Heat Transfer: Convection

Convection is the mode of heat transfer that occurs between a surface and a moving fluid. It is the mechanism behind a radiator warming air in a room, water carrying heat away from an engine jacket, and ambient wind cooling a hot pipe. In practice, convection is rarely solved from first principles for real equipment. Engineers typically rely on heat transfer coefficients and empirical correlations that capture how geometry, flow conditions, and fluid properties shape the heat transfer rate.

At its core, convection links a surface temperature to the heat carried by a fluid through a thin region near the wall where velocity and temperature change rapidly. Understanding that near-wall region, and knowing how to estimate the resulting heat transfer coefficient, is the key to designing reliable thermal systems.

The Convection Heat Transfer Coefficient

Convection is commonly expressed with Newton’s law of cooling:

$$q=hA(T_s-T_{\infty })$$

where:

• $q$ is the heat transfer rate (W)
• $h$ is the convection heat transfer coefficient (W/m²·K)
• $A$ is the surface area (m²)
• $T_s$ is the surface temperature
• $T_{\infty }$ is the bulk fluid temperature away from the surface

The coefficient $h$ is not a material constant. It depends on the flow regime (laminar or turbulent), whether flow is driven by a fan/pump or buoyancy, the surface shape and orientation, and fluid properties such as viscosity and thermal conductivity. This is why convection design is correlation-driven: you determine $h$ using dimensionless groups and geometry-specific empirical fits.

Forced vs Natural Convection

Forced Convection

Forced convection occurs when an external device drives the flow, such as a fan over a heat sink or a pump through a heat exchanger. The resulting velocities are usually high enough that inertia dominates buoyancy, and heat transfer often increases strongly with flow rate. Forced convection is the workhorse of industrial thermal design because it is controllable: you can adjust velocity, mass flow, and channel dimensions.

Typical examples:

• Air forced across finned surfaces in electronics cooling
• Coolant pumped through tubes in a radiator
• Process fluids flowing in pipes and ducts

Natural (Free) Convection

Natural convection occurs when fluid motion is created by density differences due to temperature gradients. Warmer fluid near a hot surface becomes less dense and rises, while cooler fluid moves in to replace it. Natural convection is common in passive cooling scenarios, but it is more sensitive to orientation and ambient conditions.

Typical examples:

• Heat loss from a vertical hot plate to still air
• Cooling of a hot vessel in a quiescent room
• Heat transfer from a pipe when no fan is present

In many real systems both effects coexist (mixed convection). A modest ambient airflow can significantly change the heat transfer rate compared with pure natural convection.

Boundary Layers and Why Convection Correlations Exist

When fluid flows over a surface, a velocity boundary layer develops because the no-slip condition forces fluid velocity to be zero at the wall. Simultaneously, a thermal boundary layer forms because the fluid at the wall is driven toward the wall temperature. Most of the temperature drop between the wall and the bulk fluid occurs across this near-wall region. The thinner the thermal boundary layer, the higher the temperature gradient at the wall, and the larger the heat transfer coefficient.

Analytical solutions exist only for limited cases. For engineering geometries (flat plates, cylinders, tube banks, internal ducts, fin arrays), we use correlations expressed through dimensionless numbers that reflect the relative importance of inertia, viscosity, diffusion, and buoyancy.

Key Dimensionless Numbers

Convection correlations are usually framed in terms of the following:

• Nusselt number: $Nu=\frac{hL}{k}$

Represents the ratio of convective to conductive heat transfer across the boundary layer. Once you know $Nu$, you can find $h$.

• Reynolds number: $Re=\frac{\rho VL}{\mu }$

Measures the ratio of inertial to viscous forces and is central to forced convection. It helps determine laminar vs turbulent behavior.

• Prandtl number: $Pr=\frac{\mu c_p}{k}$

Compares momentum diffusivity to thermal diffusivity. It depends primarily on fluid type and temperature.

• Grashof number: $Gr=\frac{g\beta (T_s-T_{\infty })L^3}{{\nu }^2}$

Quantifies buoyancy effects and appears in natural convection.

• Rayleigh number: $Ra=GrPr$

A combined buoyancy-diffusion parameter widely used in natural convection correlations.

Here, $L$ is a characteristic length tied to geometry, $k$ is thermal conductivity, and $\nu =\mu /\rho$ is kinematic viscosity.

