pith. machine review for the scientific record. sign in

arxiv: 2605.13057 · v1 · submitted 2026-05-13 · ⚛️ physics.app-ph

Recognition: 2 theorem links

· Lean Theorem

Identifiability Limits in Ultrasonic Microstructure Characterisation: A Canonical and Stochastic Framework

Wei Yi Yeoh

Pith reviewed 2026-05-14 01:53 UTC · model grok-4.3

classification ⚛️ physics.app-ph
keywords ultrasonic microstructureidentifiability limitsforward operatorGaussian random fieldsattenuationvelocityFisher informationsensitivity analysis
0
0 comments X

The pith

Identifiability limits in ultrasonic microstructure characterization arise from forward-map structure and intrinsic variability rather than rank deficiency alone.

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

The paper directly examines what information is available in ultrasonic responses for recovering microstructure parameters. It uses a canonical pulse-echo model to derive closed-form sensitivity limits from parameter coupling and interface saturation, then extends to stochastic surrogates modeled as Gaussian random fields with parameters D for correlation length and T for texture coherence. The forward map to attenuation and velocity observables is shown to be structurally full rank yet strongly anisotropic in sensitivity. When intrinsic variability is added, a variance-weighted Fisher analysis demonstrates that practical recoverability depends on the balance of sensitivity strength against stochastic spread. Inversion experiments confirm that single observables produce poorly conditioned landscapes while combined observables improve results through complementary information.

Core claim

Within the feature-level framework, identifiability limits are governed primarily by forward-map structure and intrinsic variability. For the canonical model, closed-form sensitivity analysis reveals information limits from parameter coupling, dimensional restriction, and interface-driven saturation. For Gaussian random field surrogates, the map from D and T to attenuation and velocity remains structurally full rank, but the sensitivity geometry is anisotropic and practical identifiability decreases further when microstructural variability is incorporated; recoverability is then set by the balance between sensitivity magnitude and stochastic variability rather than structural rank alone.

What carries the argument

Variance-weighted Fisher information applied to the sensitivity geometry of the forward operator mapping correlation length D and texture-coherence T to attenuation and velocity observables.

If this is right

  • Single attenuation or velocity observables alone produce elongated and weakly constrained objective landscapes during inversion.
  • Combined use of attenuation and velocity observables improves conditioning through their complementary sensitivities.
  • Practical recoverability decreases as intrinsic microstructural variability increases, even when the forward map is full rank.
  • Observable selection and measurement design should prioritize configurations that exploit the anisotropic sensitivity structure.

Where Pith is reading between the lines

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

  • Measurement protocols could be optimized by first estimating expected variability levels to decide whether combined observables are required.
  • Similar forward-map geometry analysis might apply to other wave-based techniques such as elastic or electromagnetic scattering for microstructure recovery.
  • Sensor array designs could target frequency ranges where sensitivity to D versus T is most balanced to reduce anisotropy effects.
  • The framework suggests testing whether real-material deviations from Gaussian statistics further degrade identifiability beyond the modeled limits.

Load-bearing premise

The Gaussian random fields used as surrogate microstructures capture the essential statistical properties of real materials and the forward map remains structurally full rank under these choices.

What would settle it

Inversion trials on measured ultrasonic data from actual microstructures where the recovered uncertainty in D and T fails to match the variance-weighted Fisher predictions for the modeled attenuation and velocity observables.

Figures

Figures reproduced from arXiv: 2605.13057 by Wei Yi Yeoh.

Figure 1
Figure 1. Figure 1: Schematic of the identifiability analysis framework. Microstructure parameters are mapped to feature-level observables through wave propagation in the stochastic media. Sensitivity is quantified via the Jacobian, which is normalised using the covariance of stochastic microstructure variability. The resulting information geometry, characterised by singular values and the Fisher Information Matrix, determine… view at source ↗
Figure 2
Figure 2. Figure 2: An illustration of the canonical model configuration with a Ti-64 slab with unknown material properties embedded in water. solid block are represented by the parameter vector 𝐏 = ⎡ ⎢ ⎢ ⎢ ⎣ 𝜌 𝑑 𝛼 𝑍 ⎤ ⎥ ⎥ ⎥ ⎦ , (1) where 𝜌 is the density, 𝑑 is the thickness, 𝛼 is an effective attenuation coefficient which represents a lumped description of absorption and scattering losses, and 𝑍 is the acoustic impedance, de… view at source ↗
Figure 3
Figure 3. Figure 3: Singular value spectra of the log-parameter scaled Jacobian as a function of impedance contrast 𝑍∕𝑍𝑤. (a) Full parameterisation (𝐏 = [𝜌, 𝑑, 𝛼, 𝑍] 𝑇 ) exhibiting persistent spectral separation and a near-null direction associated with structural parameter coupling. (b) Reduced parameterisation (𝐏𝐶 = [𝛼, 𝑍] 𝑇 ) in which the weakly observable mode is removed, yielding balanced sensitivity within the identifia… view at source ↗
Figure 4
Figure 4. Figure 4: Jacobian condition number versus impedance contrast 𝑍∕𝑍𝑤 for the full (Case 1) and reduced (Case 2) parameterisations. The strong separation between the two curves shows that density–thickness coupling in Case 1 produces a near-null direction and severe ill-conditioning [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Finite-difference validation of canonical results comparing (a) Time-of-flight 𝜏 versus impedance contrast and (b) log-magnitude of the front-wall reflection coefficient log |𝑅𝑓 | versus impedance contrast. The solid curves represents the canonical results, and markers denote the FDTD measurements, with good agreement between them. small deviations at extreme contrast due to poorer parameter sensitivity. T… view at source ↗
Figure 6
Figure 6. Figure 6: Examples of surrogate microstructures generated using Gaussian random fields. (a) Low correlation length and low texture-coherence parameter (𝐷 = 0.5 mm, 𝑇 = 0.001), showing fine-scale heterogeneity and broadly distributed orientations. (b) High correlation length and higher texture-coherence parameter (𝐷 = 1.4 mm, 𝑇 = 0.01), showing coarse heterogeneity with stronger orientation coherence. (c) Orientation… view at source ↗
Figure 7
Figure 7. Figure 7: Mean and variability of ultrasonic observables obtained from Gaussian random field (GRF) surrogate microstructures across the correlation length (𝐷) and texture-coherence parameter (𝑇 ) space across 40 realisations. (a) Mean attenuation 𝛼, (b) mean longitudinal wave velocity 𝑣, (c) standard deviation of attenuation, and (d) standard deviation of velocity computed at each (𝐷, 𝑇 ) combination. where 𝜇𝑘 denot… view at source ↗
Figure 8
Figure 8. Figure 8: Comparison of raw and normalised sensitivity metrics across the (𝐷, 𝑇 ) domain. (a)–(b) shows the largest singular value (SV1), (c)–(d) the smallest singular value (SV2), and (e)–(f) the corresponding condition number of the Jacobian. The raw Jacobian reflects absolute sensitivity of the observables, while the normalised Jacobian represents proportional (scale-independent) sensitivity. The maps illustrate … view at source ↗
Figure 9
Figure 9. Figure 9: Whitened sensitivity metrics across the (𝐷, 𝑇 ) domain, incorporating intrinsic microstructure variability. (a)–(b) show the singular values of the whitened Jacobian, (c) the corresponding condition number, and (d) the angular separation between sensitivity directions. Whitening scales the Jacobian by the inverse covariance of the observables, so that sensitivity is measured relative to variability. This r… view at source ↗
Figure 10
Figure 10. Figure 10: Practical identifiability map over the (𝐷, 𝑇 ) parameter space based on a composite index combining the variance-weighted information floor and a softened anisotropy measure. Higher values indicate regions where the weakest parameter direction remains more informative and the local sensitivity geometry is more balanced, corresponding to more favourable inverse-problem conditioning. 3.3.2. Inversion of Sur… view at source ↗
Figure 11
Figure 11. Figure 11: Inversion landscapes for two representative parameter configurations using attenuation only, velocity only, and both observables for inversion. The colour map shows the least-squares objective (𝐷, 𝑇 ), with the white markers denoting the ground-truth and red markers the recovered solutions. 4.1. From Canonical To Surrogate Identifiability In the canonical slab, the limitation is explicit: the observable … view at source ↗
read the original abstract

