Forced Convection CFD Analysis
Theory and Physics
Fundamental Concepts of Forced Convection
Professor, forced convection is about "cooling by moving fluid with fans or pumps," right? What are the key points when solving it with CFD?
Exactly. In forced convection, the flow driven by external means (fans, pumps, pressure differences in pipes, etc.) is dominant, and buoyancy is negligible. The goal in CFD is to accurately predict the wall heat flux or Nusselt number after solving for the velocity and pressure fields.
What does the Nusselt number signify?
The Nusselt number $Nu$ is the ratio of convective heat transfer at the wall to the fluid's thermal conduction, serving as a dimensionless indicator of heat transfer performance.
Here, $h$ is the heat transfer coefficient, $L$ is the characteristic length, and $k_f$ is the fluid's thermal conductivity. A larger $Nu$ indicates that convective heat transfer is dominant, meaning heat is being transported efficiently.
Typical Nusselt Number Correlations
What are some experimental correlations?
Commonly used correlations for internal flow are the Dittus-Boelter equation and the Gnielinski equation.
$n = 0.4$ (heating), $n = 0.3$ (cooling). Applicable range: $Re_D > 10{,}000$, $0.6 < Pr < 160$, $L/D > 10$.
The Gnielinski equation has better accuracy over a wider range ($2300 < Re_D < 5 \times 10^6$) than Dittus-Boelter. It's standard practice to first verify CFD results against these correlations.
What about external flow?
For forced convection over a flat plate, laminar flow uses $Nu_x = 0.332 Re_x^{1/2} Pr^{1/3}$, and turbulent flow uses $Nu_x = 0.0296 Re_x^{0.8} Pr^{1/3}$. For flow around a cylinder, the Churchill-Bernstein equation is often used.
Nusselt Number Correlations are "Fossils of Empirical Equations" – Why They're Still Used in Design Today
Classical Nusselt number correlations like Dittus-Boelter are empirical equations derived from experiments in the 1930s-50s. Even with the widespread adoption of CFD, they are still used on the front lines for initial heat transfer coefficient estimation during the early design stages. The reason is speed and convenience – correlations are the best way to get a "rough idea" before running CFD calculations. However, using them outside their applicable range (Re number, Pr number, geometry) can lead to errors exceeding 50%. Before using an equation, checking "Is it okay to use under these conditions?" is a fundamental habit for practical engineers.
Physical Meaning of Each Term
- Temporal Term $\partial(\rho\phi)/\partial t$: Think of the moment you turn on a faucet. At first, water comes out erratically and unstably, but after a while, the flow becomes steady, right? This "period of change" is described by the temporal term. The pulsation of blood flow from a heartbeat, or flow fluctuations each time an engine valve opens/closes – all are unsteady phenomena. So what is steady-state analysis? It looks only at "after sufficient time has passed and the flow has settled down" – meaning this term is set to zero. This significantly reduces computational cost, so trying a steady-state solution first is a basic CFD strategy.
- Convection Term $\nabla \cdot (\rho \mathbf{u} \phi)$: What happens if you drop a leaf into a river? It gets carried downstream by the flow, right? This is "convection" – the effect where fluid motion transports things. Warm air from a heater reaching the far corner of a room is also due to air, the "carrier," transporting heat via convection. Here's the interesting part – this term contains "velocity × velocity," making it nonlinear. This means as flow speed increases, this term rapidly strengthens, making control difficult. This is the root cause of turbulence. A common misconception: "Convection and conduction are similar" → They are completely different! Convection is transport by flow, conduction is transfer by molecules. There's an order-of-magnitude difference in efficiency.
- Diffusion Term $\nabla \cdot (\Gamma \nabla \phi)$: Have you ever put milk in coffee and left it? Even without stirring, it naturally mixes after a while. That's molecular diffusion. Now a question – honey and water, which flows more easily? Obviously water, right? Honey has high viscosity ($\mu$), so it flows poorly. Higher viscosity strengthens the diffusion term, making the fluid move "sluggishly." In low Reynolds number flows (slow, viscous), diffusion is dominant. Conversely, in high Re number flows, convection overwhelmingly dominates, and diffusion plays a supporting role.
