pith. sign in

arxiv: 2408.06765 · v1 · submitted 2024-08-13 · ⚛️ physics.class-ph · physics.comp-ph

Efficient numerical frameworks for modelling ultrasonic beams propagating across interfaces

Andr\'e Lello de Almeida , Melody Png , Bo Lan This is my paper

Pith reviewed 2026-05-23 22:27 UTC · model grok-4.3

classification ⚛️ physics.class-ph physics.comp-ph
keywords ultrasonic beam modelingwave propagation across interfacesRayleigh-Sommerfeld Integralray tracingQuasi-Monte Carlo integrationnumerical efficiencynon-destructive testingmulti-layer simulation
0
0 comments X

The pith

Ray tracing model is more efficient than RSI-based model for ultrasonic fields through many interfaces

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

The paper develops two numerical frameworks for computing ultrasonic wave fields from a transducer propagating through one or more interfaces, both evaluated using Quasi-Monte Carlo integration. The first extends the Rayleigh-Sommerfeld Integral while the second applies a high-frequency ray-tracing approximation. Direct comparisons on multiple use cases show that the RSI approach scales better when many field points must be computed, such as a full image, whereas the ray-tracing approach scales better when the number of interfaces is large, such as thin-layered media, because it skips intermediate field propagations between layers. The distinction matters for practical simulations where computational cost depends on whether the task requires dense sampling or passage through many boundaries.

Core claim

The authors formulate an improved RSI-based model and a ray-tracing high-frequency model for transducer-generated ultrasonic fields crossing interfaces. They demonstrate that the RSI-based model performs well when a large number of field points is required, while the ray-tracing formulation is most efficient for a large number of interfaces because it is unnecessary to propagate the field between all the interfaces first; this advantage is especially clear when the field needs evaluation at only a few points beyond multiple interfaces.

What carries the argument

Two alternative numerical frameworks (extended Rayleigh-Sommerfeld Integral and high-frequency ray tracing) evaluated by Quasi-Monte Carlo integration

If this is right

  • RSI-based model is the practical choice when a full image of the field must be computed
  • Ray-tracing formulation becomes preferable once the number of interfaces grows large
  • Efficiency advantage of ray tracing appears most clearly when only a few observation points lie beyond the final interface
  • Both models can treat multiple interfaces but differ sharply in how cost scales with interface count versus point count

Where Pith is reading between the lines

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

  • The efficiency comparison supplies a simple rule for selecting the model according to whether interface count or field-point count dominates the problem size
  • The same scaling distinction could be tested in three-dimensional geometries or with curved interfaces to see if the crossover point shifts
  • Hybrid schemes that switch between the two formulations depending on local interface density become a natural next step

Load-bearing premise

The high-frequency approximation underlying the ray-tracing model remains accurate for the ultrasonic frequencies and interface configurations considered, and skipping intermediate field propagation between interfaces does not introduce significant errors.

What would settle it

A side-by-side comparison of measured ultrasonic pressure at a few points after a stack of ten or more thin layers against the ray-tracing predictions that omit all intermediate propagations would show whether the approximation holds.

Figures

Figures reproduced from arXiv: 2408.06765 by Andr\'e Lello de Almeida, Bo Lan, Melody Png.

Figure 1
Figure 1. Figure 1: Rayleigh-Sommerfeld diffraction by a semi-transparent plane interface. The total field at [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Modeling of an acoustic field through an interface using the RSI formulation. The total [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Representation of the geometry used for the system in each particular study case: oblique [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Wave field generated by a transducer at oblique incidence and transmission through one [PITH_FULL_IMAGE:figures/full_fig_p011_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Detailed representation of the geometry used in the FEM (a), and plot of the wave field [PITH_FULL_IMAGE:figures/full_fig_p012_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Wave field generated by a focused transducer at normal incidence and transmission [PITH_FULL_IMAGE:figures/full_fig_p014_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Wave field generated by a transducer at oblique incidence and transmission through [PITH_FULL_IMAGE:figures/full_fig_p015_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Computation time as a function of the number of field points for several different sce [PITH_FULL_IMAGE:figures/full_fig_p016_8.png] view at source ↗
read the original abstract