Ultrasound for microstructure characterisation is increasingly studied and is often assessed through inversion performance. However, the framework is fundamentally constrained by the information content available in the measured response. Hence, this work examines identifiability directly by analysing the geometry of the forward operator in both a canonical pulse-echo model and a stochastic surrogate microstructure. For the canonical model, a closed-form sensitivity analysis reveals information limits arising from parameter coupling, dimensional restriction, and interface-driven saturation. For the surrogate microstructures represented by Gaussian random fields, the forward map from correlation length $D$ and texture-coherence parameter $T$ to the attenuation and velocity observables remains structurally full rank. However, the sensitivity geometry is strongly anisotropic, with uneven parameter influence across the observable space. When intrinsic microstructural variability is incorporated, practical identifiability is further reduced. A variance-weighted Fisher framework shows that recoverability is governed by the balance between sensitivity magnitude and stochastic variability, rather than by structural rank alone. Inversion results confirm this behaviour: single observables produce elongated and weakly constrained objective landscapes, whereas combined observables improve conditioning through complementary sensitivities. These results show that, within the feature-level framework considered here, identifiability limits are governed primarily by forward-map structure and intrinsic variability, with direct implications for observable selection and measurement design.

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 / 1 minor

Summary. The manuscript examines identifiability limits in ultrasonic microstructure characterisation by analysing the geometry of the forward operator. In a canonical pulse-echo model, closed-form sensitivity analysis identifies limits from parameter coupling, dimensional restriction, and interface saturation. For stochastic surrogates based on Gaussian random fields with parameters D (correlation length) and T (texture coherence), the forward map to attenuation and velocity observables is shown to be structurally full rank but strongly anisotropic in sensitivity. A variance-weighted Fisher framework demonstrates that intrinsic microstructural variability further reduces practical recoverability, with numerical inversions confirming that single observables yield elongated objective landscapes while combined observables improve conditioning. The central conclusion is that identifiability is governed primarily by forward-map structure and variability, with implications for observable selection and measurement design.

Significance. If the results hold, this work provides a useful theoretical lens on why ultrasound inversion for microstructure parameters is often ill-conditioned, directly informing experimental choices in nondestructive evaluation. The combination of analytical closed-form results, Fisher-information geometry, and supporting inversions is a strength, as is the explicit separation of structural rank from practical recoverability due to variability. The framework yields concrete guidance on when and why combining observables helps, which could be tested experimentally.

major comments (2)
  1. [Stochastic surrogate microstructures] Stochastic surrogate section (abstract and main text): The conclusion that identifiability limits are governed by forward-map structure and intrinsic variability, with direct implications for real materials, rests on the premise that Gaussian random fields with parameters D and T adequately capture the essential statistical properties of real microstructures. No comparison is provided to measured statistics such as grain-size distributions or spatial correlations from EBSD maps, which is load-bearing for the claim that the reported anisotropy and rank properties reflect physical limits rather than surrogate artifacts.
  2. [Variance-weighted Fisher framework] Variance-weighted Fisher framework (abstract): The statement that recoverability is governed by the balance between sensitivity magnitude and stochastic variability requires explicit derivation of the weighting procedure and the resulting information matrix; without this, it is unclear whether the reduction in practical identifiability is quantitatively demonstrated or follows by construction from the chosen variance model.
minor comments (1)
  1. [Notation and model definition] Clarify the precise definition of the texture-coherence parameter T and its relation to the correlation function in the Gaussian random field model to avoid ambiguity in the forward-map description.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive and insightful comments. We address each major comment point by point below and indicate the planned revisions.

read point-by-point responses

  1. Referee: [Stochastic surrogate microstructures] Stochastic surrogate section (abstract and main text): The conclusion that identifiability limits are governed by forward-map structure and intrinsic variability, with direct implications for real materials, rests on the premise that Gaussian random fields with parameters D and T adequately capture the essential statistical properties of real microstructures. No comparison is provided to measured statistics such as grain-size distributions or spatial correlations from EBSD maps, which is load-bearing for the claim that the reported anisotropy and rank properties reflect physical limits rather than surrogate artifacts.