- Pressure Term $-\nabla p$: When you push a syringe plunger, liquid shoots out forcefully from the needle tip, right? Why? The plunger side is high pressure, the needle tip is low pressure – this pressure difference provides the force pushing the fluid. Dam discharge works on the same principle. On a weather map, where isobars are tightly packed? That's right, strong winds blow. "Flow arises where there is a pressure difference" – this is the physical meaning of the pressure term in the Navier-Stokes equations. A point of confusion here: "Pressure" in CFD is often gauge pressure, not absolute pressure. When switching to compressible analysis, if results suddenly become strange, it might be due to mixing up absolute/gauge pressure.
- Source Term $S_\phi$: Warmed air rises – why? Because it becomes lighter (lower density) than its surroundings, so buoyancy pushes it upward. This buoyancy is added to the equation as a source term. Other examples: chemical reaction heat from a gas stove flame, Lorentz force acting on molten metal in a factory's electromagnetic pump... These are all actions that "inject energy or force into the fluid from the outside," expressed by the source term. What happens if you forget the source term? In natural convection analysis, forgetting buoyancy means the fluid won't move at all – a physically impossible result, like turning on a heater in a winter room but the warm air doesn't rise.
Assumptions and Applicability Limits
- Continuum Assumption: Valid for Knudsen number Kn < 0.01 (mean free path ≪ characteristic length)
- Newtonian Fluid Assumption: Shear stress and strain rate have a linear relationship (non-Newtonian fluids require viscosity models)
- Incompressibility Assumption (for Ma < 0.3): Density is treated as constant. For Mach number ≥ 0.3, compressibility effects must be considered
- Boussinesq Approximation (Natural Convection): Density variation is considered only in the buoyancy term; constant density is used in other terms
- Non-applicable Cases: Rarefied gas (Kn > 0.1), supersonic/hypersonic flow (shock capturing required), free surface flow (requires VOF/Level Set, etc.)
Dimensional Analysis and Unit Systems
| Variable | SI Unit | Notes / Conversion Memo |
|---|---|---|
| Velocity $u$ | m/s | When converting from volumetric flow rate for inlet conditions, pay attention to cross-sectional area units |
| Pressure $p$ | Pa | Distinguish between gauge and absolute pressure. Use absolute pressure for compressible analysis |
| Density $\rho$ | kg/m³ | Air: ~1.225 kg/m³ @20°C, Water: ~998 kg/m³ @20°C |
| Viscosity $\mu$ | Pa·s | Be careful not to confuse with kinematic viscosity $\nu = \mu/\rho$ [m²/s] |
| Reynolds Number $Re$ | Dimensionless | $Re = \rho u L / \mu$. Criterion for laminar/turbulent transition |
| CFL Number | Dimensionless | $CFL = u \Delta t / \Delta x$. Directly related to time-step stability |
Numerical Methods and Implementation
Wall Treatment Selection
I've heard that wall mesh treatment is especially important in forced convection CFD.
That's correct. To accurately predict wall heat transfer, the thermal boundary layer must be sufficiently resolved. There are two main approaches.
(1) Low-Reynolds number approach: Place the first cell near the wall at $y^+ \approx 1$ and solve directly from the viscous sublayer. Combining this with the SST k-ω model is standard.
(2) Wall Function approach: Place the first cell at $y^+ \approx 30$–$300$ and approximate the near-wall region with the log law. Fluent's Enhanced Wall Treatment is practical as it automatically switches based on $y^+$.
Which is better for heat transfer prediction?
For quantitative Nu number prediction, the Low-Re approach with $y^+ \approx 1$ is recommended. Wall functions can be used to grasp flow trends, but they risk over/underestimating Nu by 10–20%. Accuracy degrades particularly for flows with separation (backward-facing step, bluff bodies).
Turbulence Model Comparison
Which turbulence model is suitable for forced convection?
| Turbulence Model | Internal Flow | External Flow | Separated Flow | Computational Cost |
|---|---|---|---|---|
| Standard k-ε | Acceptable | Acceptable | Not Suitable | Low |
| Realizable k-ε | Good | Good | Acceptable | Low |
| SST k-ω | Good | Good | Good | Low–Medium |
| Transition SST | Good in Transition Region | Good in Transition Region | Good | Medium |
| k-ω BSL RSM | When Anisotropy is Important | Good | Good | High |
Is standard k-ε okay for the fully developed region of internal flow?
For Nu number prediction in fully developed turbulent pipe flow, standard k-ε can agree with the Dittus-Boelter equation within 5%. However, for entrance regions or pipes with sudden contraction/expansion, SST k-ω is safer. For transition regions ($Re \approx 2300$–$10000$), the Transition SST model becomes necessary.