Two different frameworks are developed to model the wave field generated by a transducer and propagating through one or more interfaces, and a Quasi-Monte Carlo (QMC) integration scheme is used to numerically evaluate their results. The first method is based on the Rayleigh-Sommerfeld Integral (RSI), further developing a formulation in the literature and improving its capabilities, while the second relies on a high-frequency approximation, using a ray tracing principle. The advantages and limitations of each model are then compared via in-depth investigations on several use cases, culminating in an efficiency and scope assessment. It was found that the RSI-based model performs well if a large number of field points is needed, such as when modelling a full image of the field. Conversely, for a large number of interfaces, such as when modelling the field through a thin-layered material, the most efficient model was the ray tracing formulation, since it was unnecessary to propagate the field between all the interfaces first. This was especially noticeable for applications requiring only the evaluation of the field at a few points on the other side of multiple interfaces.

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 develops two numerical frameworks for modeling ultrasonic wave fields from a transducer propagating across one or more interfaces, both evaluated via Quasi-Monte Carlo integration. The first extends the Rayleigh-Sommerfeld Integral (RSI); the second employs a high-frequency ray-tracing approximation. Comparative investigations on several use cases yield an efficiency and scope assessment: the RSI model is found preferable when many field points are required (e.g., full-field imaging), while the ray-tracing model is more efficient for many interfaces (e.g., thin-layered media) because intermediate field propagation between interfaces can be skipped.

Significance. If the efficiency rankings and underlying accuracy assumptions are substantiated by quantitative benchmarks, the work could supply practical guidance for method selection in ultrasonic NDT and imaging of layered structures. The improved RSI formulation and QMC integration scheme represent incremental methodological advances, but the absence of validation data currently limits the contribution.

major comments (2)
  1. [Abstract] Abstract (final paragraph): The central claim that 'the most efficient model was the ray tracing formulation, since it was unnecessary to propagate the field between all the interfaces first' is presented without any reported computation times, error metrics, or direct RSI-versus-ray-tracing comparisons at evaluation points behind multiple thin layers. This quantitative gap is load-bearing for the efficiency and scope assessment.
  2. [Abstract] Abstract (final paragraph): The high-frequency approximation underlying the ray-tracing model is asserted to remain accurate for the ultrasonic frequencies and interface configurations considered, yet no verification against the RSI benchmark, frequency/layer-thickness regime analysis, or error bounds on the skipped propagation steps is supplied. This assumption directly supports the ranking of models for large numbers of interfaces.
minor comments (1)
  1. [Abstract] The abstract would benefit from explicit mention of the specific numbers of interfaces, field points, and frequencies used in the 'in-depth investigations' to allow readers to assess the scope of the reported comparisons.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments. We address each major comment below and outline the revisions that will be made to strengthen the quantitative support and validation in the manuscript.

read point-by-point responses

  1. Referee: [Abstract] Abstract (final paragraph): The central claim that 'the most efficient model was the ray tracing formulation, since it was unnecessary to propagate the field between all the interfaces first' is presented without any reported computation times, error metrics, or direct RSI-versus-ray-tracing comparisons at evaluation points behind multiple thin layers. This quantitative gap is load-bearing for the efficiency and scope assessment.

    Authors: We agree that the abstract would benefit from explicit quantitative backing for the efficiency claim. The manuscript body (Sections 4 and 5) already contains comparative investigations on multiple use cases, including timing data and direct RSI-versus-ray-tracing results for scenarios with several interfaces. To resolve the gap identified, we will revise the abstract to include key metrics (e.g., relative computation times and error values) drawn from those sections, with particular emphasis on evaluation points behind multiple thin layers. revision: yes

  2. Referee: [Abstract] Abstract (final paragraph): The high-frequency approximation underlying the ray-tracing model is asserted to remain accurate for the ultrasonic frequencies and interface configurations considered, yet no verification against the RSI benchmark, frequency/layer-thickness regime analysis, or error bounds on the skipped propagation steps is supplied. This assumption directly supports the ranking of models for large numbers of interfaces.

