Non-classical scaling of strength with size in marine biological fibers

Haneesh Kesari; Sayaka Kochiyama

REVIEW 4 major objections 4 minor 8 references

Non-classical scaling of strength with size in marine biological fibers

T0 review · 4 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read Sponge glass fiber strength scales as the inverse square of fiber radius.

desk verdict New spicule strength data with a strong size effect, but the b^-2 scaling is asserted, not fitted. read the letter →

arxiv 2411.10672 v1 pith:VSPCAEUV submitted 2024-11-16 physics.bio-ph

Sayaka Kochiyama , Haneesh Kesari This is my paper · ORCID

classification physics.bio-ph

keywords Euplectellaaspergillumbasaliaspiculesizeeffecttensilestrengthlinearelasticfracturemechanicsflawscalingbiogenicsilicainverse-square

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 reports tensile tests on 38 anchor spicules of the marine sponge Euplectella aspergillum and finds that their strength rises as the inverse square of the fiber radius. That is a much faster size effect than the classical linear elastic fracture mechanics prediction, in which strength grows as the inverse square root of the specimen dimension. The paper argues that the faster scaling is still ordinary fracture mechanics provided the internal flaw that triggers failure shrinks faster than the fiber itself, and it uses a crack-size reconstruction to show that the inferred flaw ratio falls with radius. The smallest specimen reached a tensile strength near 1.5 GPa, which places these biogenic glass fibers among the strongest natural fibers and suggests that deliberately shrinking flaw size relative to fiber size could make engineered glass stronger.

What carries the argument

The load-bearing tool is a published stress-intensity-factor solution for a cylindrical rod of radius $b$ containing an internal axisymmetric crack of radius $a$, combined with a fracture toughness value for the spicule material. The paper substitutes each measured strength and radius into this relation to back out the crack size $a$; the resulting decrease of the ratio $a/b$ with decreasing $b$ is what reconciles the observed $b^{-2}$ strength scaling with classical fracture mechanics. The physical floor on flaw size, tied to the 50–200 nm colloidal silica spheres that make up the spicule's layered structure, explains why smaller specimens can be relatively flaw-free.

What would settle it

Measure the diameter and gauge length of every spicule and check whether the 14 specimens with randomly chosen gauge lengths show a correlation between diameter and length; if thinner spicules are systematically shorter, retest with a fixed gauge length over the full diameter range. A direct check is to test fibers of identical diameter at several gauge lengths: if strength drops as gauge length grows, the inverse-square law is at least partly a volume effect rather than a pure diameter effect.

Watch

Extended reading notes

Core claim

The central claim is that the tensile strength of Euplectella basalia spicules follows $\sigma_s^t \propto b^{-2}$, where $b$ is the specimen radius, rather than the classical $\sigma_s^t \propto b^{-1/2}$ size effect. The paper shows that this non-classical scaling is consistent with linear elastic fracture mechanics if the crack responsible for failure shrinks faster than the specimen: plugging measured strengths and radii into a stress-intensity solution for a cylinder with an internal axisymmetric crack, and using published fracture toughness values, gives an estimated crack-radius-to-specimen-radius ratio $a/b$ that decreases as $b$ decreases. The measured maximum strength, 1.5 GPa for the smallest specimen, is presented as the natural outcome of that scaling, and the inferred crack sizes remain in a physically reasonable range tied to the size of the silica particles that build the spicule.

Load-bearing premise

The result assumes that specimen diameter, rather than gauge length or volume, is the variable driving the strength increase; the paper does not test whether thinner spicules were also shorter, which could produce the same apparent scaling through weakest-link statistics.

Editorial extensions

If this is right

If the inverse-square scaling holds, halving a spicule's radius should roughly quadruple its tensile strength, predicting much stronger fibers at small diameters than the classical square-root law would allow.
Because the data are consistent with linear elastic fracture mechanics, the non-classical scaling does not require a new failure mechanism, only that flaw size diminishes faster than specimen size.
The 1.5 GPa peak strength places these spicules among the strongest natural fibers, comparable to spider silk and bamboo fiber, despite the spicule's untreated exposure to the ocean.
The spicule's layered assembly offers a manufacturing lesson: processes that keep internal flaws small relative to fiber diameter could push engineered glass fibers toward higher strength.

