Modeling the Fatigue Behavior of Amorphous Polymers

Alessio Zaccone; Oleg V. Gendelman; Valeriy V. Ginzburg

REVIEW 4 major objections 5 minor 7 references

Modeling the Fatigue Behavior of Amorphous Polymers

T0 review · 4 major / 5 minor · reviewed 2026-08-02 · deepseek-v4-flash

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

Pith's one-line read For amorphous polymers under symmetric cyclic loading, the authors derive Basquin's law from a linear viscoelastic model and show the exponent is fixed at m=3, with the prefactor given by measurable material properties.

desk verdict A clean, explicitly derived m=3 Basquin law for amorphous polymers, with a prefactor that is only as good as the fitted tanδ and a thermal-reset assumption the authors themselves call artificial. read the letter →

arxiv 2607.14994 v1 pith:SW4OHEIZ submitted 2026-07-16 cond-mat.mtrl-sci cond-mat.soft

Valeriy V. Ginzburg , Oleg V. Gendelman , Alessio Zaccone This is my paper · ORCID

classification cond-mat.mtrl-scicond-mat.soft

keywords fatigueBasquin'slawamorphouspolymersS-NcurveviscoelasticitylosstangentMaxwellmodelglassy

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper tries to explain the empirical Basquin law of fatigue—N S^m = constant—for amorphous polymers, where N is cycles to failure and S is stress amplitude. Using a linear Maxwell-type viscoelastic model with a single alpha relaxation time, the authors show that viscous dissipation in the first half of each cycle slightly heats the material, speeding relaxation in the second half and leaving a small residual viscoplastic strain each cycle. Summing this damage to a critical failure strain gives m = 3 exactly, and a closed-form prefactor A that depends on loss tangent, modulus, heat capacity, activation energy, and ultimate strength. The predicted S-N curves reproduce experimental data for PS, PMMA, and PVC with only the loss tangent as a fitting parameter. A sympathetic reader would care because this turns fatigue lifetime prediction into a matter of simple one-time material tests rather than long, expensive testing campaigns.

What carries the argument

The machinery is a linear Maxwell element for the viscoplastic strain rate, dε_VP/dt = σ/(E τ), together with an Arrhenius-type shift of the alpha relaxation time with temperature. The dissipation per half-cycle is calculated, converted to a temperature rise through the heat capacity, and the resulting asymmetry between forward and reverse relaxation gives a per-cycle residual strain. The loss tangent, tan δ = (ωτ_α)^{-1}, translates the model to realistic materials. This mechanism converts continuous cyclic loading into a systematic accumulation of infinitesimal irreversible strain, which then crosses a failure threshold.

What would settle it

Measure the surface temperature of an amorphous polymer specimen during high-cycle R=-1 fatigue using a high-speed infrared camera. If the temperature shows a monotonic rise with no per-cycle oscillation, or if the observed S-N slope is significantly different from 3 for a clean, defect-free material, the central mechanism is contradicted. Alternatively, measure A from DMA/tensile parameters and compare predicted lifetimes to experiment: a systematic deviation beyond a factor of a few would falsify the prefactor expression.

Watch

Extended reading notes

Core claim

The central claim is that fatigue failure under high-cycle, fully reversed (R = -1) sinusoidal loading is driven by dissipation-induced breaking of microscopic reversibility. Because the material dissipates energy in the first half-cycle, its temperature rises slightly; the alpha relaxation time shortens in the second half-cycle, so the viscoplastic strain accumulated during loading is not fully erased during unloading. The per-cycle residual strain scales as [tan(δ)]^2 S^3 / (E^2 C T) times (E_a / R T), and equating the accumulated strain after N cycles to a critical failure strain ε* yields N S^3 = (2/π)(R T / E_a)(tan δ)^{-2} E C T σ*, with σ* = E ε*. Thus the exponent m is fixed at 3 and

Load-bearing premise

The derivation assumes that the heat generated in the first half of each load cycle raises the temperature for the second half, and that the temperature then returns to its baseline before the next cycle—if the temperature instead rises monotonically over many cycles, the per-cycle residual strain and the m=3 law do not follow.