    Authors: The ray-tracing model employs the standard high-frequency approximation, which is cross-checked against the RSI reference in the single-interface cases presented. We acknowledge, however, that explicit verification, regime analysis, and error bounds for the multi-interface skipped-propagation steps are not currently supplied. We will add a dedicated subsection that performs direct RSI-versus-ray-tracing comparisons on a representative multi-layer test case, includes a wavelength-to-layer-thickness regime analysis, and reports quantitative error bounds on the approximation. revision: yes

Circularity Check

0 steps flagged

No circularity; standard numerical frameworks compared via direct computation

full rationale

The paper constructs RSI and ray-tracing models from established wave-propagation principles (Rayleigh-Sommerfeld integral and high-frequency ray approximation), then evaluates both with QMC integration on concrete use cases. Efficiency rankings are obtained from explicit runtime and accuracy measurements across varying numbers of field points and interfaces; these outcomes are not forced by any self-definition, fitted parameter renamed as prediction, or self-citation chain. The abstract and described investigations treat the two formulations as independent implementations whose relative performance is measured externally, satisfying the self-contained criterion.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Abstract-only review; no explicit free parameters, invented entities, or ad-hoc axioms stated. Relies on standard domain assumptions of linear acoustics and validity of the cited integral and ray approximations.

axioms (1)
  • domain assumption Ultrasonic wave propagation through interfaces can be modeled using the Rayleigh-Sommerfeld Integral or high-frequency ray tracing approximations.
    Central to both frameworks described in the abstract.

pith-pipeline@v0.9.0 · 5719 in / 1248 out tokens · 27857 ms · 2026-05-23T22:27:37.200232+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

44 extracted references · 44 canonical work pages

  1. [1]

    Schmerr, L. W. (2016) Fundamentals of Ultrasonic Nondestructive Evaluation: A Modeling Approach. 2nd ed. New York, Springer

    work page 2016

  2. [2]

    & Alnuaimi, N

    Banerjee, S., Kundu, T. & Alnuaimi, N. A. (2007) DPSM technique for ultrasonic field modelling near fluid-solid interface. Ultrasonics. 46, 235-250. doi:10.1016/j.ultras.2007.02.003

    work page doi:10.1016/j.ultras.2007.02.003 2007

  3. [3]

    & Jeong, H

    Zhang, S., Huang, Y., Li, X. & Jeong, H. (2021) Modeling of wave fields generated by ul- trasonic transducers using a quasi-Monte Carlo method. J. Acoust. Soc. Am . 149 (1), 7-15. doi:10.1121/10.0002972

    work page doi:10.1121/10.0002972 2021

  4. [4]

    Williams Jr., A. O. (1951) The piston source at high frequencies. J. Acoust. Soc. Am . 23, 1-6. doi:10.1121/1.1906722

    work page doi:10.1121/1.1906722 1951

  5. [5]

    Rogers, P. H. & Van Buren, A. L. (1974) An exact expression for the Lommel diffraction cor- rection integral. J. Acoust. Soc. Am. 55 (4), 724-728. doi:10.1121/1.1914589. Review of Progress in Quantitative Nondestructive Evaluation . Boston, Springer. 19

    work page doi:10.1121/1.1914589 1974

  6. [6]

    W., Lerch, T

    Schmerr, L. W., Lerch, T. P. & Sedov, A. (1995) Modeling the Radiation of Focused and Unfocused Ultrasonic Transducers Through Planar Interfaces. In: Thompson, D. O. & Chimenti, D. E. (eds.). New York, Plenum Press

    work page 1995

  7. [7]