Using Empirical Correlations in Practice

The General Workflow

A practical convection calculation typically follows a consistent sequence:

1. Define geometry and characteristic length

For a flat plate, $L$ may be plate length in the flow direction. For internal flow in ducts, $L$ is often the hydraulic diameter.

1. Identify convection type and regime

Decide whether the situation is forced, natural, or mixed convection. For forced convection, estimate $Re$ to infer laminar or turbulent flow.

1. Evaluate fluid properties at an appropriate temperature

Many correlations assume properties at a film temperature, often approximated as $T_f=(T_s+T_{\infty })/2$. This matters because viscosity and conductivity can change significantly with temperature.

1. Select a correlation for the geometry and boundary condition

Correlations differ for constant wall temperature vs constant heat flux, for external flow vs internal flow, and for specific shapes.

1. Compute __MATH_INLINE_22__, then __MATH_INLINE_23__

Use $h=Nuk/L$, then evaluate heat transfer with $q=hA\Delta T$.

This approach is intentionally empirical. The correlations embed experimental behavior like transition to turbulence and the influence of developing flow.

Common Geometry Categories

External Flow Over Flat Plates and Surfaces

External forced convection on surfaces is common in cooling of panels, enclosures, and moving air streams. Correlations typically relate $Nu$ to $Re$ and $Pr$, with different expressions for laminar and turbulent regimes and for average vs local heat transfer coefficients. Surface roughness and flow disturbances can increase heat transfer by promoting turbulence, but they can also increase pressure drop.

Crossflow Over Cylinders and Tube Banks

Pipes and tubes exposed to airflow (or any crossflow) show behavior influenced by separation and wake formation. Tube banks in heat exchangers add complexity because upstream tubes alter the flow field for downstream tubes. Correlations account for tube spacing, arrangement, and Reynolds number range. In design, the convection coefficient is often paired with pressure-drop estimates because the fan or blower must supply the required flow.

Internal Flow in Pipes and Ducts

Internal convection in pipes is foundational in heat exchanger design. Here the heat transfer depends strongly on whether the flow is fully developed (velocity and temperature profiles stabilized) or developing (entrance region). Turbulent internal flow typically yields higher $h$ but at the cost of greater pumping power. The hydraulic diameter concept allows similar methods to be applied to non-circular ducts.

A key practical point: improving $h$ by increasing velocity may quickly become expensive because pumping power often rises steeply with flow rate. Convection design is usually an optimization between heat transfer and pressure drop.

Natural Convection Correlations and Orientation Effects

Natural convection depends heavily on how buoyant plumes form and rise. A vertical plate, a horizontal plate facing upward, and a horizontal plate facing downward can have noticeably different heat transfer because the buoyancy-driven flow organizes differently. Correlations typically use $Ra$ and reflect changes in flow regime as buoyancy becomes stronger.

In passive thermal systems, seemingly small changes like adding a shroud, moving nearby obstructions, or changing surface orientation can alter natural convection paths and thus the effective heat transfer coefficient.

Practical Insight: What Controls Convection Most

Across forced and natural convection applications, a few recurring levers dominate:

• Velocity and turbulence: Higher velocity and turbulence generally increase $h$ by thinning the boundary layer.
• Geometry and characteristic length: Smaller hydraulic diameters and well-designed fin structures can raise $Nu$, but may also increase pressure drop.
• Fluid properties: Higher thermal conductivity and appropriate viscosity improve convection performance. Property variation with temperature is often a first-order effect.
• Surface condition: Roughness can increase turbulence; fouling and deposits add thermal resistance and effectively reduce heat transfer.
• Temperature difference: For natural convection, $\Delta T$ directly influences buoyancy via $Gr$ and $Ra$, changing the flow strength and $h$.

Conclusion

Convection is the practical bridge between surface temperatures and real heat transfer rates in flowing fluids. Because the convection heat transfer coefficient is highly dependent on flow, geometry, and fluid properties, engineers rely on empirical correlations built around dimensionless numbers like $Re$, $Pr$, and $Ra$. The most effective convection designs treat heat transfer and fluid mechanics together: selecting the right correlation for the right geometry, evaluating properties carefully, and balancing improved heat transfer against the cost of moving fluid.