Investigation of Aeroacoustics and In-flight Particle Transport in Thermal Spray Supersonic Jets
D. Rahmat Samii, M. Tembely · 2026
Read Investigation of Aeroacoustics and In-flight Particle Transport in Thermal Spray Supersonic Jets by D. Rahmat Samii, M. Tembely (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-05547.txt.
Introduction
Samii and Tembely look into aeroacoustics and particle transport in thermal spray supersonic jets, whose sound pattern varies with working conditions. They build a math model linking chamber and nozzle settings with far-field sound levels, tuned to cut error from left-out turbulence. A setup maps supersonic jet flow and particle transport, using time-varying simulations with hybrid turbulence models to capture the shock-filled jet. Far-field sound follows the Ffowcs Williams-Hawkings analogy, with Lagrangian tracking of particles from nozzle exit, matching mic spectra and flux readings.
Catalog record
| CREATOR | D. Rahmat Samii, M. Tembely |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 108028 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.05547
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Investigation of Aeroacoustics and In-flight Particle Transport in Thermal Spray Supersonic Jets D. Rahmat Samii, M. Tembely Abstract The acoustic signature of thermal spray processes is known to vary with changes in operating conditions, which also influence particle in-flight velocity and distribution. Building on this idea, the present work first develops an analytical model that links chamber/nozzle parameters to far-field acoustic levels using gas-dynamics relations and simplified acoustic power propagation; the model is then calibrated to reduce systematic error associated with neglected turbulence effects and to improve agreement across operating conditions. In addition, a numerical framework is implemented to complement the analytical model and to resolve supersonic jet flow and in-flight particle transport. The second part of the study uses unsteady compressible simulations with hybrid turbulence modeling such as Unsteady Reynolds-Averaged Navier-Stokes(URANS) and Delayed Detached Eddy Simulation(DDES) to capture the development of the shock-containing jet and the associated near-field pressure fluctuations. Far-field sound is predicted using the Ffowcs Williams–Hawkings acoustic analogy, while a Lagrangian approach tracks particles injected at the nozzle exit to quantify velocity evolution, radial spreading, and downstream flux distributions. The influence of operating conditions (e.g. chamber pressure and temperature) is assessed, and predictions are evaluated against published microphone spectra and particle-flux measurements. Overall, the combined analytical and numerical approach captures how changes in nozzle operating conditions affect jet unsteadiness and mixing, leading to measurable shifts in acoustic level and spectral content. These results suggest that aeroacoustic signatures could be used as a non-intrusive pathway to monitor and potentially control thermal spray operating conditions. Keywords Thermal spray \cdot Aeroacoustics \cdot Computational fluid dynamics \cdot Supersonic jet \cdot Lagrangian particle tracking \cdot Ffowcs Williams–Hawkings \cdot Delayed Detached Eddy Simulations 1 Introduction It is well known in the thermal spray community that the sound of the torch is a significant indicator of spraying conditions. A skilled thermal spray specialist can often determine the overall quality of a coating solely based on the sound produced by the operating torch. In this study, the computational fluid dynamics(CFD) package OpenFOAM together with the Ffowcs Williams–Hawkings analogy, coupled with a Lagrangian particle-tracking method, is employed to simulate cold spray operation and capture pressure-induced noise fluctuations. Furthermore, the particle distribution downstream of the nozzle exit is analyzed. A complete CFD setup and configuration that can be used in future research to accurately model in-flight particle distributions, predict the generated acoustic signals, and establish a link between the sound characteristics and particle behavior is described. In section 2 , a brief overview of the existing literature on the CFD simulation of jet aeroacoustic noise is provided. In section 3 , an analytical model is developed, linking various design parameters of the nozzle and operating conditions to pressure fluctuations. In section 4 , the CFD governing equations, boundary conditions, discretization techniques, and simulation procedures are discussed. The methods and results of model verification and validation are detailed in section 5 . Finally, the CFD results are presented and examined in section 6 . 