Reading between the lines

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

If the flaw-size floor is set by gaps between the 50–200 nm colloidal silica particles, then below some small radius the ratio $a/b$ should stop decreasing and the strength scaling should revert toward $b^{-1/2}$; tensile tests on sub-10-$\mu$m spicules or synthetic analogs would reveal that limit.
The paper uses a single fracture toughness value for all specimens; measuring toughness on the tested fibers themselves would show whether the inferred $a/b$ trend is robust or partly an artifact of that choice.
The subset with random gauge lengths could be analyzed as a weakest-link experiment: if strength at fixed diameter falls with increasing gauge length, then diameter scaling and volume scaling are entangled, and the reported $b^{-2}$ law would need to be replaced by a coupled size-length law.

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.

Desk Editor's Note

read the letter

The paper reports tensile tests on 38 Euplectella basalia spicules and finds that strength increases sharply with decreasing diameter, reaching 1.5 GPa for the thinnest specimen. That dataset is a genuine contribution, and the paper is honest about the scatter. The authors also usefully note that Griffith's own fiber data already show a D^-0.73 scaling, which puts their claim in context. Experimentally, the work seems careful: they checked for tapering, used slip marks, and clearly describe the loading setup.

The problem is the central claim. The b^-2 scaling is imposed, not estimated. They fix the exponent to -2 and draw a curve through scattered points, with no free-exponent fit, no confidence intervals, and no comparison against weaker scalings like b^-1 or b^-3/2. You cannot tell from what is presented whether the data are actually inconsistent with a shallower exponent. The classical b^-1/2 line is fitted to a post-hoc subset of 35 specimens with b > 25 μm, which excludes the small specimens that most strongly drive the steep slope. That makes the comparison look worse than a fair fit would.

There is also the gauge-length confound. Fourteen of 38 specimens had randomly chosen gauge lengths between 1.1 and 5.4 cm. If thinner spicules were also shorter, the apparent size effect could be a volume effect from weakest-link statistics. The authors do not test this, and the fixed-gauge subset (N=24) is not analyzed separately for scaling. That is a real gap.

The crack-size interpretation is circular. They infer flaw size a from the same strength measurements and an assumed toughness KIC, then note that a/b decreases with b. That trend is mathematically tied to the strength scaling, so it is not independent evidence. The KIC value is borrowed from one prior study, and the other two values are simply assumed; there is no direct toughness measurement.

On the positive side, the paper is clearly written and the authors engage with the relevant literature, though the Benthem 1972 solution is mentioned without a reference in the list. Raw data are not provided, which is a problem for a scaling claim.

Who is this for? Researchers interested in biological fibers and size effects in strength. The dataset is worth having, but the analysis needs substantial revision before the b^-2 scaling is credible. I would send this to peer review because the experimental work is solid and the question is interesting, but I would require the authors to fit a free exponent with uncertainty, report the fixed-gauge subset separately or include gauge length as a covariate, provide raw data, and properly cite Benthem. If the free-exponent fit still shows an exponent clearly steeper than -1/2, the paper becomes compelling.

Referee Report

4 major / 4 minor

Summary. The paper reports tensile tests on 38 basalia spicules from the marine sponge Euplectella aspergillum and claims that tensile strength scales as the inverse square of the spicule radius, b^-2, which is a much stronger size effect than the classical Griffith square-root scaling. The authors compare their data with a b^-1/2 curve fitted to a subset of 35 larger specimens, use Benthem's stress-intensity solution to estimate crack sizes from the measured strengths for three assumed fracture-toughness values, and argue that the estimated crack sizes are consistent with the observed scaling. They also place their measured strengths in the context of other natural fibers and technical glasses.

