Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 2610.28908

Monocular 3D bubble reconstruction: DNS-informed ray tracing and deep learning

Chaitanya S. Nayak·Faizaan Mohammed·Vivek Kumar·Shivam Prajapati·Suhas S. Jain·Cyrus K. Aidun
View original source ↗Read source document ↗Export metadata ↓

Abstract

Bubble size, depth, and spatial distribution govern transport and uniformity in multiphase flows. A single shadowgraphic view cannot separate bubble diameter from focal-plane displacement, as image signatures depend on both, while multicamera systems add optical access, calibration, and cost. We develop a single-camera physics-guided framework combining direct numerical simulation (DNS), validated Monte Carlo ray tracing, preprocessing, and focus-conditioned multitask deep learning to recover three-dimensional bubble position and physical-equivalent diameter from two-dimensional shadowgraphs. DNS-resolved bubble label images rendered at multiple focal-plane settings. The ray tracer reproduces shadowgraphic signatures with errors below $3\%$ (intensity), $1\%$ (apparent size), $15\%$ (blur), $6\%$ (numerical aperture), and $4\%$ (edge-spread-function shape), running $\sim30\times$ faster than the CPU implementation. The dataset has $900$ frames and $87{,}728$ labeled instances, partitioned by DNS timestamp to prevent leakage and expanded to over $10^{5}$ crops by broad-depth controlled-density rendering. A multitask network predicts signed depth and diameter from crops and focal-plane coordinates, achieving $\sim97.6\%$ normalized depth accuracy on held-out data; a closed-loop DNS-to-image-to-reconstruction test of $180$ bubbles gives diameter error below $9.71\%$ ($\sim90.29\%$ accuracy). On experimental images at four radial planes ($142$--$971$ bubbles each), PDF overlap is $68\%$--$92\%$ and log-PDF correlation is $0.72$--$0.87$. At the focal reference plane ($11{,}679$ bubbles), overlap is $\sim86\%$, correlation is $0.93$, and full-distribution, $D_{99.8}$, and composite errors are below $7.7\%$, increasing with focal-plane distance and defining a recoverability limit. Without stereoscopic hardware, the framework offers a scalable diagnostic for dense bubbly flows.

Explore related subjects

physics.flu-dyn

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Chaitanya S. Nayak, Faizaan Mohammed, Vivek Kumar, Shivam Prajapati, Suhas S. Jain, Cyrus K. Aidun. 2026-10-20. Monocular 3D bubble reconstruction: DNS-informed ray tracing and deep learning. https://arxiv.org/abs/2610.28908

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Data-Free Finite-Volume Convolutional Neural Networks for Steady Compressible Euler Flow around a Circular Cylinder

A data-free finite-volume convolutional neural network (FV-CNN) is formulated for steady Euler flow around a cylinder at eight Mach numbers, 2--15. Shared-face fluxes define conservative spatial balances, while automatic differentiation supplies parameter gradients. The derivative mismatch in the earlier automatic-differentiation-based convolutional neural network (AD-CNN) is derived. No computational fluid dynamics (CFD) or experimental field targets enter training; reference-independent selection of coefficients, seeds, and checkpoints is not established. Billig-based upstream fixing conditions shock-location agreement.For eight adopted fields, density and pressure relative $L_2$ discrepancies from finite-resolution CFD are $5.43$--$18.69\%$ and $6.25$--$19.70\%$. Closed-band defects are smaller than AD-CNN in 32/32 self-centered and 22--31/32 CFD-centered comparisons over seven half-widths, 0.06--0.12 cylinder diameters. AD-CNN retains smaller enthalpy root-mean-square (RMS) discrepancies in 7/8 cases, shock-position RMS at Mach 11 and 15, and temperature and transverse-velocity discrepancies at Mach 3.Ten new Mach-15 initializations and adopted seed 42 yield density discrepancies of $12.04$--$43.70\%$ (mean $24.34\%$, sample standard deviation $11.78$ percentage points); the new ten alone give $25.04\%$ and $12.17$ points. Dispersion exceeds the historical adopted-field gap of $3.40$ points; six of eleven outcomes are below AD-CNN. Transition-width ratios of AD-CNN to FV-CNN remain $4.07$--$4.66$, and all 44 self-centered seed--component balances at half-width 0.10 cylinder diameters are smaller for FV-CNN. Local sharpness and this balance metric persist, while initialization-robust accuracy superiority is not established.Component interventions remain one-seed conditional responses; Mach 2 uses a separate budget-limited protocol.

physics.flu-dyn↗
arXiv2026

Tangential and normal partial slip at the liquid-fluid interfaces: An application to a small liquid droplet, gas bubble, and aerosol

An analytical solution is obtained for the problem of the slow movement of a small drop of a fluid in another immiscible fluid in an infinitely large reservoir with the boundary condition of the normal slip and(or) tangential partial slip at the liquid-liquid or gas-liquid interface. That generalizes the conventional Navier and Maxwellian boundary conditions of partial slip. Normal slip is accompanied by the density gradient in the fluid and is applicable only if one of the phases in contact at the interface is a gas. At the first time the generalization of the Hadamard-Rybczynski equation (HRE) is discussed within the formalism of two slip lengths. It is proven that each of the two fluids separated by an interface has its own slip length. The slip lengths of the fluids that contact each other through an interface are related by the ratio of the viscosities of these fluids. New equations describing the terminal velocity of gas bubble rise and aerosol falling have been obtained. The theoretical results are in good agreement with the experiments at moderate Knudsen numbers, Kn < 0.25. It has been shown that the gas density within a rising bubble and around a falling droplet in the air is not uniform. The relative magnitude of the density increment increases with the size of the bubble or aerosol. Presumably, the best applicability of the generalized HRE should be expected for the interface of hydrophobic liquid and hydrophilic one (water - hydrocarbons, water - higher alcohols, in general: aqueous emulsions, water - lipophilic organic liquids and oils, etc.). These are quite important emulsions in practical terms, for example, for the oil industry and medicine. Experimental methods for determining the slip length are considered.

physics.flu-dyn↗
arXiv2026

Mixing and spreading of gravity currents in heterogeneous porous media

We analyze the mixing, migration and spreading of a gravity current in a heterogeneous porous medium using high-fidelity numerical simulations. Heterogeneity is represented by log-normal permeability fields of varying correlation lengths and variance. Stable and unstable density stratification scenarios are considered through linear and non-monotonic density laws, respectively. Heterogeneity reduces dissolution and increases the speed of the gravity current proportionally to the Rayleigh number. In the unstable case, heterogeneity accelerates the onset of convection. Convection-driven dissolution slows down the gravity current and counteracts the dispersive effect of heterogeneity resulting in a narrower interface and higher dissolution than in the stable case. Permeability anisotropy reduces dissolution because of the barrier effect of low permeability regions, except when blobs of buoyant fluid are trapped in low permeability structures and rapidly dissolve. The variance of the log-permeability field enhances dissolution. However, the homogeneous case outperforms heterogeneous cases except when Rayleigh number is small. This suggest an interaction between the size of the instabilities, the correlation length of the permeability field and the dispersive and barrier effects of the permeability field that controls dissolution efficiency.

physics.flu-dyn↗