Editorial extensions

If this is right

Fatigue lifetimes at millions of cycles can be extrapolated from short, single-cycle measurements: elastic modulus and ultimate strength from a tensile test, loss tangent from DMA, and heat capacity from calorimetry.
The exponent m = 3 is a prediction, so the slope of log-log S-N plots for amorphous polymers under R=-1 high-cycle loading should cluster near 3, not vary arbitrarily between 3 and 12 as is often observed.
The model predicts an explicit dependence of the prefactor A on temperature, frequency, and material aging via the loss tangent, explaining why nominally identical polymers can show very different fatigue lifetimes.
Because damage accumulates linearly, the model gives a direct route to compare testing at different frequencies and temperatures and to design accelerated fatigue tests.
The microscopic interpretation identifies dissipation-induced heating as the physical origin of fatigue damage, in line with experiments showing reversible nonaffine rearrangements under cyclic shear.

Reading between the lines

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

A direct test of the mechanism would be to measure the sample's temperature during a fatigue cycle with high time resolution; the model predicts a small periodic rise-and-fall within each cycle rather than a monotonic drift.
If the temperature accumulates monotonically instead of resetting each cycle, as the authors themselves note, the m=3 law would likely break down; checking this could either validate or bound the model's regime of validity.
The same machinery could be extended to variable-amplitude loading or R≠-1 by tracking per-cycle temperature asymmetry, potentially yielding a history-dependent Basquin exponent.
For notched or pre-cracked samples, the local stress amplitude S varies spatially; using the same per-cycle residual-strain formula with a local S could connect the model to crack growth and Kitagawa-type diagrams.

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

Short version: this paper derives Basquin's law N = A S^{-3} from a linear Maxwell element plus a half-cycle temperature asymmetry. The m=3 exponent is parameter-free and follows from the scaling of eqs 5-10. That is a real, teachable result and the paper should get a fair shot.

What it does well: the derivation is transparent, the comparison with PMMA, PS, PVC shows the slope is consistent with m=3, and the authors are honest about the 'one artificial assumption'—the temperature jump in the first half-cycle that resets before the next cycle. The microscopic interpretation (reversible rearrangements becoming irreversible due to a small thermal asymmetry) is a useful framing, even if it isn't a separate test.

Where it gets soft. First, the prefactor A is not an out-of-sample prediction. The loss tangent is the only fitted parameter, chosen per polymer to match the SN curves; E, C, σ*, and Ea are set to common estimates. So the prefactor agreement is partly a fit. Second, eq 5 has a factor-of-two error: the dissipation rate for a Maxwell dashpot is σ εdot = σ^2/(Eτ), not half of that; as written, ΔQ and A are low by a factor of 2. Easy fix, but it should be fixed. Third, and most important, the thermal reset assumption is load-bearing. The stress-test note is right: if the baseline temperature drifts monotonically, the per-cycle strain increment grows with cycle number, and the failure condition N ε_VP,1 = ε* no longer yields NS^3 = A. The authors' hand-wavy response—that with slow heat exchange each second half is still slightly hotter than the first—does not by itself save the conclusion; you need to bound the drift over N cycles. That doesn't kill the paper, but it needs to be stated carefully. Also, the abstract refers to Long's plasticity model while the text uses a Maxwell element; the mismatch should be cleaned up.

Bottom line: the exponent is new and defensible within the model; the prefactor is conditional on fitted inputs and a premise the authors themselves flag as artificial. Worth engaging seriously rather than dismissing.

Referee Report

4 major / 5 minor

Summary. This manuscript proposes a mean-field mechanism for high-cycle fatigue of glassy polymers under sinusoidal R = -1 loading. Starting from a linear Maxwell element, it computes the viscoplastic strain generated during one half-cycle, assumes the dissipation-induced temperature rise accelerates relaxation in the second half-cycle, and obtains a net residual strain per cycle that scales as σ0^3 (Ea/(RT)) (tanδ)^2 / (E^2 C T). Equating N times this per-cycle damage to a fixed accumulated strain gives Basquin's law NS^3 = A, with A = (2/π)(RT/Ea)(tanδ)^-2 E C T σ*. The formula is compared to published SN data for PMMA, PS, and PVC using tanδ as the sole adjustable parameter, together with common values for E, C, Ea, and σ*, and is reported to agree well. The authors emphasize that the exponent m = 3 is not fitted.