Significance. If the claimed b^-2 scaling were established, it would be a striking deviation from classical fracture-mechanics size effects and could motivate new bioinspired design strategies. The paper has genuine strengths: the experimental campaign is non-trivial for biological fibers, the diameter measurements include a quantified taper check (reported end-to-end diameter ratio 1.00 +/- 0.039), and the authors contextualize their results against Griffith's original fiber data and modern technical glasses. However, the central claim is currently supported only by a fixed-exponent fit imposed on highly scattered data, with no statistical evidence that the exponent is distinguishable from weaker scalings, and with a possible gauge-length confound. The crack-size 'consistency' argument is partly circular because the crack sizes are computed from the same strength measurements used to infer the scaling. Thus the significance is conditional on a reanalysis that the current manuscript does not provide.

major comments (4)

[Section 6, Fig. 4] The central claim that strength scales as b^-2 is asserted, not estimated. The solid curve in Fig. 4 is y = alpha x^-2 with alpha = 557, but the exponent -2 is fixed a priori; no free-exponent fit (e.g., sigma = alpha b^m with m free) is reported, and no confidence intervals, standard errors, or goodness-of-fit measures are given. Given the reported scatter (mean 0.55 +/- 0.28 GPa over N = 38), the data may well be consistent with m = -1 or m = -3/2, which would substantially alter the paper's novelty claim. The authors should fit the exponent freely and report confidence intervals or a model-comparison statistic.
[Sections 3 and 5] The gauge-length confound is not addressed. Section 3 states that 14 of 38 specimens had randomly chosen gauge lengths between 1.1 and 5.4 cm, while only 24 specimens had a fixed 1.5 cm gauge. Section 5 reports only aggregate means for the fixed-gauge subset (0.51 +/- 0.32 GPa, N = 24) and does not present a scaling fit for that subset. If gauge length correlates with radius, weakest-link volume effects could produce a spurious diameter scaling. A regression with gauge length as a covariate, or a fit restricted to the 24 fixed-gauge specimens, is necessary to exclude this confound.
[Section 6, Fig. 4] The classical b^-1/2 comparison is fitted only to a post-hoc subset of 35 specimens with radius b > 25 um. This exclusion removes the smallest specimens, which are precisely the points that most strongly drive the inverse-square trend. The dashed curve therefore does not represent a fair comparison between the two scaling hypotheses. The authors should either fit both curves to the full dataset or provide a clear rationale for excluding the three smallest specimens.
[Section 5, Section 6, Fig. 5] The claimed consistency of the b^-2 scaling with linear elastic fracture mechanics is circular in an important respect. Crack sizes a are computed from the measured strengths and an assumed KIC using Eq. (4.1); when sigma_t is assumed to scale as b^-2 and KIC is fixed, the small-crack limit of Eq. (4.1) forces a ~ (KIC / sigma)^2 ~ b^4, so that a/b ~ b^3 decreases with decreasing b. The trend in Fig. 5(B)-(D) is therefore a mathematical consequence of the assumed strength scaling rather than an independent confirmation. The authors should either show a direct comparison between predicted and measured strengths for independently known flaw sizes, or clearly state that the crack-size estimates are derived from the scaling being tested.

minor comments (4)

[Section 2.1] There are duplicated words in the text: 'resolution to this this apparent contradiction' and 'the the fracture stress' should be corrected.
[Section 7] The phrase 'much moreEa. spicules' appears to be a typographical error; it should read 'much more in Ea. spicules' or similar.
[Figure 1] The caption uses 'Ea.' without defining it, although the definition appears in Section 1; captions should be self-contained for readers who browse figures.
[Section 5] The Results section reports only mean and standard deviation values; providing a supplementary table with individual diameter, gauge length, and strength measurements would substantially aid the reader's ability to assess the scaling claims.

Circularity Check

1 steps flagged · score 6.0 of 10

The b^-2 scaling is an empirical fit; the claimed 'consistency' with fracture mechanics is a rearrangement of the same strength measurements through Eq. (4.1), not an independent prediction.

fitted input called prediction [Section 6, Fig. 5, and Eq. (4.1)]
"While this rate of increase in the σs t value is much faster than what is classically expected, it is consistent with the crack length estimate shown in Fig. 5, computed according to Eq. (4.1). In the figures, it can be noted that the ratio of crack diameter to specimen diameter, a/b, decreases with decreasing radius b; this is as opposed to it being a constant value over varying specimen dimension, which would be expected in the case σs t scaled as b−1/2."