    Authors: We agree that direct validation against experimental microstructure statistics (e.g., EBSD grain-size distributions or spatial correlations) would strengthen the link to real materials. The Gaussian random field model is adopted as a canonical, analytically tractable surrogate that encodes the key statistical features of correlation length and texture coherence, enabling the closed-form sensitivity analysis. The reported anisotropy and structural rank properties derive directly from the geometry of the forward map for this class of models rather than from specific distributional details. In the revision we will expand the discussion of the surrogate's scope and limitations, explicitly noting that the framework is extensible to other microstructure representations while clarifying that the present results pertain to this standard stochastic class. revision: partial

  2. Referee: [Variance-weighted Fisher framework] Variance-weighted Fisher framework (abstract): The statement that recoverability is governed by the balance between sensitivity magnitude and stochastic variability requires explicit derivation of the weighting procedure and the resulting information matrix; without this, it is unclear whether the reduction in practical identifiability is quantitatively demonstrated or follows by construction from the chosen variance model.

    Authors: We will make the derivation explicit in the revised manuscript. The variance-weighted Fisher information matrix is obtained by premultiplying the conventional Fisher matrix by the inverse of the covariance matrix of the microstructure-induced fluctuations in the observables. This weighting quantitatively attenuates the information content in directions where stochastic variability is large relative to sensitivity, producing the observed reduction in practical recoverability. The effect is not tautological but follows from embedding the measured variance structure into the information geometry; we will include the full algebraic steps and a brief numerical illustration in a dedicated subsection or appendix. revision: yes

Circularity Check

0 steps flagged

No significant circularity: identifiability limits derived from independent forward-operator geometry and Fisher analysis

full rationale

The paper derives its identifiability conclusions through explicit mathematical examination of the forward operator in a canonical pulse-echo model (closed-form sensitivity analysis revealing parameter coupling and saturation) and in Gaussian random field surrogates (structural rank and anisotropic sensitivity of the map from D and T to attenuation/velocity). The variance-weighted Fisher framework then incorporates intrinsic variability to assess practical recoverability. None of these steps reduce by construction to fitted parameters renamed as predictions, self-definitions, or load-bearing self-citations; the rank and anisotropy properties follow directly from the stated model definitions and are presented as verifiable via the described numerical inversions and objective landscapes. The framework is therefore self-contained against external benchmarks of the forward map.

Axiom & Free-Parameter Ledger

2 free parameters · 1 axioms · 0 invented entities

The framework relies on standard assumptions from wave propagation and stochastic process modeling; no new entities are postulated.

free parameters (2)
  • correlation length D
    Microstructural parameter whose sensitivity to observables is analyzed in the stochastic surrogate model.
  • texture-coherence parameter T
    Microstructural parameter whose sensitivity to observables is analyzed in the stochastic surrogate model.
axioms (1)
  • domain assumption Microstructures can be represented by Gaussian random fields
    Used to construct the stochastic surrogate for which the forward map is examined.

pith-pipeline@v0.9.0 · 5520 in / 1383 out tokens · 50463 ms · 2026-05-14T01:53:56.001342+00:00 · methodology

discussion (0)

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

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

Reference graph

Works this paper leans on

80 extracted references · 80 canonical work pages

  1. [1]

    Rose.Unbounded Isotropic and Anisotropic Media

    Joseph L. Rose.Unbounded Isotropic and Anisotropic Media. Cambridge University Press, 1999

    work page 1999

  2. [2]

    Stanke and G

    Fred E. Stanke and G. S. Kino. A unified theory for elastic wave propagation in polycrystalline materials.The Journal of the Acoustical Society of America, 75(3):665–681, 1984. ISSN 0001-4966. doi: 10.1121/1.390577

    work page doi:10.1121/1.390577 1984

  3. [3]

    Richard L. Weaver. Diffusivity of Ultrasound in Polycrystals.Journal of the Mechanics and Physics of Solids, 38(1):55–86, 1990. doi: 10.1007/978-3-662-44324-8_1289

    work page doi:10.1007/978-3-662-44324-8_1289 1990

  4. [4]

    Modeling Ultrasonic Microstructural Noise in Titanium Alloys

    F J Margetan and R B Thompson. Modeling Ultrasonic Microstructural Noise in Titanium Alloys. InReview of Progress in Quantitative Nondestructive Evaluation, volume 12, pages 1735–1742, 1993

    work page 1993

  5. [5]

    BackscatteredMicrostructuralNoiseinUltrasonicToneburstInspections.Journal of Nondestructive Evaluation, 13(3):111–136, 1994

    F.J.Margetan,R.B.Thompson,andI.Yalda-Mooshabad. BackscatteredMicrostructuralNoiseinUltrasonicToneburstInspections.Journal of Nondestructive Evaluation, 13(3):111–136, 1994. ISSN 1573-4862. doi: 10.1007/BF00728250

    work page doi:10.1007/bf00728250 1994

  6. [6]

    Numericalandanalyticmodellingofelastodynamicscatteringwithinpolycrystalline materials.The Journal of the Acoustical Society of America, 143(4):2394–2408, 2018

    A.VanPamel,G.Sha,M.J.S.Lowe,andS.I.Rokhlin. Numericalandanalyticmodellingofelastodynamicscatteringwithinpolycrystalline materials.The Journal of the Acoustical Society of America, 143(4):2394–2408, 2018. ISSN 0001-4966. doi: 10.1121/1.5031008

    work page doi:10.1121/1.5031008 2018

  7. [7]

    Effectoftextureandgrainshapeonultrasonicbackscatteringinpolycrystals.Ultrasonics,54(7):1789–1803,

    J.Li,L.Yang,andS.I.Rokhlin. Effectoftextureandgrainshapeonultrasonicbackscatteringinpolycrystals.Ultrasonics,54(7):1789–1803,

    work page

  8. [8]

    doi: 10.1016/j.ultras.2014.02.020

    work page doi:10.1016/j.ultras.2014.02.020 2014

  9. [9]

    Shape and size evaluations of elongated grains using phased array ultrasound and directional backscattering method.NDT & E International, 129:102634, 2022