2 Literature Review Lighthill’s groundbreaking work in 1952, which developed the acoustic analogy approach by reformulating the Navier-Stokes equations to separate sound creation from propagation, laid the theoretical groundwork for aeroacoustic prediction [ 29 ] . In 1969, Ffowcs Williams and Hawkings expanded this innovative framework to include solid boundaries and moving surfaces [ 19 ] , resulting in what is still the most popular approach for far-field noise prediction in contemporary computational aeroacoustics. As noted by Petrosino and Barbarino, since these early advancements in the 1950s, research into jet-noise generation has continued as more advanced engine designs require faster and more accurate simulation and prediction techniques [ 38 ] . Early computational attempts to predict jet noise relied heavily on Reynolds-Averaged Navier-Stokes (RANS) methods due to their computational efficiency. Nevertheless, these methods turned out to be essentially insufficient for capturing the unsteady turbulent structures responsible for noise generation. Tyacke et al. [ 47 ] demonstrated that low-order RANS methods are inadequate for complex jets, as they inherently average out the temporal fluctuations crucial for acoustic radiation. This limitation prompted a paradigm shift toward Large-Eddy Simulation (LES), which directly resolves the energy-containing turbulent scales while modeling solely the smallest, dissipative scales. The transition to LES marked a fundamental improvement in predictive capability. Brès and Lele describe how advances in meshing, numerical schemes, and subgrid modeling now allow LES to represent nozzle-shape variations and capture turbulent boundary layers at the nozzle exit, significantly improving quantitative accuracy in far-field sound pressure and spectral shape [ 7 ] . The superiority of LES over RANS was definitively demonstrated by Murugu et al., who compared URANS, LES, and DES turbulence models for a Mach 0.8 chevron nozzle jet, finding that LES and DES predictions agreed well with experimental measurements, whereas URANS substantially under-predicted the noise [ 37 ] . A critical consideration in LES implementation is the choice between structured and unstructured grid strategies. Fosso Pouangué et al. conducted a comparison of two LES methodologies for a Mach 0.9 cold jet, contrasting block-structured grids with low-dissipative finite-volume schemes against fully unstructured tetrahedral grids employing low-dissipative Taylor-Galerkin finite-element schemes [ 21 ] . Both methods used LES to calculate acoustic sources and the FW-H analogy to propagate noise, showing that unstructured grids are just as accurate as structured methods at handling intricate noise-reduction devices like chevrons and dual-stream nozzles. This validation was essential for extending LES to industrially relevant configurations with geometric complexity. Despite LES’s success, the requirement to resolve thin near-wall layers makes pure LES computationally unfeasible for complex installations and high Reynolds-number jets. Due to this difficulty, hybrid RANS/LES techniques have been developed, in which LES resolves the free shear layers where the majority of noise is produced while RANS is used in the near-wall region [ 47 ] . Further advances include Wall-Modeled LES (WMLES), which Stich et al. describe as applying an analytical wall-stress law to eliminate the need for y^{+}\approx 1 resolution, enabling better aspect ratios and improved acoustic predictions within practical turnaround times [ 43 ] . For far-field noise prediction, hybrid acoustic methods have become standard practice. The resolved turbulent field is generally enclosed by a permeable boundary, and small-amplitude acoustic perturbations are transmitted analytically via the Ffowcs Williams-Hawkings equation. Positioning the FW-H surface excessively near the turbulent region eliminates source contributions, whilst positioning it too distantly results in numerical errors. The "end-caps" method has been developed to mitigate this issue by using several surfaces and phase averaging to reduce spurious noise [ 7 ] . The creation of specialized software platforms has made it easier to put these cutting-edge techniques into practice. Commercial CFD software has shown to be very capable. Accordingly, West and Caraeni performed LES of a Mach 0.75 jet from a round nozzle using STAR-CCM+ and achieved agreement with experimental data to within 1-2 dB for both near-field turbulence and far-field noise [ 52 ] . Similarly, Prasad et al. investigated passive and active flow-control strategies in supersonic jets using STAR-CCM+, demonstrating significant alterations in turbulence development and far-field sound pressure levels [ 40 , 41 ] . Dewan has also effectively utilized ANSYS Fluent, leveraging hybrid DDES formulations to capture important flow characteristics while producing accurate acoustic