Equation (4.1) defines KIC in terms of σs_t, a, and b. For fixed KIC, substituting the measured σs_t and b yields a, so a/b(b) is a deterministic rearrangement of the very same strength data used to establish the b^-2 trend. With σs_t ∝ b^-2, a/b must decrease with decreasing b; with σs_t ∝ b^-1/2 it is constant. Thus the observed a/b trend is a mathematical consequence of the fitted law, not an independent check. The paper's statement 'it is consistent with the crack length estimate' therefore presents a transform of the input data as corroboration, so the fracture-mechanics consistency reduces by construction.

full rationale

The paper's central observation is an empirical scaling: tensile strength decreases roughly as b^-2. That claim is not itself circular; it is a fit to measured data, though the fixed exponent and lack of a free-exponent comparison are statistical weaknesses rather than circularity. The genuinely circular step is the supporting argument in Section 6: the crack sizes a are estimated from the same measured strengths and radii via Eq. (4.1) with an assumed KIC, and the resulting a/b trend is then offered as evidence that the b^-2 law is consistent with fracture mechanics. Because a/b(b) is obtained by substituting the measured σs_t(b) into a fixed formula, the observed trend is a mathematical transform of the input data, not an independent confirmation. No load-bearing self-citation is present: the self-cited work [3] is used only for structural context, and the KIC value is taken from an external source [4]. The gauge-length variation in 14 of 38 specimens is a potential confounding issue but is a correctness risk, not a circularity. Overall, the empirical scaling has independent content, but one of the paper's main 'consistency' claims reduces by construction, giving partial circularity.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central scaling claim itself requires only the strength measurements and a two-parameter fit, but the explanatory mechanism (flaw size decreasing faster than specimen size) rests on several strong assumptions: the Benthem crack geometry, constant KIC, and the absence of confounding variables such as gauge length. The fracture toughness values are taken from a single literature source and varied over a factor of two, which changes the estimated crack size ranges substantially.

free parameters (3)

α (scale factor for b^-2 fit) = 557
Fitted to all 38 strength-radius data points in Figure 4; the value is not used to derive anything independently.
β (scale factor for b^-1/2 fit) = 2.95
Fitted to the 35 specimens with radius > 25 μm for the classical scaling comparison.
Fracture toughness KIC = 0.43, 0.30, 0.60 MPa·m^1/2
Assumed values used to estimate crack sizes; 0.43 is taken from ref [4], the others are chosen as bounds. The crack-size estimates and the a/b trend depend directly on this choice.

assumptions (4)

standard math Griffith's energy balance fracture criterion (Eq 2.1)
Recapitulated in Section 2.1; used as the classical baseline for size scaling.
standard math Benthem's stress intensity factor solution for an internal axisymmetric crack in a cylinder (Eq 4.1)
Used in Section 4 to compute crack sizes from measured strength and assumed KIC; the solution is cited but not derived, and its reference is missing.
domain assumption The spicule can be modeled as a isotropic elastic cylinder with a single dominant internal axisymmetric crack
Adopted in Section 4; the lamellar, biogenic structure may not satisfy this idealization, and failure may originate at surface flaws.
domain assumption KIC is a constant material property equal to the value for silica at the relevant scale
Used in Section 4 and Figure 5; no measurement of KIC for these specific spicules is reported in this paper.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Non-classical scaling of strength with size in marine biological fibers." pith.science (2026). https://pith.science/paper/VSPCAEUV