    P., Schmerr, L

    Lerch, T. P., Schmerr, L. W. & Sedov, A. (1995) The Paraxial Approximation for Radiation of a Planar Ultrasonic Transducer at Oblique Incidence Through an Interface. In: Thompson, D. O. & Chimenti, D. E. (eds.) Review of Progress in Quantitative Nondestructive Evaluation . Boston, Springer

    work page 1995

  8. [8]

    Wen, J. J. & Breazeale, M. A. (1988) A diffraction beam field expressed as the superposition of Gaussian beams. J. Acoust. Soc. Am. 83 (5), 1752-1756. doi:10.1121/1.396508

    work page doi:10.1121/1.396508 1988

  9. [9]

    Kim, H.-J., Schmerr, L. W. & Sedov, A. (2006) Generation of the basis sets for multi- Gaussian ultrasonic beam models - An overview. J. Acoust. Soc. Am . 119 (4), 1971-1978. doi:10.1121/1.2169921

    work page doi:10.1121/1.2169921 2006

  10. [10]

    & Wolf, E

    Born, M. & Wolf, E. (1980) Principles of Optics . 6th ed. Oxford, Pergamon Press Ltd

    work page 1980

  11. [11]

    Goodman, J. W. (2017) Introduction to Fourier Optics . 4th ed. New York, W. H. Freeman

    work page 2017

  12. [12]

    & Ludwig, R

    Lord, W., You, Z., Lusk, M. & Ludwig, R. (1988) Numerical predictions of surface wave phenomena for ultrasonic NDE. In IEEE 1988 Ultrasonics Symposium Proceedings . 1065-1068. doi:10.1109/ULTSYM.1988.49541

    work page doi:10.1109/ultsym.1988.49541 1988

  13. [13]

    L., Vaughan, D

    Wojcik, G. L., Vaughan, D. K., Abboud, N. & Mould, J. (1993) Electromechanical modeling using explicit time-domain finite elements. In IEEE 1993 Ultrasonics Symposium Proceedings . 1107-1112. doi:10.1109/ULTSYM.1993.339594

    work page doi:10.1109/ultsym.1993.339594 1993

  14. [14]

    (2014) Finite Element Procedures

    Bathe, K.-J. (2014) Finite Element Procedures. 2nd ed. Watertown, MA, Klaus-J¨ urgen Bathe

    work page 2014

  15. [15]

    & Karal, Jr., F

    Alterman, Z. & Karal, Jr., F. C. (1968) Propagation of elastic waves in layered media by finite difference methods. Bull. Seism. Soc. Am . 58, 367-398. doi:10.1785/BSSA0580010367

    work page doi:10.1785/bssa0580010367 1968

  16. [16]

    Scandrett, C. L. & Achenbach, J. D. (1987) Time-domain finite difference calculations for interaction of an ultrasonic wave with a surface-breaking crack. Wave Motion . 9 (2), 171-190. doi:10.1016/0165-2125(87)90051-5

    work page doi:10.1016/0165-2125(87)90051-5 1987

  17. [17]

    & Bouaoua, N

    Alia, A., Djelouah, H. & Bouaoua, N. (2004) Finite difference modeling of the ul- trasonic field radiated by circular transducers. J. Comput. Acoust . 12 (4), 475-499. doi:10.1142/S0218396X04002365

    work page doi:10.1142/s0218396x04002365 2004

  18. [18]

    & Kitahara, M

    Niwa, Y., Hirose, S. & Kitahara, M. (1986) Application of the boundary integral equation (BIE) method to transient response analysis of inclusions in a half space. Wave Motion . 8 (1), 77-91. doi:10.1016/0165-2125(86)90007-7

    work page doi:10.1016/0165-2125(86)90007-7 1986

  19. [19]

    D., Andrew, G., P’an., C., Grills, B., Mullis, R

    Aldrin, J., Achenbach, J. D., Andrew, G., P’an., C., Grills, B., Mullis, R. T., Spencer, F. W. & Golis, M. (2001) Case study for the implementation of an automated ultrasonic technique to detect fatigue cracks in aircraft weep holes. Materials evaluation. 59 (11), 1313-1319

    work page 2001

  20. [20]