    Yu Liu, Qiang Tian, Ping Yu, Jingjing He, and Xuefei Guan. Shape and size evaluations of elongated grains using phased array ultrasound and directional backscattering method.NDT & E International, 129:102634, 2022. doi: 10.1016/j.ndteint.2022.102634

    work page doi:10.1016/j.ndteint.2022.102634 2022

  10. [11]

    A new correlation model for ultrasonic attenuation in polycrystals with broad grain size distributions

    Ningyue Sheng and Shahram Khazaie. A new correlation model for ultrasonic attenuation in polycrystals with broad grain size distributions. Ultrasonics, 160:107924, 2026. doi: 10.1016/j.ultras.2025.107924

    work page doi:10.1016/j.ultras.2025.107924 2026

  11. [12]

    Propagation and scattering of ultrasonic waves in macroscopically anisotropic polycrystalline materials with fiber texture.Ultrasonics, 164:108026, 2026

    Juan Camilo Victoria-Giraldo, Denis Solas, Jérôme Laurent, Alain Lhémery, and Bing Tie. Propagation and scattering of ultrasonic waves in macroscopically anisotropic polycrystalline materials with fiber texture.Ultrasonics, 164:108026, 2026. doi: 10.1016/j.ultras.2026.108026

    work page doi:10.1016/j.ultras.2026.108026 2026

  12. [13]

    Ultrasonicbackscatteringinpolycrystalswithelongatedsinglephaseandduplexmicrostructures

    O.I.Lobkis,L.Yang,J.Li,andS.I.Rokhlin. Ultrasonicbackscatteringinpolycrystalswithelongatedsinglephaseandduplexmicrostructures. Ultrasonics, 52(6):694–705, 2012. ISSN 0041624X. doi: 10.1016/j.ultras.2011.12.002

    work page doi:10.1016/j.ultras.2011.12.002 2012

  13. [14]

    Ultrasonic backscattering model for rayleigh waves in polycrystals with born and independent scattering approximations.Ultrasonics, 140:107297, 2024

    Shan Li, Ming Huang, Yongfeng Song, Bo Lan, and Xiongbing Li. Ultrasonic backscattering model for rayleigh waves in polycrystals with born and independent scattering approximations.Ultrasonics, 140:107297, 2024. doi: 10.1016/j.ultras.2024.107297

    work page doi:10.1016/j.ultras.2024.107297 2024

  14. [15]

    Van Pamel, G

    A. Van Pamel, G. Sha, S. I. Rokhlin, and M. J. S. Lowe. Finite-element modelling of elastic wave propagation and scattering within heterogeneous media.Proceedings of the Royal Society A: Mathematical, Physical and Engineering Science, 473(2197), 2017. ISSN 1364-

    work page 2017

  15. [16]

    doi: 10.1098/rspa.2016.0738

    work page doi:10.1098/rspa.2016.0738 2016

  16. [17]

    Numericalestimationofultrasonic phase velocity and attenuation for longitudinal and shear waves in polycrystalline materials.Ultrasonics, 148:107517, 2025

    VincentDorval,NicolasLeymarie,AlexandreImperiale,EdouardDemaldent,andPierre-EmileLhuillier. Numericalestimationofultrasonic phase velocity and attenuation for longitudinal and shear waves in polycrystalline materials.Ultrasonics, 148:107517, 2025. doi: 10.1016/j.ultras.2024.107517

    work page doi:10.1016/j.ultras.2024.107517 2025

  17. [18]

    Comparison of ultrasonic attenuation within two- and three-dimensional polycrystalline media

    X Bai, B Tie, J Schmitt, and D Aubry. Comparison of ultrasonic attenuation within two- and three-dimensional polycrystalline media. Ultrasonics, 100(December 2018):105980, 2020. ISSN 0041-624X. doi: 10.1016/j.ultras.2019.105980

    work page doi:10.1016/j.ultras.2019.105980 2018

  18. [20]

    Large-scale3Drandompolycrystalsforthefiniteelementmethod:Generation,meshingandremeshing

    R.Quey,P.R.Dawson,andF.Barbe. Large-scale3Drandompolycrystalsforthefiniteelementmethod:Generation,meshingandremeshing. ComputerMethodsinAppliedMechanicsandEngineering,200(17-20):1729–1745,2011. ISSN00457825. doi:10.1016/j.cma.2011.01.002

    work page doi:10.1016/j.cma.2011.01.002 2011

  19. [21]

    ISSN 0001-4966

    AndreaP.Arguelles,ChristopherM.Kube,PingHu,andJosephA.Turner.Mode-convertedultrasonicscatteringinpolycrystalswithelongated grains.The Journal of the Acoustical Society of America, 140(3):1570–1580, 2016. ISSN 0001-4966. doi: 10.1121/1.4962161

    work page doi:10.1121/1.4962161 2016

  20. [22]

    Huang, G

    M. Huang, G. Sha, P. Huthwaite, S. I. Rokhlin, and M. J. S. Lowe. Elastic wave velocity dispersion in polycrystals with elongated grains: Theoretical and numerical analysis.The Journal of the Acoustical Society of America, 148(6):3645–3662, 2020. ISSN 0001-4966. doi: 10.1121/10.0002916

    work page doi:10.1121/10.0002916 2020

  21. [23]

    Huang, P

    M. Huang, P. Huthwaite, S. I. Rokhlin, and M. J. S. Lowe. Finite-element and semi-analytical study of elastic wave propagation in strongly scattering polycrystals.Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences, 478, 2022. ISSN 14712946. doi: 10.1098/rspa.2021.0850

    work page doi:10.1098/rspa.2021.0850 2022

  22. [24]

    Wavepropagationinhighlyanisotropicpolycrystals: A numerical perspective from an unstructured-mesh-based high-order finite element method.Ultrasonics, 159:107882, 2026

    ShaojieGong,ShifengGuo,YiXiong,ShiyuanZhou,FangsenCui,andMenglongLiu. Wavepropagationinhighlyanisotropicpolycrystals: A numerical perspective from an unstructured-mesh-based high-order finite element method.Ultrasonics, 159:107882, 2026. doi: 10.1016/j.ultras.2025.107882

    work page doi:10.1016/j.ultras.2025.107882 2026

  23. [25]

