V. Tournat and V. E. Gusev, Phys. Rev. E - Stat. Nonlinear, Soft Matter Phys. 80, 1 20 (2009)

  23. [33]

    O. I. Lobkis and R. L. Weaver, Phys. Rev. Lett. 90, 4 (2003)

  24. [34]

    V. Y. Zaitsev, V. E. Nazarov, V. Tournat, V. E. Gusev, and B. Castagnède, Europhys. Lett. 70, 607 (2005)

  25. [35]

    X. Jia, T. Brunet, and J. Laurent, Phys. Rev. E - Stat. Nonlinear, Soft Matter Phys. 84, 2 (2011)

  26. [36]

    Tournat and V

    V. Tournat and V. E. Gusev, Acta Acust. United with Acust. 96, 208 (2010)

  27. [37]

    R. L. Weaver and W. Sachse, J. Acoust. Soc. Am. 97, 2094 (1995)

  28. [38]

    X. Jia, C. Caroli, and B. Velicky, Phys. Rev. Lett. 82, 1863 (1999)

  29. [39]

    Jia, Phys

    X. Jia, Phys. Rev. Lett. 93, 8 (2004)

  30. [40]

    Snieder, Pure Appl

    R. Snieder, Pure Appl. Geophys. 163, 455 (2006)

  31. [41]

    Planès and E

    T. Planès and E. Larose, Cem. Concr. Res. 53, 248 (2013)

  32. [42]

    D. M. Mueth, H. M. Jaeger, and S. R. Nagel, Phys. Rev. E - Stat. Physics, Plasmas, Fluids, Relat. Interdiscip. Top. (1998)

  33. [43]

    J. A. Turner, M. E. Chambers, and R. L. Weaver, Acustica 84, 628 (1998)

  34. [44]

    Merkel, V

    A. Merkel, V. Tournat, and V. Gusev, Phys. Rev. E - Stat. Nonlinear, Soft Matter Phys. 82, 1 (2010)

  35. [45]

    Merkel, V

    A. Merkel, V. Tournat, and V. Gusev, Phys. Rev. Lett. 107, 1 (2011)

  36. [46]

    Snieder, A

    R. Snieder, A. Grêt, H. Douma, and J. Scales, Science (80-. ). 295, 2253 (2002)

  37. [47]

    J. A. TenCate, D. Pasqualini, S. Habib, K. Heitmann, D. Higdon, and P. A. Johnson, Phys. Rev. Lett. 93, 65501 (2004)

  38. [48]

    Pasqualini, K

    D. Pasqualini, K. Heitmann, J. A. TenCate, S. Habib, D. Higdon, and P. A. Johnson, J. Geophys. Res. Solid Earth 112, B01204 (2007)

  39. [49]

    K. L. Johnson, Contact Mechanics (Cambridge University Press, Cambridge, UK, 1985)

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.