pith. sign in

arxiv: 2604.14003 · v2 · submitted 2026-04-15 · ⚛️ physics.flu-dyn

Flow Characterization of the Delft Multiphase Flow Tunnel

Lina Nikolaidou , Angeliki Laskari , Tom van Terwisga , Christian Poelma This is my paper

Pith reviewed 2026-05-10 12:05 UTC · model grok-4.3

classification ⚛️ physics.flu-dyn
keywords cavitation tunnelflow characterizationlaser Doppler anemometryturbulence intensityflow uniformityturbulent boundary layerDelft multiphase flow tunnel
0
0 comments X

The pith

The new Delft multiphase flow tunnel maintains freestream turbulence of 0.5-0.6 percent and uniform mean flow to within 1 percent across most of the test section.

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

This paper presents velocity measurements taken in a newly built cavitation tunnel to document its flow quality before use in experiments. Laser Doppler anemometry recorded speeds from 2.13 to 9 m/s and established a linear link between freestream velocity and thruster rotation rate. After noise correction, turbulence intensity stayed between 0.5 and 0.6 percent with no evidence of large-scale mean-velocity drifts over long records. Local surveys showed the core flow remains uniform to better than 1 percent of the free-stream speed, while thin turbulent boundary layers near the walls produce the only notable deviations. The boundary layer itself begins upstream of the test section and grows at different rates on the side and top walls.

Core claim

After noise removal from the LDA records, the freestream turbulence intensity lies between 0.5 percent and 0.6 percent throughout the test section. The mean flow is spatially uniform with local deviations below 1 percent of the freestream velocity except in the immediate vicinity of the side walls. Long-duration center-line records show no large-scale temporal fluctuations of the mean velocity. The turbulent boundary layer originates upstream of the test section, its thickness is smaller on the side walls than on the top wall, and its growth deviates from canonical zero-pressure-gradient behavior.

What carries the argument

Laser Doppler anemometry (LDA) point measurements performed in multiple horizontal and vertical planes plus extended time series at the test-section center to quantify turbulence intensity, spatial uniformity, and boundary-layer profiles.

If this is right

  • The reported turbulence level and uniformity allow experiments that require low-disturbance incoming flow.
  • Boundary-layer corrections are needed only within a few centimeters of the walls.
  • Flow speed can be set accurately by controlling thruster frequency via the measured linear relation.
  • Absence of large-scale unsteadiness supports repeatable time-averaged statistics.
  • Non-canonical boundary-layer growth implies that inlet geometry upstream of the test section must be considered in future modeling.

Where Pith is reading between the lines

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

  • The same LDA protocol can be repeated after future modifications to the facility to quantify any changes in flow quality.
  • Numerical simulations of the entire tunnel circuit could be validated directly against the measured mean-velocity and turbulence profiles.
  • The thinner side-wall boundary layers suggest that corner flows or secondary motions influence the development and could be studied with stereo-PIV.
  • The low turbulence value sets a baseline for assessing how cavitation or multiphase phenomena interact with the incoming flow.

Load-bearing premise

Subtracting the estimated measurement noise from the LDA velocity variance yields an unbiased value for the true fluid turbulence intensity.

What would settle it

Independent hot-wire anemometer measurements at the same locations and speeds, processed with a separate noise-subtraction method, would show whether turbulence intensity remains inside the 0.5-0.6 percent band.

Figures

Figures reproduced from arXiv: 2604.14003 by Angeliki Laskari, Christian Poelma, Lina Nikolaidou, Tom van Terwisga.