    Li and S

    J. Li and S. I. Rokhlin. Propagation and scattering of ultrasonic waves in polycrystals with arbitrary crystallite and macroscopic texture symmetries.Wave Motion, 58:145–164, 2015. ISSN 01652125. doi: 10.1016/j.wavemoti.2015.05.004

    work page doi:10.1016/j.wavemoti.2015.05.004 2015

  24. [26]

    W. Y. Yeoh, B. Lan, and M. J. S. Lowe. Investigation of the influence of macrozones in titanium alloys on the propagation and scattering of ultrasound.Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences, 479, 9 2023. ISSN 1364-5021. doi: 10.1098/rspa.2023.0176. URLhttps://royalsocietypublishing.org/doi/10.1098/rspa.2023.0176

    work page doi:10.1098/rspa.2023.0176 2023

  25. [27]

    Ultrasonic backscattering model of lamellar duplex phase microstructures in polycrystalline materials.Ultrasonics, 149:107581, 2025

    Zenghua Liu, Jinlong Li, Yang Zheng, and Cunfu He. Ultrasonic backscattering model of lamellar duplex phase microstructures in polycrystalline materials.Ultrasonics, 149:107581, 2025. doi: 10.1016/j.ultras.2025.107581. 20

    work page doi:10.1016/j.ultras.2025.107581 2025

  26. [28]

    Veres, and Martin Ryzy

    Tomáš Grabec, István A. Veres, and Martin Ryzy. Surface acoustic wave attenuation in polycrystals: Numerical modeling using a statistical digital twin of an actual sample.Ultrasonics, 119:106585, 2022. doi: 10.1016/j.ultras.2021.106585

    work page doi:10.1016/j.ultras.2021.106585 2022

  27. [29]

    W. Y. Yeoh, B. Lan, and M. J. S. Lowe. Representative microstructures for two-dimensional computational studies of ultrasonic wave propagationintitaniumalloys.ProceedingsoftheRoyalSocietyA:Mathematical,PhysicalandEngineeringSciences,481(2323):20240776,

    work page

  28. [30]

    doi: 10.1098/rspa.2024.0776

    work page doi:10.1098/rspa.2024.0776 2024

  29. [31]

    Kalkowski, Ming Huang, Michael J.S

    Yuan Liu, Michał K. Kalkowski, Ming Huang, Michael J.S. Lowe, Vykintas Samaitis, Vaidotas Cic˙enas, and Andreas Schumm. Can ultrasound attenuation measurement be used to characterise grain statistics in castings?Ultrasonics, 115, 2021. ISSN 0041624X. doi: 10.1016/j.ultras.2021.106441

    work page doi:10.1016/j.ultras.2021.106441 2021

  30. [32]

    Rodríguez, Heriberto Granados Becerra, and Jin-Yeon Kim

    Alberto Ruiz, Vania M. Rodríguez, Heriberto Granados Becerra, and Jin-Yeon Kim. Ultrasonic characterization of microstructural changes due to static recrystallization and grain growth in inconel 625 and inconel 718 superalloys.Ultrasonics, 142:107383, 2024. doi: 10.1016/j.ultras.2024.107383

    work page doi:10.1016/j.ultras.2024.107383 2024

  31. [33]

    Ultrasonicbackscatteringmethodforcharacterizingthenon-uniformmicrostructure of polycrystals.International Journal of Mechanical Sciences, 306:110831, 2025

    BohanLiu,MingHuang,DehanZhang,andXudongYu. Ultrasonicbackscatteringmethodforcharacterizingthenon-uniformmicrostructure of polycrystals.International Journal of Mechanical Sciences, 306:110831, 2025. doi: 10.1016/j.ijmecsci.2025.110831

    work page doi:10.1016/j.ijmecsci.2025.110831 2025

  32. [34]

    Effectofmicrostructureandtexture gradient on the backscattered ultrasound amplitude of ti-6al-4v bars for aeroengine blade.Journal of Alloys and Compounds, 1010:177604,

    LeiLi,YangYing,MinZhou,ZuhanCao,DiziGuo,HaiyingYang,LiDing,FeixiaoHan,andJiVincent. Effectofmicrostructureandtexture gradient on the backscattered ultrasound amplitude of ti-6al-4v bars for aeroengine blade.Journal of Alloys and Compounds, 1010:177604,

    work page

  33. [35]

    doi: 10.1016/j.jallcom.2024.177604

    work page doi:10.1016/j.jallcom.2024.177604 2024

  34. [36]

    Characterizationofmicrotextureinti-6al-4vusingultrasonicpolar forward scattering.Ultrasonics, 165:108113, 2026

    IIIVela,Ramon,NathanialJ.Matz,WaledHassan,andJosephA.Turner. Characterizationofmicrotextureinti-6al-4vusingultrasonicpolar forward scattering.Ultrasonics, 165:108113, 2026. doi: 10.1016/j.ultras.2026.108113

    work page doi:10.1016/j.ultras.2026.108113 2026

  35. [37]

    Experimentalandcomputationalstudiesofultrasoundwavepropagationinhexagonalclose-packed polycrystals for texture detection.Acta Materialia, 63:107–122, 2014

    BLan,M.J.S.Lowe,andF.P.E.Dunne. Experimentalandcomputationalstudiesofultrasoundwavepropagationinhexagonalclose-packed polycrystals for texture detection.Acta Materialia, 63:107–122, 2014. ISSN 1359-6454. doi: 10.1016/j.actamat.2013.10.012

    work page doi:10.1016/j.actamat.2013.10.012 2014

  36. [38]

    Determiningthecrystallographic orientation of hexagonal crystal structure materials with surface acoustic wave velocity measurements.Ultrasonics, 108:106171, 2020

    PaulDryburgh,RichardJ.Smith,PaulMarrow,StevenJ.Lainé,SteveD.Sharples,MattClark,andWenqiLi. Determiningthecrystallographic orientation of hexagonal crystal structure materials with surface acoustic wave velocity measurements.Ultrasonics, 108:106171, 2020. doi: 10.1016/j.ultras.2020.106171

    work page doi:10.1016/j.ultras.2020.106171 2020

  37. [39]