Significance. The paper is potentially significant: if the m = 3 exponent and the prefactor formula are valid, it would provide a simple predictive route from DMA and tensile data to fatigue lifetimes, and it would identify dissipation-induced local heating as a concrete mechanism for irreversible damage. The derivation is transparent, the m = 3 exponent is parameter-free, and the comparison to three independent experimental datasets is a strength. However, the central result is derived under an explicitly artificial thermal-reset assumption, and the prefactor agreement relies on a fitted loss tangent. The paper is therefore an interesting candidate mechanism rather than a fully validated law; its usefulness depends on whether the thermal condition can be made quantitative and whether the prefactor can be tested independently.

major comments (4)

[Discussion ('One artificial assumption…'); Eqs. (5)–(10)] The derivation of Eq. (10) assumes that the per-cycle residual strain is independent of N. This requires the specimen temperature to return to its baseline before every cycle, an assumption the Discussion explicitly calls 'artificial.' The following sentence—that if heat exchange is slow the approach remains valid—does not rescue the derivation: with slow heat exchange the baseline T_n increases with cycle number, so τ_1 in Eq. (8a) decreases and |ε_VP,1| in Eq. (9b) increases as (ωτ_α)^-2. The total damage is then a sum of increasing increments, not N times a constant, and NS^3 = A does not follow from the model. The paper needs a quantitative condition (e.g., thermal relaxation time << cycle period, or an upper bound on accumulated ΔT) under which the per-cycle increment is approximately constant; otherwise m = 3 is conditional on the reset assumption rather than a robust prediction.
[Eq. (5)] The dissipation rate is written as Eτ(εdot_VP)^2/2. From Eq. (3), Eτ εdot_VP = σ, so the mechanical dissipation rate is σ εdot_VP = Eτ(εdot_VP)^2. The extra factor 1/2 in Eq. (5) propagates through Eq. (9) and halves the prefactor A in Eq. (12): the corrected value is A = (1/π)(RT/Ea)(tanδ)^-2 E C T σ*, not (2/π) times that combination. The exponent m = 3 is unaffected, but the prefactor formula needs correction.
[Table 1 and 'only fitting parameter'] The prefactor validation is partly circular. The loss tangent is introduced as 'the only fitting parameter' and the best-fit values in Table 1 are selected to match the experimental SN curves. Since A is proportional to tanδ^-2, the agreement in A is in part constructed: a factor-of-two uncertainty in tanδ changes A by a factor of four. The paper should compare the fitted tanδ with independently measured DMA loss tangents at the same temperature, frequency, and aging state, and should report sensitivity to the common choices E = 3 GPa, σ* = 80 MPa, Ea = 570 kJ/mol, and C = 1.3×10^6 J/(K m^3). This would let the reader judge how predictive Eq. (12) actually is.
[Eqs. (8)–(9) and Table 1] The model identifies τ with the α-relaxation time and Ea with the effective activation energy near Tg, but the experiments are at 293–298 K, well below Tg. At these temperatures the α time is many orders of magnitude longer than the (ω tanδ)^-1 ~ 0.01–1 s implied by the fitted loss tangents; the measured tanδ in glassy polymers also includes secondary relaxations. The authors should clarify whether τ in Eq. (8) is an effective Maxwell relaxation time chosen to reproduce the DMA loss tangent, or the actual segmental relaxation time. If the latter, the claim that the mechanism is α-relaxation-driven needs quantitative support; if the former, the physical interpretation in the 'Microscopic Interpretation' section should be revised accordingly.

minor comments (5)

[Eq. (10)] Eq. (9) yields a negative ε_VP,1 (the minus sign is discussed), but Eq. (10) writes the same quantity as positive. Please use |ε_VP,1| or otherwise state that damage is taken as the absolute value.
[Abstract] The abstract refers to 'Long's plasticity model,' but the text does not define or cite this model; the derivation uses a linear Maxwell element. Either add the appropriate reference or remove the name.
[Figures 1–4] The text refers to colored curves ('the orange curve,' 'the blue curve'). In grayscale printing these are not identifiable; use distinct line styles and/or labels.
[Figs. 2–4] The fitted slopes are said to be 'reasonably consistent' with m = 3, but no slope estimates or confidence intervals are given. A simple regression of log N vs log S for each dataset, restricted to the high-cycle regime, would strengthen the claim.
[Table 1] The values E = 3 GPa, σ* = 80 MPa, Ea = 570 kJ/mol, and C = 1.3×10^6 J/(K m^3) are shared by all three polymers. This is a reasonable first approximation, but a short sensitivity analysis (e.g., factor-of-two variations) would help readers assess the uncertainty in A.

