Heat Transfer · Design background

Designing the Heated-Rod Energy Balance

Before reaching for RK4 or a root-finder, the actual engineering work is deciding what physics belongs in the model at all — and in what form.

▶ Play the lab 📖 Design background

1Choose a control volume and lump it

The rod is long and thin, with high conductivity (copper) — exactly the regime where the lumped capacitance assumption is justified: internal conduction is fast enough that the whole cross-section is at one temperature at any instant. That turns a 2-D (or 3-D) conduction problem into a single ODE for $T(t)$, with all the spatial detail absorbed into geometric factors (surface area, cross-sectional area).

2Write the energy balance

For a control volume of unit length, energy stored = energy generated − energy lost:

$$ \rho c_p A \frac{dT}{dt} = \dot E_g - \dot E_{out} = I^2R - \pi D\,h\,(T-T_\infty) - \pi D\,\varepsilon\,\sigma\,(T^4-T_{surr}^4) $$

Three terms on the right: Joule heating $I^2R$ (constant, since current and resistance are fixed), convective loss (linear in temperature difference — Newton's law of cooling), and radiative loss (a difference of fourth powers — Stefan-Boltzmann). The combination of a linear term and a $T^4$ term is what makes this nonlinear and gives it no closed-form solution.

3Find steady state first — it's just root-finding

Before solving the time-dependent problem, set $dT/dt=0$. What's left is a single nonlinear algebraic equation in $T$:

$$ I^2R = \pi D\,h\,(T_{ss}-T_\infty) + \pi D\,\varepsilon\,\sigma\,(T_{ss}^4-T_{surr}^4) $$

The right-hand side is strictly increasing in $T$ (more loss at higher temperature), so there's exactly one root — the same bracket-and-bisect approach as the root-finding lab. Solving for the steady state before the transient gives you a sanity check: the time-marching solution has to asymptote to exactly this value, no matter how it gets there.

▶ In the lab

The "steady-state T" stat is computed this way, independent of the time integration — and it matches the original paper's reported values exactly (786.2 K, 583.6 K, 308.8 K for the three cases). Open the lab →

4Integrate the transient with RK4

With steady state known as a target, integrate $dT/dt = f(T)$ forward in time from $T(0)=T_\infty$ using the same 4th-order Runge-Kutta scheme as the ODE Solver lab. There's nothing special about this equation that requires a different method — it's a textbook first-order ODE, just with a physically rich right-hand side.

5Decide which loss mechanism actually matters

This is the real design question the original problem is asking. Compare the convective and radiative loss terms at steady state:

Case	What changes	Effect
Bare copper, natural convection	ε=0.04 (low), h=20 (natural)	Convection dominates; radiation is nearly irrelevant at these temperatures and this emissivity.
Coated / oxidized surface	ε=0.8 (high), h=20	Radiation becomes a comparable or dominant loss path — steady-state temperature drops substantially.
Forced convection (fan)	ε=0.04, h=1000	Convection alone swamps both heating and radiation — rod barely warms, reaches steady state in seconds.

None of this is obvious without doing the calculation — emissivity and convection coefficient interact with the fourth-power and linear terms in ways that aren't intuitive from the formulas alone. That's exactly the kind of "what if" exploration this problem was built to support, then and now.

Key takeaways

Lumped capacitance turns a spatial conduction problem into a single ODE — valid when internal conduction is fast relative to surface losses.
Combine linear (convective) and $T^4$ (radiative) loss terms in one energy balance, and the result is nonlinear — solved numerically, not algebraically.
Always solve for steady state first (a root-finding problem) before integrating the transient — it's both a sanity check and a faster way to answer "how hot does it get."
Whether convection or radiation dominates depends entirely on the temperature range and surface emissivity — check both, don't assume.

▶ Play

Open the heated-rod lab

Solar Heat Exchanger Design

EngineeringCandy · Learn · the design thinking behind the playground