    Characterizingpolycrystallinemicrostructuresbyreconstructinggrainboundaries and velocity map from ultrasonic surface wave field.Ultrasonics, 137:107175, 2024

    ZeqingSun,ShangziWu,AbhishekSaini,andZhengFan. Characterizingpolycrystallinemicrostructuresbyreconstructinggrainboundaries and velocity map from ultrasonic surface wave field.Ultrasonics, 137:107175, 2024. doi: 10.1016/j.ultras.2023.107175

    work page doi:10.1016/j.ultras.2023.107175 2024

  38. [40]

    Ben Britton, Tea-Sung Jun, Weimin Gan, Michael Hofmann, Fionn P.E

    Bo Lan, T. Ben Britton, Tea-Sung Jun, Weimin Gan, Michael Hofmann, Fionn P.E. Dunne, and Michael J.S. Lowe. Direct volumetric measurementofcrystallographictextureusingacousticwaves.ActaMaterialia,159:384–394,2018. ISSN13596454. doi:10.1016/j.actamat. 2018.08.037

    work page doi:10.1016/j.actamat 2018

  39. [41]

    Yu Liu, Xinyan Wang, J. P. Oliveira, Jingjing He, and Xuefei Guan. Spatial and directional characterization of wire and arc additive manufactured aluminum alloy using phased array ultrasonic backscattering method.Ultrasonics, 132:107024, 2023. doi: 10.1016/j.ultras. 2023.107024

    work page doi:10.1016/j.ultras 2023

  40. [42]

    Metallicmaterialmicrostructuregrainsizemeasurementsfrombackscatteredultrasonicarraydata using full matrix capture.NDT & E International, 149:103251, 2025

    WeiWang,JieZhang,andPaulD.Wilcox. Metallicmaterialmicrostructuregrainsizemeasurementsfrombackscatteredultrasonicarraydata using full matrix capture.NDT & E International, 149:103251, 2025. doi: 10.1016/j.ndteint.2024.103251

    work page doi:10.1016/j.ndteint.2024.103251 2025

  41. [43]

    Combined effect of frequency, grain size, and sound path on phased array ultrasonic spectrum and attenuation.Ultrasonics, 149:107593, 2025

    Yu Liu, Qiang Tian, Shuzeng Zhang, and Xuefei Guan. Combined effect of frequency, grain size, and sound path on phased array ultrasonic spectrum and attenuation.Ultrasonics, 149:107593, 2025. doi: 10.1016/j.ultras.2025.107593

    work page doi:10.1016/j.ultras.2025.107593 2025

  42. [44]

    Rapid measurement of volumetric texture using resonant ultrasound spectroscopy.Scripta Materialia, 157:44–48, 2018

    Bo Lan, Michael A Carpenter, Weimin Gan, Michael Hofmann, Fionn P E Dunne, and Michael J S Lowe. Rapid measurement of volumetric texture using resonant ultrasound spectroscopy.Scripta Materialia, 157:44–48, 2018. ISSN 1359-6462. doi: 10.1016/j.scriptamat.2018.07

    work page doi:10.1016/j.scriptamat.2018.07 2018

  43. [45]

    URLhttps://doi.org/10.1016/j.scriptamat.2018.07.029

    work page doi:10.1016/j.scriptamat.2018.07.029 2018

  44. [46]

    Howdoesgrazingincidenceultrasonicmicroscopy work? a study based on grain-scale numerical simulations.Ultrasonics, 114:106387, 2021

    MichałK.Kalkowski,MichaelJ.S.Lowe,MartinBarth,MarekRjelka,andBerndKöhler. Howdoesgrazingincidenceultrasonicmicroscopy work? a study based on grain-scale numerical simulations.Ultrasonics, 114:106387, 2021. doi: 10.1016/j.ultras.2021.106387

    work page doi:10.1016/j.ultras.2021.106387 2021

  45. [47]

    Average grain size evaluation using scattering-induced attenuation of coda waves.Ultrasonics, 141:107334, 2024

    Jingjing He, Chenjun Gao, Xun Wang, Jinsong Yang, Qiang Tian, and Xuefei Guan. Average grain size evaluation using scattering-induced attenuation of coda waves.Ultrasonics, 141:107334, 2024. doi: 10.1016/j.ultras.2024.107334

    work page doi:10.1016/j.ultras.2024.107334 2024

  46. [48]

    Identifyinggrainsizeinastma36steelusingultrasonicbackscatteredsignalsand machine learning.NDT & E International, 2024

    M.C.A.Viana,P.Pereira,A.A.Buenos,andA.A.Santos. Identifyinggrainsizeinastma36steelusingultrasonicbackscatteredsignalsand machine learning.NDT & E International, 2024. In press

    work page 2024

  47. [49]

    Autonomouscharacterizationofgrainsizedistributionusingnonlinearlambwaves based on deep learning.The Journal of the Acoustical Society of America, 152(3):1913–1921, 2022

    LishuaiLiu,PengWu,YanxunXiang,andFu-ZhenXuan. Autonomouscharacterizationofgrainsizedistributionusingnonlinearlambwaves based on deep learning.The Journal of the Acoustical Society of America, 152(3):1913–1921, 2022. doi: 10.1121/10.0014289

    work page doi:10.1121/10.0014289 1913

  48. [50]

    Deep learning based inversion of locally anisotropic weld properties from ultrasonic array data.Applied Sciences, 12(2):532, 2022

    Jonathan Singh, Katherine Tant, Anthony Mulholland, and Charles MacLeod. Deep learning based inversion of locally anisotropic weld properties from ultrasonic array data.Applied Sciences, 12(2):532, 2022. doi: 10.3390/app12020532

    work page doi:10.3390/app12020532 2022

  49. [51]

    Developingneuralnetworkstorapidlymapcrystallographicorientationusinglaser ultrasound measurements.Scripta Materialia, 256:116415, 2025

    RikeshPatel,WenqiLi,RichardJ.Smith,andMattClark. Developingneuralnetworkstorapidlymapcrystallographicorientationusinglaser ultrasound measurements.Scripta Materialia, 256:116415, 2025. doi: 10.1016/j.scriptamat.2024.116415

    work page doi:10.1016/j.scriptamat.2024.116415 2025

  50. [52]