Circularity Check

1 steps flagged · score 6.0 of 10

Prefactor validation reduces to a tanδ fit, while the m=3 exponent is an independent, non-fitted result.

fitted input called prediction [Results and Discussion, paragraph following Table 1]
"The only fitting parameter was the loss tangent, and the best-fit values (between 0.03 and 0.15) are indeed consistent with the experimentally observed loss tangents for glassy polymers."

Eq. 12 defines A=(2/pi)(RT/Ea)[tan(delta)]^(-2) E C T sigma*, so the prefactor entering the SN curve is inversely proportional to [tan(delta)]^2. The paper determines tan(delta) as the 'only fitting parameter' to make the theoretical curves match the experimental SN data, then presents the same comparison as successful validation. Hence the prefactor comparison is forced by the fit rather than an independent prediction. The exponent m=3 is not fitted and remains an independent consequence of the model, so this is partial circularity, not complete.

full rationale

The derivation from the Maxwell element model to epsilon_VP,1 proportional to sigma0^3/[omega*tau]^2 is algebraically self-contained; substituting tan(delta)=(omega*tau)^(-1) is exact for the Maxwell element, and the m=3 exponent follows without fitting. The comparison to PS/PMMA/PVC uses external experimental data. However, the prefactor validation is not independent because tan(delta) is fit to the same SN curves (A is proportional to tan(delta)^(-2)). The Discussion's 'one artificial assumption'—that temperature resets each half-cycle—is an acknowledged physical limitation, not a circularity; it conditions the m=3 result but does not make the derivation equivalent to its inputs. No load-bearing self-citation chain is present.

Assumptions & free parameters 7 free parameters · 6 assumptions · 0 invented entities

The m=3 exponent is robust: it follows from per-cycle damage ∝ σ0^3 regardless of the exact prefactor. The prefactor A, however, rests on several assumed material constants and a fitted tanδ, plus the artificial half-cycle temperature asymmetry.

free parameters (7)

tan δ (PMMA) = 0.033
Best-fit loss tangent; enters A as tanδ^{-2}; used to match PMMA SN data.
tan δ (PS) = 0.15
Best-fit loss tangent; enters A as tanδ^{-2}; used to match PS SN data.
tan δ (PVC) = 0.076
Best-fit loss tangent; enters A as tanδ^{-2}; used to match PVC SN data.
Activation energy Ea = 570 kJ/mol
Estimated from fragility m≈100-150 via Ea≈2.303mRTg; same for all polymers; enters A linearly; not independently measured for the tested samples.
Ultimate strength σ* = 80 MPa
Assumed identical for PS, PMMA, PVC; enters A linearly; not measured per sample.
Young's modulus E = 3e3 MPa
Assumed common value; enters A linearly.
Heat capacity per volume C = 1.3e6 J/(K m3)
Assumed common value; enters A linearly.

assumptions (6)

domain assumption Maxwell element relation Eτ dε_VP/dt = σ(t) describes the viscoplastic strain.
Used in eqs 3a-3b; no derivation from polymer microstructure; the abstract mentions Long's plasticity model but the text does not define it.
ad hoc to paper Temperature increases by ΔT=ΔQ/C after the first half-cycle and returns to baseline before the next cycle.
The authors themselves call this 'artificial' in the Discussion; without this asymmetry the per-cycle residual strain is exactly zero.
ad hoc to paper Failure occurs when accumulated viscoplastic strain reaches a pre-determined ε*.
Introduced before eq 10; no fracture-mechanics derivation; σ*=Eε* is set to 80 MPa for all polymers.
ad hoc to paper The α-relaxation time follows an Arrhenius law with a single activation energy Ea near Tg.
Eqs 8a-8c; the text notes Ea is actually non-Arrhenius, temperature-dependent, and history-dependent.
domain assumption tanδ = (ωτ_alpha)^{-1} generalizes the Maxwell result to real polymers.
Used to convert eq 9a to 9b; real loss tangent is frequency and history dependent.
domain assumption Damage accumulates linearly: N ε_VP,1 = ε*.
Used in eq 10; acceleration of damage at late stages is neglected, as the authors acknowledge.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Modeling the Fatigue Behavior of Amorphous Polymers." pith.science (2026). https://pith.science/paper/SW4OHEIZ