    & Kundu, T

    Placko, D. & Kundu, T. (2001) A theoretical study of magnetic and ultrasonic sensors: depen- dence of magnetic potential and acoustic pressure on the sensor geometry. Advanced NDE for structural and biological health monitoring. In Proceedings of SPIE, SPIE’s 6th Annual Interna- tional Symposium on NDE for Health Monitoring and Diagnostics . 52-62. doi:10...

    work page doi:10.1117/12.434201 2001

  21. [21]

    & Talvard, M

    Lh´ emery, A., Paradis, L., Rizzo, L. & Talvard, M. (1999) Multiple-technique NDT simulations of realistic configurations at the French Atomic Energy Comission (CEA). Rev. Prog. Quant. Nondestr. Eval. Vol. 18, 671-678. doi:10.1049/ic:19990107

    work page doi:10.1049/ic:19990107 1999

  22. [22]

    J., Mukherjee, S., Nishimura, N., Schanz, M., Ye, W., Sutradhar, A., Pan, E., Dumont, N

    Liu, Y. J., Mukherjee, S., Nishimura, N., Schanz, M., Ye, W., Sutradhar, A., Pan, E., Dumont, N. A., Frangi, A. & Saez, A. (2011) Recent advances and emerging applications of the boundary element method. Appl. Mech. Rev. 64 (3), 030802. doi:10.1115/1.4005491

    work page doi:10.1115/1.4005491 2011

  23. [23]

    (2014) Accelerated finite element elastodynamic simulations using the GPU

    Huthwaite, P. (2014) Accelerated finite element elastodynamic simulations using the GPU. J. Comput. Phys. A . 257, 687-707. doi:10.1016/j.jcp.2013.10.017

    work page doi:10.1016/j.jcp.2013.10.017 2014

  24. [24]

    (2022) Modeling ultrasonic wave fields us- ing a Quasi-Monte Carlo method: Wave transmission through complicated interfaces

    Zhang, S., Cheng, C., Li, X., Huang, Y., & Jeong, H. (2022) Modeling ultrasonic wave fields us- ing a Quasi-Monte Carlo method: Wave transmission through complicated interfaces. J. Acoust. Soc. Am. 152 (2), 994-1002. doi:10.1121/10.0013411

    work page doi:10.1121/10.0013411 2022

  25. [25]

    Rayleigh, J. W. S. (1945) The Theory of Sound , Vol. II. New York, Dover

    work page 1945

  26. [26]

    King, L. V. (1934) Can. J. Res . 11, 135-146

    work page 1934

  27. [27]

    (1941) Akust

    Schoch, A. (1941) Akust. Z. 6, 318-326

    work page 1941

  28. [28]

    Harris, G. R. (1981) Review of transient field theory for a baffled planar piston. J. Acoust. Soc. Am. 70 (1), 10-20. doi:10.1121/1.386687

    work page doi:10.1121/1.386687 1981

  29. [29]

    Nayfeh, A. H. & Chimenti, D. E. (1984) Reflection of finite acoustic beams from loaded and stiffened half-spaces. J. Acoust. Soc. Am. 75 (5), 1360-1368. doi:10.1121/1.390857

    work page doi:10.1121/1.390857 1984

  30. [30]

    Pialucha, T. P. (1992) The reflection coefficient from interface layers in NDT of adhesive joints. PhD thesis. Imperial College London

    work page 1992

  31. [31]

    (1917) Ann

    Rubinowicz, A. (1917) Ann. Physik. 53, 257

    work page 1917

  32. [32]

    Marchand, E. W. & Wolf, E. (1962) Boundary diffraction wave in the domain of the Rayleigh- Kirchhoff diffraction theory. J. Opt. Soc. Am . 52 (7), 761-767. doi:10.1364/JOSA.52.000761

    work page doi:10.1364/josa.52.000761 1962

  33. [33]

    Miller, D. A. B. (1991) Huygens’s wave propagation principle corrected. Opt. Lett. 16 (18), 1370-1372. doi:10.1364/OL.16.001370

    work page doi:10.1364/ol.16.001370 1991

  34. [34]