    KhemrajShukla,AmeyaD.Jagtap,JamesL.Blackshire,DanielSparkman,andGeorgeEmKarniadakis. Aphysics-informedneuralnetwork for quantifying the microstructural properties of polycrystalline nickel using ultrasound data: A promising approach for solving inverse problems.IEEE Signal Processing Magazine, 39(1):68–77, 2022. doi: 10.1109/MSP.2021.3118904

    work page doi:10.1109/msp.2021.3118904 2022

  51. [53]

    S. I. Rokhlin, G. Sha, J. Li, and A. L. Pilchak. Inversion methodology for ultrasonic characterization of polycrystals with clusters of preferentially oriented grains.Ultrasonics, 115(March), 2021. ISSN 0041624X. doi: 10.1016/j.ultras.2021.106433

    work page doi:10.1016/j.ultras.2021.106433 2021

  52. [54]

    Decouplinganalysisofultrasonicscatteringcharacteristicsinporouspolycrystallinematerials using phase field and finite element methods.Ultrasonics, 159:107864, 2026

    ZixinGuo,YongfengSong,andXiongbingLi. Decouplinganalysisofultrasonicscatteringcharacteristicsinporouspolycrystallinematerials using phase field and finite element methods.Ultrasonics, 159:107864, 2026. doi: 10.1016/j.ultras.2025.107864

    work page doi:10.1016/j.ultras.2025.107864 2026

  53. [55]

    Bellman and K

    R. Bellman and K. J. Åström. On structural identifiability.Mathematical Biosciences, 7(3-4):329–339, 1970. ISSN 0025-5564. doi: 10.1016/0025-5564(70)90132-X

    work page doi:10.1016/0025-5564(70)90132-x 1970

  54. [56]

    Cobelli and J

    C. Cobelli and J. J. DiStefano, III. Parameter and structural identifiability concepts and ambiguities: a critical review and analysis.American Journal of Physiology-Regulatory, Integrative and Comparative Physiology, 239(1):R7–R24, 1980. doi: 10.1152/ajpregu.1980.239.1.R7

    work page doi:10.1152/ajpregu.1980.239.1.r7 1980

  55. [57]

    Joseph H. A. Guillaume, John D. Jakeman, Stefano Marsili-Libelli, Michael Asher, Philip Brunner, Barry Croke, Mary C. Hill, Anthony J. Jakeman, Karel J. Keesman, Saman Razavi, and Johannes D. Stigter. Introductory overview of identifiability analysis: A guide to evaluating 21 whether you have the right type of data for your modeling purpose.Environmental ...

    work page 2019

  56. [58]

    doi: 10.1016/j.envsoft.2019.07.007

    work page doi:10.1016/j.envsoft.2019.07.007 2019

  57. [59]

    Rose.Reflection and Refraction

    Joseph L. Rose.Reflection and Refraction. Cambridge University Press, 2014

    work page 2014

  58. [60]

    R. M. Alford, K. R. Kelly, and D. M. Whitmore. Accuracy of finite-difference modeling of the acoustic wave equation.Geophysics, 39(6): 834–842, 1974. doi: 10.1190/1.1440470

    work page doi:10.1190/1.1440470 1974

  59. [61]

    Phd thesis, Curtin University of Technology, May 2003

    Ahmad Zakaria.Numerical Modelling of Wave Propagation Using Higher Order Finite-Difference Formulas. Phd thesis, Curtin University of Technology, May 2003. URLhttps://espace.curtin.edu.au/handle/20.500.11937/190

    work page 2003

  60. [62]

    Numerical time-domain modeling of linear and nonlinear ultrasonic wave propagation using finite integration tech- niques––theory and applications.Ultrasonics, 42(1-9):221–229, 2004

    Frank Schubert. Numerical time-domain modeling of linear and nonlinear ultrasonic wave propagation using finite integration tech- niques––theory and applications.Ultrasonics, 42(1-9):221–229, 2004. ISSN 0041-624X. doi: 10.1016/j.ultras.2004.01.013

    work page doi:10.1016/j.ultras.2004.01.013 2004

  61. [63]

    Anstett-Collin, L

    F. Anstett-Collin, L. Denis-Vidal, and G. Millérioux. A priori identifiability: An overview on definitions and approaches.Annual Reviews in Control, 50:139–149, 2020. ISSN 1367-5788. doi: 10.1016/j.arcontrol.2020.10.006

    work page doi:10.1016/j.arcontrol.2020.10.006 2020

  62. [64]

    Kube and Joseph A

    Christopher M. Kube and Joseph A. Turner. Voigt, Reuss, Hill, and Self-Consistent Techniques for Modeling Ultrasonic Scattering. In41st Annual Review of Progress in Quantitative Nondestructive Evaluation, volume 1650, pages 926–934, 2015. doi: 10.1063/1.4914698

    work page doi:10.1063/1.4914698 2015

  63. [65]

    Advancesingaussianrandomfieldgeneration:Areview.IEEEAccess,7:153123–153139,2019

    YangLiu,JingfaLi,ShuyuSun,andBoYu. Advancesingaussianrandomfieldgeneration:Areview.IEEEAccess,7:153123–153139,2019. doi: 10.1109/ACCESS.2019.2948616

    work page doi:10.1109/access.2019.2948616 2019

  64. [66]

    Bo Lan, Michael J. S. Lowe, and Fionn P. E. Dunne. A spherical harmonic approach for the determination of HCP texture from ultrasound: A solution to the inverse problem.Journal of the Mechanics and Physics of Solids, 83:179–198, 2015. ISSN 00225096. doi: 10.1016/j.jmps.2015.06.014. URLhttp://dx.doi.org/10.1016/j.jmps.2015.06.014

    work page doi:10.1016/j.jmps.2015.06.014 2015

  65. [67]

    Accelerated finite element elastodynamic simulations using the GPU.Journal of Computational Physics, 257:687–707,

    Peter Huthwaite. Accelerated finite element elastodynamic simulations using the GPU.Journal of Computational Physics, 257:687–707,

    work page

  66. [68]

    doi: 10.1016/j.jcp.2013.10.017

    ISSN 00219991. doi: 10.1016/j.jcp.2013.10.017

    work page doi:10.1016/j.jcp.2013.10.017 2013

  67. [69]

