FLOW Lab UC Berkeley

What we study

Research Areas

Multiphase flows for energy production, transportation, industrial, biological, and environmental applications — the overarching, unifying theme of our research.

The goal of our research is to advance the physical understanding of high-Reynolds-number single- and multiphase flows, primarily through experiments and through the development and use of advanced measurement techniques. Multiphase flows appear in almost every aspect of modern life: offshore applications, biological flows, energy production, chemical processing, and naval hydrodynamics.

Specific topics include reducing the drag of marine vehicles, mitigating damage and noise caused by cavitation in naval and industrial applications, and efficient handling of flow in energy production. For details, see Publications and our YouTube channel.

Theme

Decarbonizing marine transport

Shipping moves ~80% of world trade and burns fuel mostly to overcome friction. We study the multiphase physics that could cut that bill.

Active2010–present

Multiphase flows for frictional drag reduction

Frictional drag accounts for roughly 60% of a typical cargo ship's propulsive power. Air-layer drag reduction could cut a ship's frictional resistance significantly — with major economic and environmental impact. We study how air layers form and persist, including gas injection from discrete ports rather than continuous slots.

CT scan showing gas trapped on a superhydrophobic surface

CT scan of gas trapped on a superhydrophobic surface. Left: axial cut (dark spots are plastrons). Right: gas interfaces seen through the wall, showing coverage.

Key papers

Active2018–present

Physics of superhydrophobic-surface drag reduction

Superhydrophobic surfaces (SHS) may reduce frictional drag on ships and in pipelines. Teaming with leading materials groups who develop the surfaces, we investigate the physical mechanisms of SHS drag reduction — and the limits of the technique.

Schematic of a turbulent boundary layer interacting with a deformable free surface, showing eddies producing surface scars and upwellings

Schematic: wall turbulence meeting a deformable free surface. The largest eddies reach the interface, leaving scars and upwellings that in turn feed back on the boundary layer.

Key papers

Active2024–present

Turbulent boundary layers meeting a free surface

Air lubrication ultimately succeeds or fails at an interface. We measure how a turbulent boundary layer and a free surface exchange momentum and vorticity — how eddies deform the surface, and how that deformation feeds back on the near-wall turbulence. This is the canonical problem sitting underneath the applied air-layer work.

Theme

Multiphase flows in energy & industrial systems

Bubbles, cavities and phase change govern heat transfer, loading and noise in reactors, pumps, valves and propulsors.

Schematic of two parallel subchannels connected by a narrow gap, with coherent structures driving cross-gap mixing, measured by wire-mesh sensors and X-ray densitometry

Schematic: two parallel subchannels connected by a narrow gap. Coherent structures in the gap drive cross-channel exchange; we resolve them with wire-mesh sensors and X-ray densitometry. See Int. J. Multiphase Flow (2023).

Key papers

Active2013–present

Multiphase mixing through narrow gaps

Mixing between adjacent flow channels connected by a narrow gap governs heat and mass transfer in the rod-bundle geometries central to nuclear thermal-hydraulics. We measure single- and two-phase mixing for both balanced and unbalanced channel flows, resolving the coherent structures within the gap and producing data intended for code validation.

Schematic of bubbly cross-flow over a circular cylinder: bubbles collect in the cores of shed vortices and separation moves aft

Schematic: bubbles injected into cross-flow over a cylinder collect in the low-pressure vortex cores, shift the shedding frequency, and move the separation point aft. See Int. J. Multiphase Flow (2026).

Key papers

Active2019–present

Bubble–vortex interaction and flow–structure coupling

Adding bubbles to a separated flow changes it qualitatively, not just quantitatively. In cross-flow over a cylinder we find that bubbles accumulate in vortex cores, shift the shedding frequency, and can trigger transition to the supercritical regime far below the single-phase Reynolds number — with direct consequences for drag, loading and noise.

Cavitation of a Newtonian fluid in a sudden gap expansion. (Silent video.)

Key papers

Active2010–present

Cavitation for industrial, transportation & medical applications

When local pressure drops below vapor pressure, a liquid can vaporize — cavitation. It appears in valves, on ship propellers, and even in the human body, and the inception pressure depends on the liquid's nuclei content and on nuclei at adjacent surfaces. We study cavitation in both Newtonian and non-Newtonian fluids.

Theme

Phase change & transport for health and environment

The same nucleation and transport physics decides whether an organ survives freezing and where an airborne droplet lands.

Schematic of photon-counting multi-energy X-ray computed tomography of a rigid constant-volume chamber during vitrification

Schematic: photon-counting, multi-energy X-ray CT of a rigid constant-volume chamber. Holding volume fixed raises pressure during cooling and suppresses ice. See Cryobiology (2024) — Arthur W. Rowe Best Paper Honorable Mention, Society for Cryobiology.

Key papers

Active2023–present

Phase change, vitrification and cryobiology

Vitrifying a biological sample without forming ice is fundamentally a nucleation and heat-transfer problem. With the Rubinsky group we use photon-counting, multi-energy X-ray computed tomography to observe isochoric (constant-volume) vitrification directly — resolving the ice- and cavity-free end state that isobaric cooling cannot reach.

Schematic of a turbulent particle-laden jet: large droplets fall ballistically while small droplets evaporate into suspended aerosol

Schematic: in a particle-laden turbulent jet, large droplets fall ballistically while small ones evaporate into aerosol carried by coherent structures. See Aerosol Sci. Technol. (2021).

Key papers

Earlier work2020–2022

Droplet and aerosol transport

Where a droplet ends up depends on its size and on the coherent structures carrying it. We study particle-laden turbulent jets — from coughs to laboratory spills — measuring how large droplets fall ballistically while smaller ones evaporate into aerosol that follows the flow, and how that partition sets deposition and exposure.

Theme

Measurement science: seeing the inaccessible

Many flows that matter are opaque, fast, or sealed inside metal. We build the instruments that make them measurable — and those instruments now travel far beyond fluids.

X-ray measurement of void fraction distribution in a shedding cavity

X-ray void-fraction fields in a shedding cavity at initial collapse and at the start of a new cycle. 1 ms exposure; 4% void-fraction and 0.5 mm spatial resolution. Void fractions span 0% in the free stream to nearly 100% in the sheet cavity. Flow right to left.

Key papers

Active2017–present

Advanced X-ray & optical flow diagnostics

Many of the most important multiphase flows are optically opaque. We develop and use time-resolved X-ray densitometry, multi-spectral computed tomography, X-ray particle velocimetry and tracking (XPV/XPTV), and tomographic PIV to measure what conventional optics cannot see.

Schematic comparing a conventional frame camera with an event camera, which reports per-pixel brightness changes asynchronously

Schematic: an event camera reports per-pixel brightness changes asynchronously with microsecond latency, rather than full frames at a fixed rate. See Meas. Sci. Technol. (2026).

Key paper

Active2023–present

Event-based (neuromorphic) cameras

Event cameras report only what changes, pixel by pixel, with microsecond latency — a natural match for sparse, fast phenomena such as cavitation inception, bubble passage and vortex shedding. We characterise when their frequency estimates can be trusted, and build triggering and detection systems around them.

Watch

The lab in motion

More on YouTube →
Droplet generation from a simulated laboratory accident: a 96-well plate dropped from counter height.
Lagrangian tracking of bubbles entrained by a plunging jet (A. Goujon, G. Chamoulaud & S. Mäkiharju).