    & Truell, R

    Seki, H., Granato, A. & Truell, R. (1956) Diffraction Effects in the Ultrasonic Field of a Piston Source and Their Importance in the Accurate Measurement of Attenuation. J. Acoust. Soc. Am. 28 (2), 230-238. doi:10.1121/1.1908249

    work page doi:10.1121/1.1908249 1956

  35. [35]

    & Peyton, A

    Weston, M., Mudge, P., Davis, C. & Peyton, A. (2012) Time efficient auto-focussing algorithms for ultrasonic inspection of dual-layered media using Full Matrix Capture.NDT&E International. 47, 43-50. doi:10.1016/j.ndteint.2011.10.006

    work page doi:10.1016/j.ndteint.2011.10.006 2012

  36. [36]

    Morokoff, W. J. & Caflisch, R. E. (1995) Quasi-Monte Carlo Integration. J. Comput. Phys . 122, 218-230. doi:10.1006/jcph.1995.1209

    work page doi:10.1006/jcph.1995.1209 1995

  37. [37]

    O’Neill, B. E. & Maev, R. G. (1998) Integral approximation method for calculat- ing ultrasonic beam propagation in anisotropic materials. Phys. Rev. B . 58, 5479-5485. doi:10.1103/PhysRevB.58.5479. 21

    work page doi:10.1103/physrevb.58.5479 1998

  38. [38]

    Every, A. G. & Amulele, G. M. (2002) Angular spectrum method and ray tracing algorithm for the acoustic field of a focusing transducer in an anisotropic solid. IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control. 49 (3), 307-318. doi:10.1109/58.990942

    work page doi:10.1109/58.990942 2002

  39. [39]

    & Achenbach, J

    Wang, C.-Y. & Achenbach, J. D. (1995) Three-dimensional time-harmonic elastody- namic Green’s functions for anisotropic solids. Proc. R. Soc. Lond. A . 449, 441-458. doi:10.1098/rspa.1995.0052

    work page doi:10.1098/rspa.1995.0052 1995

  40. [40]

    & Rose, J

    Tverdokhlebov, A. & Rose, J. (1988) On Green’s functions for elastic waves in anisotropic media. J. Acoust. Soc. Am. 83, 118-121. doi:10.1121/1.396437

    work page doi:10.1121/1.396437 1988

  41. [41]

    & Tverdokhlebov, A

    Rose, J., Balasubramaniam, K. & Tverdokhlebov, A. (1989) A numerical integration Green’s function model for ultrasonic field profiles in mildly anisotropic media. J. Nondestruct. Eval . 8, 165-179. doi:10.1007/BF00570885

    work page doi:10.1007/bf00570885 1989

  42. [42]

    & Reddyhoff, T

    Jia, Y., Dou, P., Zheng, P., Wu, T., Yang, P., Yu, M. & Reddyhoff, T. (2022) High-accuracy ultrasonic method for in-situ monitoring of oil film thickness in a thrust bearing. Mechanical Systems and Signal Processing . 180. doi.org/10.1016/j.ymssp.2022.109453

    work page doi:10.1016/j.ymssp.2022.109453 2022

  43. [43]

    Wilcox, P. D. & Velichko, A. (2010) Efficient finite element modeling of scattering for 2D and 3D problems. Proc. SPIE 7650, Health Monitoring of Structural and Biological Systems 2010 . 76501E. doi.org/10.1117/12.847569

    work page doi:10.1117/12.847569 2010

  44. [44]

    A., Pettit, J., Lowe, M

    Choi, W., Skelton, E. A., Pettit, J., Lowe, M. J. S. & Craster, R. V. (2016) A generic hybrid model for the simulation of three-dimensional bulk elastodynamics for use in nondestructive evaluation. IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control . 63 (5), 726-736. doi.org/10.1109/TUFFC.2016.2535369. 22

    work page doi:10.1109/tuffc.2016.2535369 2016