Figure 1
Figure 1. Figure 1: Sketch of the Multiphase Flow Tunnel. Flow in the test section is from right to left. The [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Schematic representation of the “fringe model” in two planes. [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Components of our LDA system. Adapted from [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Schematic representation of the different signal components. Source: [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Schematic representation of the flow uniformity measurement positions (not to scale). Flow [PITH_FULL_IMAGE:figures/full_fig_p005_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: TBL measurements for U∞ = 5 m/s: (a) top and side wall TBL thickness measurements in three streamwise positions and (b) top wall normalized mean streamwise velocity profiles in the same positions. Boundary layer measurements of the top wall were performed for U∞ of 3.5, 5 and 7 m/s (or 200, 285 and 400 RPM). The boundary layer thickness was defined as the wall normal distance where the velocity is 99% of t… view at source ↗
Figure 7
Figure 7. Figure 7: Local measurements of the mean streamwise velocity [PITH_FULL_IMAGE:figures/full_fig_p008_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Local mean streamwise u (a, b, c) and vertical v (d, e, f) velocity profiles in three streamwise locations for 285 RPM or U∞ = 5 m/s. See also [PITH_FULL_IMAGE:figures/full_fig_p008_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Local turbulence intensity of the streamwise (a, b, c) and vertical (d, e, f) velocity in three [PITH_FULL_IMAGE:figures/full_fig_p009_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Ratio of the local streamwise velocity and freestream velocity for [PITH_FULL_IMAGE:figures/full_fig_p010_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: PDF of the local streamwise velocity u for U∞ = 3.5 m/s or 200 RPM (a, b, c), U∞ = 5 m/s or 285 RPM (d, e, f) and U∞ = 7 m/s or 400 RPM (g, h, i) in three streamwise locations. the time signal. Additionally, a local minimum is observed near the tunnel turnover frequency in the intermediate U∞ case. However, these features are subtle and may result from signal processing effects such as filtering or spectr… view at source ↗
Figure 12
Figure 12. Figure 12: Original and filtered velocity signal for three [PITH_FULL_IMAGE:figures/full_fig_p012_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: Turbulent intensity in the freestream (x = 0.05L) for a wide range of RPM of the pump (125-500 RPM), corresponding to U∞ ranging from 2-9 m/s. 4.4.2 RPM - U∞ correlation The same measurements were used in order to establish the relation between the RPM and the freestream velocity at the tunnel ( [PITH_FULL_IMAGE:figures/full_fig_p013_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: Correlation of the freestream velocity and the rotation frequency of the VIT. [PITH_FULL_IMAGE:figures/full_fig_p013_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: Correlation of the freestream velocity at [PITH_FULL_IMAGE:figures/full_fig_p014_15.png] view at source ↗
Figure 16
Figure 16. Figure 16: Local measurements of the mean streamwise velocity [PITH_FULL_IMAGE:figures/full_fig_p016_16.png] view at source ↗
Figure 17
Figure 17. Figure 17: Local measurements of the mean streamwise velocity [PITH_FULL_IMAGE:figures/full_fig_p016_17.png] view at source ↗
Figure 18
Figure 18. Figure 18: Local turbulence intensity of the streamwise (a, b, c) and vertical (d, e, f) velocity in three [PITH_FULL_IMAGE:figures/full_fig_p017_18.png] view at source ↗
Figure 19
Figure 19. Figure 19: Local turbulence intensity of the streamwise (a, b, c) and vertical (d, e, f) velocity in three [PITH_FULL_IMAGE:figures/full_fig_p017_19.png] view at source ↗
read the original abstract

At the end of 2020, a new cavitation tunnel was commissioned at the Ship Hydrodynamics laboratory of TU Delft, replacing its 1960s predecessor. Since this was a new facility, a flow characterization campaign was performed to investigate the flow quality in the test section. To that end, velocity measurements were performed in the test section using Laser Doppler Anemometry. Velocities in the range of 2.13 m/s to 9 m/s were measured and a linear relation was found between the freestream velocity and the rotational frequency of the thruster. Long term measurements at the center of the test section, did not reveal any large scale fluctuations of the mean velocity. The freestream turbulence intensity was found to lie between 0.5% - 0.6% throughout the test section, after removing the measurement noise. Local measurements in various planes in the test section confirmed that the flow is uniform ($u_{local}< U_{\infty} \times 1\%$), with few outliers near the side walls, due to the turbulent boundary layer. Finally, preliminary measurements of the turbulent boundary layer (TBL) indicated that the TBL originates upstream of the test section and its growth is not strictly canonical. Smaller TBL thickness was found in the side wall compared to the top wall.

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 / 2 minor

Summary. The manuscript reports an experimental flow characterization campaign in the newly commissioned Delft Multiphase Flow Tunnel using Laser Doppler Anemometry. Measurements across freestream velocities of 2.13–9 m/s establish a linear relation with thruster frequency, confirm the absence of large-scale mean-velocity fluctuations, report freestream turbulence intensity of 0.5–0.6 % after noise removal, demonstrate flow uniformity (local velocity deviations <1 % of U_∞ except near walls), and provide preliminary observations that the turbulent boundary layer originates upstream of the test section with non-canonical growth and differing thicknesses on the side and top walls.

Significance. If the reported uniformity and turbulence levels hold after detailed validation of the data-processing steps, the work supplies a valuable, directly measured baseline for a new multiphase/cavitation facility. The purely experimental character, with all quantities obtained from LDA data without fitted parameters or circular derivations, strengthens its utility as a reference for the hydrodynamics community.

major comments (2)
  1. [Abstract / freestream turbulence results] The central claim that freestream turbulence intensity lies between 0.5 % and 0.6 % throughout the test section rests on an unspecified noise-removal procedure applied to the LDA time series (Abstract). Standard LDA noise models require either an independent calibration (stationary target or zero-lag autocorrelation) or a validated statistical model; without a description of the exact method, its assumptions, or an independent cross-check (hot-wire, Monte-Carlo simulation of processor settings), it is impossible to confirm that the subtracted variance is unbiased at these low levels.
  2. [Measurement and data-processing sections] No details are provided on particle seeding, LDA burst processing, filtering, or uncertainty quantification (error bars, confidence intervals) for the reported velocities, uniformity metric, or turbulence intensities. These omissions directly affect the load-bearing claims of u_local < U_∞ × 1 % and the 0.5–0.6 % TI range.
minor comments (2)
  1. [Abstract / TBL results] The abstract states that the turbulent boundary layer “originates upstream of the test section” and shows non-canonical growth; a brief sketch of the upstream geometry or reference to facility drawings would help readers interpret the preliminary TBL thickness differences between side and top walls.
  2. [Long-term velocity measurements] Long-term measurements at the test-section center are said to show “no large scale fluctuations”; a quantitative statement (e.g., standard deviation of the running mean or power spectrum) would strengthen this negative finding.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful and constructive review, which highlights the potential value of the flow characterization for the hydrodynamics community. We agree that the major comments identify areas where additional detail is required to support the reported claims, and we will revise the manuscript to address them.