@misc{pith2026241110672,
  author       = {Pith},
  title        = {Pith review of: Non-classical scaling of strength with size in marine biological fibers},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VSPCAEUV}},
  note         = {Machine review of arXiv:2411.10672}
}

read the original abstract

Intriguing physical phenomena observed in natural materials have inspired the development of several engineering materials with dramatically improved performance. Marine sponge glass fibers, for instance, have attracted interest in recent decades. We tested the glass fibers in tension and observed that the strength of these fibers scales inversely with their size. While it is expected that the strength of a material scales inversely with its size, the scaling is generally believed to be inversely proportional to the square root of the specimen dimension. Interestingly, we found that the marine sponge glass fibers' strength scaled much faster, and was inversely proportional to the square of the specimen dimension. Such non-classical scaling is consistent with the experimental measurements and classical linear elastic fracture mechanics. We hypothesize that this enhanced scaling is due to the flaw size decreasing faster than the size of the specimen. The tensile strength, as a result of non-classical, higher-order scaling, reached a value as large as 1.5 GPa for the smallest diameter specimen. The manufacturing processes through which the spicules are made might hold important lesson for further enhancing the strength of engineering materials.

Figures

Figures reproduced from arXiv: 2411.10672 by the authors.

**Figure 1.** Euplectella aspergillum sponge and its basalia spicules. (A) The entire skeletal structure of an Ea. sponge; the basalia spicules are identified with a white arrow (modified from [3], with permission from National Academy of Sciences). (B) Lateral view of an Ea. basalia spicule taken with an optical microscope. (C) A close-up view revealing the lamellar structure within the spicule. (D) Cross-section of an Ea. basal… view at source ↗

**Figure 2.** (A) Tensile strength of glass specimens in Griffith’s experiments is plotted against 1/ √ ac, where ac is the length of the crack artificially introduced to the specimens and therefore is a known quantity. The measured values lie on a linear line and show that indeed, the measured tensile strength values scale as predicted by Eq. (2.1) The slope of the line of fit (gray curve) is γ1 = 8.28. (B) Tensile strength of t… view at source ↗

**Figure 3.** Schematic of the conducted tensile test on [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Figures from the paper (3 more)

**Figure 4.** Figure 4: Tensile strength σ s t measured for 38 Ea. basalia spicules. The solid curve represents the nonlinear fitting of all 38 data points, where α = 557, and the dashed curve represents the fitting of the 35 specimens whose radius b is greater than 25 µm, where β = 2.95. 6. …

**Figure 5.** Figure 5: (A) Tensile strength measured for 38 Ea. basalia spicules, each sample marked with a unique marker (the presented data is identical to that shown in Fig.4; the plot is presented here again to facilitate comparison with the subfigures (B)–(D)). (B)–(D) Ratio of half cra…

**Figure 6.** Figure 6: (A) Schematic showing one way in which the colloidal silica spheres could be arranged within Ea. basalia spicules. When the spheres are arranged in this manner, the size of the ”gap” (indicated by dashed enclosure) between the spheres is equal to the sphere diameter. (…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

8 extracted references · 8 canonical work pages

[1]

Structural and mechan- ical properties of bamboo fiber bundle and fiber /bundle reinforced composites: a review

Gao, X., Zhu, D., Fan, S., Rahman, M.Z., Guo, S., Chen, F., 2022. Structural and mechan- ical properties of bamboo fiber bundle and fiber /bundle reinforced composites: a review. Journal of Materials Research and Technology 19, 1162–1190

work page 2022
[2]

Gri ffith, A.A., 1921. Vi. the phenomena of rupture and flow in solids. Philosophical transactions of the royal society of london. Series A, containing papers of a mathematical or physical character 221, 163–198

work page 1921
[3]

New functional insights into the internal architecture of the laminated anchor spicules of Euplectella as- pergillum

Monn, M.A., Weaver, J.C., Zhang, T., Aizenberg, J., Kesari, H., 2015. New functional insights into the internal architecture of the laminated anchor spicules of Euplectella as- pergillum. Proceedings of the National Academy of Sciences 112, 4976–4981

work page 2015
[4]

Tensile and fracture behavior of silica fibers from the venus flower basket (eu- plectella aspergillum)

