Modeling Flow Regime Transitions in Thermal and Multiphase Systems

43 concepts18 papers1 proposal

In fluid dynamics, systems rarely behave linearly across all scales. Instead, they undergo dramatic, abrupt transformations when specific physical thresholds are crossed. A fluid that quietly transfers heat via static molecular conduction can suddenly erupt into dynamic, buoyancy-driven loops. Similarly, solid particles suspended in a gas stream can abruptly break away from fluid streamlines, transforming from a well-behaved mixture into an abrasive bombardment on containment walls.

Predicting these boundaries requires moving beyond brute-force simulation. Instead, researchers rely on dimensionless numbers and simplified physical approximations to pinpoint exactly why and whereto a flow regime shifts.

The Conduction-to-Convection Transition and Buoyancy Modeling

When a fluid is heated from below or near a localized source, thermal expansion creates density gradients. Gravity acts on these gradients, generating buoyancy forces. If these forces overcome the fluid's viscous resistance, the primary mechanism of heat transfer transitions from conduction to natural convection.

To simulate this transition without overwhelming computational resources, researchers frequently turn to the Boussinesq approximation. This mathematical simplification treats fluid density as a constant in all terms of the governing Navier-Stokes equations except for the gravity-dependent buoyancy term. As demonstrated by Haijiang He et al. in their study of Rayleigh–Taylor turbulence, this approach isolates the essential physics of density-driven mixing while discarding negligible acoustic and compressible complexities (Entropy Generation Rates in Two-Dimensional Rayleigh–Taylor Turbulence Mixing). The same principle allowed Judith Hilton et al. to efficiently model how complex terrain alters buoyancy-driven vortex pairs within massive wildfire plumes (Wind-terrain effects on firebrand dynamics).

Understanding this transition is critical because natural convection fundamentally changes how heat distributes. In phase change materials (PCMs) used for thermal energy storage, ignoring buoyancy-driven flow leads to massive inaccuracies. Ying Xu et al. showed that applying the Boussinesq approximation to liquid phases reveals a significantly higher liquid fraction and more uniform heat distribution than conduction-only models can predict (Heat transfer characteristics of cascade phase change energy storage composite pipeline). This convective dominance alters the liquid-solid interface early in the melting process, a phenomenon detailed by Yan Yang et al. when evaluating nanoparticle-enhanced PCMs (Nanoparticles to Enhance Melting Performance of Phase Change Materials for Thermal Energy Storage).

Low Richardson Number (Ri << 1)       High Richardson Number (Ri >> 1)
   [Forced Convection]                   [Natural Convection]
 ┌───────────────────────┐             ┌───────────────────────┐
 │ ───►  ───►  ───►  ───►│             │     ┌───►───┐         │
 │                       │             │     │       │         │
 │ ───►  ───►  ───►  ───►│             │     └───◄───┘         │
 └───────────────────────┘             └───────────────────────┘
  Driven by mechanical forces           Driven by buoyancy forces
  (Smooth, parallel isotherms)          (Distorted, looping isotherms)

The arbiter of this transition is often a dimensionless ratio. In enclosed environments with moving boundaries, M. Jahirul Haque Munshi et al. demonstrated that the Richardson number ($Ri$) dictates the flow field (A Numerical Study of Mixed Convection in Square Lid-Driven with Internal Elliptic Body and Constant Flux Heat Source on the Bottom Wall). When mechanical forcing dominates ($Ri \ll 1$), isotherms remain smooth and parallel. As buoyancy forces take over ($Ri \gg 1$), the flow field shifts to large, recirculating loops that distort isotherms near the heat source.

A similar convective mode transition occurs during "thermal soak" in automotive engine bays. Zhenhai Gao et al. captured the abrupt shift that occurs when a vehicle shuts down: forced mechanical cooling ceases, and the system transitions to a regime driven entirely by natural buoyancy and radiation, causing localized heat accumulation (Numerical investigation of thermal soak within engine bay using lattice Boltzmann method). Even in high-energy applications like gas metal arc welding, Chunsheng Wang et al. observed that convective fluid flow within the molten pool quickly dominates over transient thermal arc effects, making the pool's internal temperature distribution remarkably resilient to rapid current fluctuations (Molten Pool Behaviors in Double-Sided Pulsed GMAW of T-Joint: A Numerical Study).

The Anomalous Case of Water

This transition becomes highly non-linear in water-based systems due to water's unique density maximum at 4°C. As water cools toward freezing, it becomes denser until it hits 4°C, below which it expands and becomes lighter. This density inversion completely flips the direction of natural convection currents.