read point-by-point responses

  1. Referee: [Abstract / freestream turbulence results] The central claim that freestream turbulence intensity lies between 0.5 % and 0.6 % throughout the test section rests on an unspecified noise-removal procedure applied to the LDA time series (Abstract). Standard LDA noise models require either an independent calibration (stationary target or zero-lag autocorrelation) or a validated statistical model; without a description of the exact method, its assumptions, or an independent cross-check (hot-wire, Monte-Carlo simulation of processor settings), it is impossible to confirm that the subtracted variance is unbiased at these low levels.

    Authors: We acknowledge that the noise-removal procedure was described only briefly in the original manuscript. In the revised version we will add a dedicated subsection detailing the exact method applied to the LDA time series (including the statistical model or calibration approach used), its assumptions, and any validation performed. If an independent cross-check (such as zero-lag autocorrelation or Monte-Carlo simulation of processor settings) was carried out during the campaign, it will be reported explicitly; otherwise the limitations of the procedure at these low turbulence levels will be discussed openly. revision: yes

  2. Referee: [Measurement and data-processing sections] No details are provided on particle seeding, LDA burst processing, filtering, or uncertainty quantification (error bars, confidence intervals) for the reported velocities, uniformity metric, or turbulence intensities. These omissions directly affect the load-bearing claims of u_local < U_∞ × 1 % and the 0.5–0.6 % TI range.

    Authors: We agree that the measurement and data-processing sections lack sufficient detail. The revised manuscript will expand these sections to include: the seeding particles (material, nominal diameter, concentration and injection method), LDA system parameters and burst-processing settings (detection thresholds, validation criteria, and processor configuration), any temporal or spatial filtering applied to the velocity records, and the uncertainty quantification procedure (including how random and systematic uncertainties were estimated and propagated to the uniformity metric and turbulence intensity). This will directly support the reported values of local velocity deviation and freestream turbulence intensity. revision: yes

Circularity Check

0 steps flagged

No circularity: purely experimental measurements with no derivations or self-referential steps

full rationale

This is a flow characterization study reporting direct LDA velocity data from a new facility. Key results (freestream TI of 0.5-0.6% after noise removal, local uniformity u_local < U_∞ × 1%, TBL observations) are empirical outputs from time-series processing. No equations, models, fitted parameters, or predictions are presented that reduce to inputs by construction. The noise-removal procedure is a methodological detail whose correctness is an external assumption, not a circular reduction. No self-citations, uniqueness theorems, or ansatzes are used in a load-bearing manner. The paper contains no derivation chain to analyze; all claims trace to raw measurements.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

Claims rest on standard experimental fluid mechanics assumptions about LDA accuracy and noise handling rather than new postulates.

axioms (2)
  • domain assumption Laser Doppler Anemometry yields accurate point velocity measurements in seeded water flows when optical access and seeding are adequate.
    Implicit in all reported velocity and turbulence results.
  • domain assumption Measurement noise can be identified and subtracted to recover the true freestream turbulence intensity.
    Directly invoked to arrive at the 0.5-0.6% value.

pith-pipeline@v0.9.0 · 5539 in / 1255 out tokens · 43613 ms · 2026-05-10T12:05:10.824573+00:00 · methodology

discussion (0)

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

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Drag reduction regimes in air lubrication

    physics.flu-dyn 2026-04 unverdicted novelty 5.0

    Experimental characterization identifies three air lubrication regimes and proposes a scaling for the critical air flow rate marking the transition to the air layer regime with up to 60% drag reduction.

Reference graph

Works this paper leans on

4 extracted references · 4 canonical work pages · cited by 1 Pith paper

  1. [1]

    and George, W

    Buchhave, P. and George, W. K. (1978).Bias corrections in turbulence measurements by the laser Doppler anemometer. State University of New York at Buffalo, Faculty of Engineering and Applied

    work page 1978

  2. [2]

    Clauser, F. H. (1956). The turbulent boundary layer.Advances in applied mechanics, 4:1–51. ITTC (2006).Testing and Extrapolation Methods, General Density and Viscosity of Water. ITTC - Recommended Procedures and Guidelines. International Towing Tank Conference. Mayo Jr, W., Shay, M., and Riter, S. (1974). Digital estimation of turbulence power spectra fro...

    work page 1956

  3. [3]

    and Passchier, D

    Tummers, M. and Passchier, D. (2001). Spectral analysis of biased lda data.Measurement Science and Technology, 12(10):1641

    work page 2001

  4. [4]

    White, F. M. and Majdalani, J. (2006).Viscous fluid flow, volume 3. McGraw-Hill New York. 18

    work page 2006