@misc{pith2026260714994,
  author       = {Pith},
  title        = {Pith review of: Modeling the Fatigue Behavior of Amorphous Polymers},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SW4OHEIZ}},
  note         = {Machine review of arXiv:2607.14994}
}

read the original abstract

Prediction of material durability is both very important and very difficult. In many cases, material durability is measured by subjecting a sample to repeating oscillatory cycles (in shear or tension-compression) until it fails in either ductile or brittle fashion. Typically, the stress amplitude is denoted S, and the number of cycles N, so the resulting dependence is known as the SN-curve. For many materials, SN curve has been shown to obey the empirical Basquin's law, N = AS^(-m), where the prefactor A was a function of the temperature, load frequency, and sample history, and the power law m was a real number, usually between 3 and 12. Here, we derive the Basquin's law using a linearized version of the Long's plasticity model and demonstrate that within this framework, m = 3. We also derive the expression for the prefactor A. Finally, we show that our theory successfully describes experimental data for three amorphous polymers, PS, PMMA, and PVC.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

7 extracted references · 4 canonical work pages

[1]

(1) Burhan, I.; Kim, H. S. S -n Curve Models for Composite Materials Characterisation: An Evaluative Review. Journal of Composites Science 2018, 2 (3),

2018
[38]

(2) Sauer, J

https://doi.org/10.3390/jcs2030038. (2) Sauer, J. A.; Richardson, G. C. Fatigue of Polymers. Int. J. Fract. 1980, 16 (6), 499–532. (3) Riddell, M. N.; Koo, G. P .; O’Toole, J. L. Fatigue Mechanisms of Thermoplastics. Polym. Eng. Sci. 1966, 6 (4), 363–368. https://doi.org/10.1002/pen.760060414. (4) Maity, S.; Bhaumik, H.; Athani, S.; Sastry, S. Fatigue Fai...

work page doi:10.3390/jcs2030038 1980
[2013]

Slow Stretched -Exponential and Fast Compressed -Exponential Relaxation from Local Event Dynamics

(25) Trachenko, K.; Zaccone, A. Slow Stretched -Exponential and Fast Compressed -Exponential Relaxation from Local Event Dynamics. Journal of Physics: Condensed Matter 2021, 33 (31), 315101. https://doi.org/10.1088/1361-648X/ac04cd. (26) Ginzburg, V. V; Gendelman, O. V; Zaccone, A. Unifying Physical Framework for Stretched - Exponential, Compressed-Expone...

work page doi:10.1088/1361-648x/ac04cd 2021
[2019]

On the Inverse Power Laws for Accelerated Random Fatigue Testing

(10) Allegri, G.; Zhang, X. On the Inverse Power Laws for Accelerated Random Fatigue Testing. Int. J. Fatigue 2008, 30, 967–977. https://doi.org/10.1016/j.ijfatigue.2007.08.023. (11) Saad, N. The Fatigue Behavior of Composite Materials for High -Temperature Applications. In Lightweight Composite Structures in Transport; Elsevier, 2016; pp 239–266. (12) Ri...

work page doi:10.1016/j.ijfatigue.2007.08.023 2008
[2024]

(5) Hadi, S.; Hardjito, A.; Wicaksono, H.; Cahyono, T

https://doi.org/10.1038/s41567-026-03174-x. (5) Hadi, S.; Hardjito, A.; Wicaksono, H.; Cahyono, T. S. A.; Wirawan, W.; Emzain, Z. F. Fatigue Life Prediction of Injection Molded Polymer Materials; 2023; pp 235 –248. https://doi.org/10.2991/978-94-6463-358-0_24. (6) Janssen, R. P . M.; de Kanter, D.; Govaert, L. E.; Meijer, H. E. H. Fatigue Life Predictions...

arXiv 2023
[2026]

Predicting the Brittle-to-Ductile Transition in Amorphous Polymers

https://doi.org/https://doi.org/10.48550/arXiv.2605.04753. (20) Okeke, C. P .; Thite, A. N.; Durodola, J. F.; Greenrod, M. T. A Novel Test Rig for Measuring Bending Fatigue Using Resonant Behaviour. In Procedia Structural Integrity; Elsevier B.V., 2018; Vol. 13, pp 1470–1475. https://doi.org/10.1016/j.prostr.2018.12.303. (21) Iacopi, A. V.; White, J. R. R...

work page Pith review arXiv doi:10.48550/arxiv.2605.04753 2018
[3707]

B.; Simon, S

(15) McKenna, G. B.; Simon, S. L. 50th Anniversary Perspective: Challenges in the Dynamics and Kinetics of Glass-Forming Polymers. Macromolecules 2017, 50 (17), 6333–6361. (16) Mauro, J. C.; Yue, Y .; Ellison, A. J.; Gupta, P . K.; Allan, D. C. Viscosity of Glass-Forming Liquids. Proceedings of the National Academy of Sciences 2009, 106 (47), 19780–19784....

arXiv 2017

Pith tools

Reviewed August 2, 2026 · model on record in the stance chip above.