Feng Shaohui et al. found that this inversion triggers sudden internal convection loops that accelerate heat exchange far beyond standard conduction predictions, complicating the thermal stabilization of large liquid masses near freezing (Antifreeze design for Muon Detector of LHAASO). To control this in ice thermal storage systems, Chunwei Zhang et al. demonstrated that the geometric placement of a heat source can either exploit or suppress these inversion-driven currents, directly dictating how fast the surrounding ice front melts (Numerical Analysis of Heat Transfer Behaviours of Melting Process for Ice Thermal Storage Based on Various Heat Source Configurations).

Inertial Decoupling and Streamline Deviation

A parallel transition paradigm exists in multiphase and particulate transport: the threshold where momentum forces a discrete phase to decouple from the carrier fluid. This is known as inertia-driven streamline deviation.

When particles are small or the flow is slow, aerodynamic drag forces particles to follow the fluid's curved streamlines perfectly. However, if a particle's mass or velocity increases, its momentum overcomes this drag. At a critical threshold, the particles cross the fluid streamlines and collide with solid boundaries.

      Fluid Streamline ───► ───► ───► ┐
                                      │
      Low Stokes (Stk << 1)           ▼
      Particle follows flow ───► ───► ┐
                                      │
                                      ▼
      High Stokes (Stk ~ 1)
      Particle deviates ───────► ───────► [Impact Wall!]

This transition is governed by the Stokes number ($Stk$), the ratio of a particle's relaxation time to the characteristic time of the flow field. Paul T Ogunlela et al. utilized this mechanism to explain severe erosive wear in industrial riser pipes (CFD Approach to the Influence of Particle Size on Erosive Wear in Coal Riser Pipes). When the Stokes number approaches unity at pipe bends, coal particles decouple from the gas stream and strike the containment walls directly.

Interestingly, the authors uncovered a non-linear threshold effect: erosive wear does not increase indefinitely with particle size. Instead, it peaks at a critical particle diameter (300–350 µm) before declining, because larger particles reduce the total number density of impacts for a given mass flow rate (Investigating the Effect of Particle Size on Erosive Wear in Industrial Coal Pneumatic Transport Using Computational Fluid Dynamics). A similar momentum threshold was identified by Mao Sheng et al. in downhole fracturing tools, where high pumping rates give abrasive sand grains enough inertia to bypass upper outlets entirely, shifting the primary site of mechanical failure to the base of the tool (CFD analysis and field observation of tool erosion caused by abrasive waterjet fracturing).

Flow Regimes in Confined Spaces

In highly confined geometries, such as channels or porous packed beds, transitions are governed by the interplay between fluid momentum, viscous drag, and boundary geometry.

In horizontal channels, the transition from laminar to turbulent flow dictates whether particles settle out or remain suspended. Daniel Orea et al. captured this experimentally, demonstrating a critical threshold where the onset of chaotic, turbulent fluid motion effectively counteracts gravitational settling, causing particle deposition on channel walls to drop sharply (PARTICLE DEPOSITION STUDY IN A HORIZONTAL SQUARE CHANNEL USING HIGH-RESOLUTION IMAGING).

Within the tortuous voids of a packed bed reactor, the transition is mediated by the bed's void fraction and the particle Reynolds number ($Re_p$). Early photographic work by Kenji Kubo et al. mapped these internal flow patterns from smooth laminar streams to chaotic interstitial turbulence (Photographic observation of flow pattern in voids of packed bed of spheres.).

By analyzing these patterns, the authors developed a "side mixing model" that identifies a distinct dynamic threshold at $Re_p \approx 10$ (Axial Mixing Characteristics in Packed Beds). Below this threshold, fluid trapped in stagnant "side pockets" near sphere contacts remains isolated, acting as a static barrier to mass transport. Above $Re_p \approx 10$, however, the increasing momentum of the main flow begins to progressively strip away and sweep these stagnant zones into the active stream, altering the mixing efficiency of the entire system.

Key Foundational Work

Axial Mixing Characteristics in Packed Beds (1981) — contributed 2 concept elements to this direction.
Photographic observation of flow pattern in voids of packed bed of spheres. (1978) — contributed 1 concept elements to this direction.
History & citationVersion 1 · 2026-10-04

Current · Permanent link

View frozen version 1

Cite this direction

Latest version (a living page):

@misc{lacuna_boussinesq_approximation_54781006,
  title = {Boussinesq approximation},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/boussinesq-approximation/rdr_54781006d24548b6980a45975162e9bc},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-20},
}

This exact version (version 1, frozen text):

@misc{lacuna_boussinesq_approximation_54781006_v1,
  title = {Boussinesq approximation},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/boussinesq-approximation/rdr_54781006d24548b6980a45975162e9bc/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-20},
}
About

These directions overlap with this one through shared papers.

See something wrong? Report a mistake