Morankar, S., Singaravelu, A.S.S., Niverty, S., Mistry, Y ., Penick, C.A., Bhate, D., Chawla, N., 2022. Tensile and fracture behavior of silica fibers from the venus flower basket (eu- plectella aspergillum). International Journal of Solids and Structures 253, 111622

work page 2022
[5]

The strength of fused silica

Proctor, B., Whitney, I., Johnson, J., 1967. The strength of fused silica. Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences 297, 534–557

work page 1967
[6]

Changes in the tensile strength of glass caused by water immersion treatment

Stockdale, G., Tooley, F., Ying, C., 1951. Changes in the tensile strength of glass caused by water immersion treatment. Journal of the American Ceramic Society 34, 116–121

work page 1951
[7]

Mechanisms of toughening of a natural rigid com- posite

Walter, S., Flinn, B., Mayer, G., 2007. Mechanisms of toughening of a natural rigid com- posite. Materials Science and Engineering: C 27, 570–574

work page 2007
[8]

Hierarchical assembly of the siliceous skeletal lattice of the hexactinellid sponge euplectella aspergillum

Weaver, J.C., Aizenberg, J., Fantner, G.E., Kisailus, D., Woesz, A., Allen, P., Fields, K., Porter, M.J., Zok, F.W., Hansma, P.K., et al., 2007. Hierarchical assembly of the siliceous skeletal lattice of the hexactinellid sponge euplectella aspergillum. Journal of Structural Biology 158, 93–106. 10

work page 2007

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [1]

Structural and mechan- ical properties of bamboo fiber bundle and fiber /bundle reinforced composites: a review

Gao, X., Zhu, D., Fan, S., Rahman, M.Z., Guo, S., Chen, F., 2022. Structural and mechan- ical properties of bamboo fiber bundle and fiber /bundle reinforced composites: a review. Journal of Materials Research and Technology 19, 1162–1190

work page 2022

[2] [2]

Gri ffith, A.A., 1921. Vi. the phenomena of rupture and flow in solids. Philosophical transactions of the royal society of london. Series A, containing papers of a mathematical or physical character 221, 163–198

work page 1921

[3] [3]

New functional insights into the internal architecture of the laminated anchor spicules of Euplectella as- pergillum

Monn, M.A., Weaver, J.C., Zhang, T., Aizenberg, J., Kesari, H., 2015. New functional insights into the internal architecture of the laminated anchor spicules of Euplectella as- pergillum. Proceedings of the National Academy of Sciences 112, 4976–4981

work page 2015

[4] [4]

Tensile and fracture behavior of silica fibers from the venus flower basket (eu- plectella aspergillum)

Morankar, S., Singaravelu, A.S.S., Niverty, S., Mistry, Y ., Penick, C.A., Bhate, D., Chawla, N., 2022. Tensile and fracture behavior of silica fibers from the venus flower basket (eu- plectella aspergillum). International Journal of Solids and Structures 253, 111622

work page 2022

[5] [5]

The strength of fused silica

Proctor, B., Whitney, I., Johnson, J., 1967. The strength of fused silica. Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences 297, 534–557

work page 1967

[6] [6]

Changes in the tensile strength of glass caused by water immersion treatment

Stockdale, G., Tooley, F., Ying, C., 1951. Changes in the tensile strength of glass caused by water immersion treatment. Journal of the American Ceramic Society 34, 116–121

work page 1951

[7] [7]

Mechanisms of toughening of a natural rigid com- posite

Walter, S., Flinn, B., Mayer, G., 2007. Mechanisms of toughening of a natural rigid com- posite. Materials Science and Engineering: C 27, 570–574

work page 2007

[8] [8]

Hierarchical assembly of the siliceous skeletal lattice of the hexactinellid sponge euplectella aspergillum

Weaver, J.C., Aizenberg, J., Fantner, G.E., Kisailus, D., Woesz, A., Allen, P., Fields, K., Porter, M.J., Zok, F.W., Hansma, P.K., et al., 2007. Hierarchical assembly of the siliceous skeletal lattice of the hexactinellid sponge euplectella aspergillum. Journal of Structural Biology 158, 93–106. 10

work page 2007