    Huang, G

    M. Huang, G. Sha, P. Huthwaite, S. I. Rokhlin, and M. J. S. Lowe. Maximizing the accuracy of finite element simulation of elastic wave propagation in polycrystals.The Journal of the Acoustical Society of America, 148:1890–1910, 2020. ISSN 0001-4966. doi: 10.1121/10.0002102

    work page doi:10.1121/10.0002102 1910

  68. [70]

    Brett, Peter Huthwaite, and Michael J

    Anton Van Pamel, Colin R. Brett, Peter Huthwaite, and Michael J. S. Lowe. Finite element modelling of elastic wave scattering within a polycrystalline material in two and three dimensions.The Journal of the Acoustical Society of America, 138(4):2326–2336, 2015. ISSN 0001-4966. doi: 10.1121/1.4931445

    work page doi:10.1121/1.4931445 2015

  69. [71]

    Zuo, and Xiaodong Wang

    Yonghong Zhang, Yu Wang, Ming J. Zuo, and Xiaodong Wang. Ultrasonic time-of-flight diffraction crack size identification based on cross-correlation.Canadian Conference on Electrical and Computer Engineering, pages 1797–1800, 2008. ISSN 08407789. doi: 10.1109/CCECE.2008.4564854

    work page doi:10.1109/ccece.2008.4564854 2008

  70. [72]

    A generalized spherical harmonic deconvolution to obtain texture of cubic materials from ultrasonicwavespeed.JournaloftheMechanicsandPhysicsofSolids,83:221–242,2015

    Bo Lan, Michael J S Lowe, and Fionn P E Dunne. A generalized spherical harmonic deconvolution to obtain texture of cubic materials from ultrasonicwavespeed.JournaloftheMechanicsandPhysicsofSolids,83:221–242,2015. ISSN0022-5096. doi:10.1016/j.jmps.2015.06.012. URLhttp://dx.doi.org/10.1016/j.jmps.2015.06.012

    work page doi:10.1016/j.jmps.2015.06.012 2015

  71. [73]

    Cambridge University Press, 2017

    Zhilin Li, Zhonghua Qiao, and Tao Tang.Numerical Solution of Differential Equations: Introduction to Finite Difference and Finite Element Methods. Cambridge University Press, 2017. doi: 10.1017/9781316671528

    work page doi:10.1017/9781316671528 2017

  72. [74]

    Practicalidentifiabilityofparametrisedmodels:Areviewofbenefitsandlimitationsof various approaches.Mathematics and Computers in Simulation, 199:202–216, 2022

    NicholasN.Lam,PaulD.Docherty,andRuaMurray. Practicalidentifiabilityofparametrisedmodels:Areviewofbenefitsandlimitationsof various approaches.Mathematics and Computers in Simulation, 199:202–216, 2022. ISSN 0378-4754. doi: 10.1016/j.matcom.2022.03.020

    work page doi:10.1016/j.matcom.2022.03.020 2022

  73. [75]

    Ultrasonic attenuation of polycrystalline materials with a distribution of grain sizes.The Journal of the Acoustical Society of America, 141(6):4347 – 4353, 2017

    Andrea P Arguelles and Joseph A Turner. Ultrasonic attenuation of polycrystalline materials with a distribution of grain sizes.The Journal of the Acoustical Society of America, 141(6):4347 – 4353, 2017. doi: 10.1121/1.4984290

    work page doi:10.1121/1.4984290 2017

  74. [76]

    Nagy, and Peter Cawley

    Yuan Liu, Anton Van Pamel, Peter B. Nagy, and Peter Cawley. Investigation of ultrasonic backscatter using three-dimensional finite element simulations.The Journal of the Acoustical Society of America, 145(3):1584–1595, 2019. ISSN 0001-4966. doi: 10.1121/1.5094783

    work page doi:10.1121/1.5094783 2019

  75. [77]

    Xun Huan and Youssef M. Marzouk. Optimal experimental design: Formulations and computations.Acta Numerica, 33:1–140, 2024. doi: 10.1017/S0962492924000013

    work page doi:10.1017/s0962492924000013 2024

  76. [78]

    Drinkwater, and Paul D

    Caroline Holmes, Bruce W. Drinkwater, and Paul D. Wilcox. Post-processing of the full matrix of ultrasonic transmit-receive array data for non-destructive evaluation.NDT and E International, 38(8):701–711, 2005. ISSN 09638695. doi: 10.1016/j.ndteint.2005.04.002

    work page doi:10.1016/j.ndteint.2005.04.002 2005

  77. [79]

    Wilcox, Caroline Holmes, and Bruce W

    Paul D. Wilcox, Caroline Holmes, and Bruce W. Drinkwater. Advanced reflector characterization with ultrasonic phased arrays in nde applications.IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, 54(8):1541–1550, 2007. doi: 10.1109/TUFFC. 2007.424

    work page doi:10.1109/tuffc 2007

  78. [80]

    Unsupervisedlearningfromincompletemeasurementsforinverseproblems

    JulianTachella,SamuelHurault,andMikeDavies. Unsupervisedlearningfromincompletemeasurementsforinverseproblems. InAdvances in Neural Information Processing Systems, volume 35, pages 4981–4993, 2022

    work page 2022

  79. [81]

    doi: 10.1017/S0962492919000059

    SimonArridge,PeterMaass,OzanÖktem,andCarola-BibianeSchönlieb.Solvinginverseproblemsusingdata-drivenmodels.ActaNumerica, 28:1–174, 2019. doi: 10.1017/S0962492919000059

    work page doi:10.1017/s0962492919000059 2019

  80. [82]

    Brett, and Michael J.S

    Anton Van Pamel, Peter Huthwaite, Colin R. Brett, and Michael J.S. Lowe. Numerical simulations of ultrasonic array imaging of highly scattering materials.NDT and E International, 81:9–19, 2016. ISSN 09638695. doi: 10.1016/j.ndteint.2016.02.004. URLhttp: //dx.doi.org/10.1016/j.ndteint.2016.02.004. 22

    work page doi:10.1016/j.ndteint.2016.02.004 2016