[1] [1]

(1) Burhan, I.; Kim, H. S. S -n Curve Models for Composite Materials Characterisation: An Evaluative Review. Journal of Composites Science 2018, 2 (3),

2018

[2] [38]

(2) Sauer, J

https://doi.org/10.3390/jcs2030038. (2) Sauer, J. A.; Richardson, G. C. Fatigue of Polymers. Int. J. Fract. 1980, 16 (6), 499–532. (3) Riddell, M. N.; Koo, G. P .; O’Toole, J. L. Fatigue Mechanisms of Thermoplastics. Polym. Eng. Sci. 1966, 6 (4), 363–368. https://doi.org/10.1002/pen.760060414. (4) Maity, S.; Bhaumik, H.; Athani, S.; Sastry, S. Fatigue Fai...

work page doi:10.3390/jcs2030038 1980

[3] [2013]

Slow Stretched -Exponential and Fast Compressed -Exponential Relaxation from Local Event Dynamics

(25) Trachenko, K.; Zaccone, A. Slow Stretched -Exponential and Fast Compressed -Exponential Relaxation from Local Event Dynamics. Journal of Physics: Condensed Matter 2021, 33 (31), 315101. https://doi.org/10.1088/1361-648X/ac04cd. (26) Ginzburg, V. V; Gendelman, O. V; Zaccone, A. Unifying Physical Framework for Stretched - Exponential, Compressed-Expone...

work page doi:10.1088/1361-648x/ac04cd 2021

[4] [2019]

On the Inverse Power Laws for Accelerated Random Fatigue Testing

(10) Allegri, G.; Zhang, X. On the Inverse Power Laws for Accelerated Random Fatigue Testing. Int. J. Fatigue 2008, 30, 967–977. https://doi.org/10.1016/j.ijfatigue.2007.08.023. (11) Saad, N. The Fatigue Behavior of Composite Materials for High -Temperature Applications. In Lightweight Composite Structures in Transport; Elsevier, 2016; pp 239–266. (12) Ri...

work page doi:10.1016/j.ijfatigue.2007.08.023 2008

[5] [2024]

(5) Hadi, S.; Hardjito, A.; Wicaksono, H.; Cahyono, T

https://doi.org/10.1038/s41567-026-03174-x. (5) Hadi, S.; Hardjito, A.; Wicaksono, H.; Cahyono, T. S. A.; Wirawan, W.; Emzain, Z. F. Fatigue Life Prediction of Injection Molded Polymer Materials; 2023; pp 235 –248. https://doi.org/10.2991/978-94-6463-358-0_24. (6) Janssen, R. P . M.; de Kanter, D.; Govaert, L. E.; Meijer, H. E. H. Fatigue Life Predictions...

arXiv 2023

[6] [2026]

Predicting the Brittle-to-Ductile Transition in Amorphous Polymers

https://doi.org/https://doi.org/10.48550/arXiv.2605.04753. (20) Okeke, C. P .; Thite, A. N.; Durodola, J. F.; Greenrod, M. T. A Novel Test Rig for Measuring Bending Fatigue Using Resonant Behaviour. In Procedia Structural Integrity; Elsevier B.V., 2018; Vol. 13, pp 1470–1475. https://doi.org/10.1016/j.prostr.2018.12.303. (21) Iacopi, A. V.; White, J. R. R...

work page Pith review arXiv doi:10.48550/arxiv.2605.04753 2018

[7] [3707]

B.; Simon, S

(15) McKenna, G. B.; Simon, S. L. 50th Anniversary Perspective: Challenges in the Dynamics and Kinetics of Glass-Forming Polymers. Macromolecules 2017, 50 (17), 6333–6361. (16) Mauro, J. C.; Yue, Y .; Ellison, A. J.; Gupta, P . K.; Allan, D. C. Viscosity of Glass-Forming Liquids. Proceedings of the National Academy of Sciences 2009, 106 (47), 19780–19784....

arXiv 2017