Heat Transfer Coefficient Off by a Factor of 2 Due to Wall Treatment Selection Mistake
In forced convection CFD, the choice of near-wall turbulence treatment has the greatest impact. Whether using wall functions or resolving the near-wall region with a low-Re model significantly changes the heat transfer coefficient. In a car air conditioner design case, an engineer used wall functions (assuming y+≈30) with a coarse mesh, resulting in a heat transfer coefficient over 50% higher than measured. The actual y+ exceeded 200, completely outside the wall function's applicable range. "Just use wall functions by default" is one of the most costly misconceptions in forced convection.
Upwind Scheme
First-order Upwind: Large numerical diffusion but stable. Second-order Upwind: Improved accuracy but risk of oscillations. Essential for high Reynolds number flows.
Central Differencing
Second-order accurate, but numerical oscillations occur for Pe > 2. Suitable for low Reynolds number, diffusion-dominated flows.
TVD Scheme (MUSCL, QUICK, etc.)
Maintains high accuracy while suppressing numerical oscillations via limiter functions. Effective for capturing shocks or steep gradients.
Finite Volume Method vs Finite Element Method
FVM: Naturally satisfies conservation laws. Mainstream in CFD. FEM: Advantageous for complex shapes and multiphysics. Mesh-free methods like SPH are also developing.
CFL Condition (Courant Number)
Explicit method: CFL ≤ 1 is the stability condition. Implicit method: Stable even for CFL > 1, but affects accuracy and iteration count. LES: CFL ≈ 1 recommended. Physical meaning: Information should not travel more than one cell per timestep.
Residual Monitoring
Convergence is typically judged when residuals for the continuity equation, momentum, and energy drop by 3–4 orders of magnitude. The mass conservation residual is particularly important.
Relaxation Factor
Typical initial values: Pressure: 0.2–0.3, Velocity: 0.5–0.7. Reduce the factor if divergence occurs. Increase after convergence to accelerate.
Unsteady Calculation Inner Iterations
Iterate within each timestep until a steady solution converges. Inner iteration count: 5–20 is a guideline. If residuals fluctuate between timesteps, review the timestep size.
Analogy for the SIMPLE Method
The SIMPLE method is an "alternating adjustment" technique. First, velocity is tentatively calculated (predictor step), then pressure is corrected so that mass conservation is satisfied with that velocity (corrector step), and velocity is revised using the corrected pressure – this back-and-forth is repeated to approach the correct solution. It resembles two people leveling a shelf: one adjusts the height, the other balances it, repeating alternately.
Analogy for the Upwind Scheme
The upwind scheme is a method that "stands in the river flow and prioritizes upstream information." A person in the river looking downstream cannot tell where the water comes from – this discretization method reflects the physics that upstream information determines downstream conditions. It's first-order accurate but highly stable because it correctly captures flow direction.
Practical Guide
CFD Analysis of Heat Sinks
I've heard CFD is often used for heat sink design in electronic devices.
That's right. Aluminum extruded fins, pin fins, liquid-cooled cold plates, etc., all primarily rely on forced convection for cooling. Design variables include fin pitch, fin height, channel width, flow rate, etc. CFD is used to evaluate thermal resistance $R_{th} = \Delta T / Q$ and perform parametric studies.
Could you explain the specific workflow?
(1) Model only one pitch using fin symmetry (use periodic boundary conditions). (2) Set uniform velocity at inlet, pressure outlet at exit. (3) Apply constant heat flux to fin base. (4) Perform steady-state calculation with SST k-ω. (5) Calculate thermal resistance from the average temperature at the fin base.
Using just one pitch is computationally efficient. What about the effect of the entrance region?
For short heat sinks ($L/D_h < 20$), entrance effects cannot be ignored, so a full model is needed. For long heat sinks where fully developed flow dominates, one pitch with periodic conditions is sufficient. In Fluent, you can specify mass flow rate in the periodic condition, and STAR-CCM+ also allows setting via periodic interface.
Liquid-Cooled Cold Plate Design
Is CFD also used for liquid cooling design of power semiconductors?
Liquid-cooled cold plates are a typical CHT (Conjugate Heat Transfer) problem. Microchannels are machined into aluminum or copper plates, and water or coolant is flowed through them. When the hydraulic diameter $D_h$ of the channels is below 1mm, the Re number can become low, often resulting in laminar flow. In that case, a laminar model is sufficient.
Related Topics
なった
詳しく
報告