If you've sat through CHEG 3220 and felt like heat transfer, mass transfer, and momentum transfer are somehow "the same thing wearing different costumes" — you're not wrong. But that intuition only gets you so far. The real unlock isn't memorizing that Fourier's law looks like Fick's law looks like Newton's law of viscosity. It's understanding the difference between two ideas that constantly get blurred together: flux and conservation.

Get this distinction straight, and a huge chunk of transport phenomena stops being "different equations for different chapters" and starts being one idea applied three times.

Flux is local. Conservation is global.

Here's the cleanest way to say it:

A constitutive law (Fourier's, Fick's, Newton's) tells you the flux at a point, based on the gradient at that same point. A conservation law tells you what happens to a region — a control volume — by comparing what flows in against what flows out.

That's it. That's the whole distinction. Everything else follows from it.

Flux answers: at this exact location, how fast and in what direction does stuff want to move?

Conservation answers: for this whole little chunk of material, is stuff piling up, draining away, or staying balanced?

These are different questions, asked at different scales, and that's exactly why they show up in the math differently.

control volume x x x + Δx jₓ|ₓ flux in jₓ|ₓ₊⁼ₓ flux out flux law gives jₓ at one point net gain per volume = -∂jₓ/∂x

Flux is measured at a single point (the dot in the middle). Conservation compares flux at two different boundaries of a control volume — that comparison is what introduces the second derivative.

Why flux uses a first derivative

A constitutive law is a local statement. Take 1-D heat conduction:

$$q_x'' = -k \frac{dT}{dx}\bigg|_x$$

Notice that everything on the right side is evaluated at the same point, $x$. The flux at that point depends only on the temperature slope at that point. Fick's law and Newton's law of viscosity have the identical structure:

$$J_{A,x} = -D_{AB}\frac{dC_A}{dx}\bigg|_x \qquad \qquad \tau_{xy} = -\mu \frac{du_x}{dy}\bigg|_y$$

In every case: flux at a point = -(transport property)(local gradient). One location, one slope, one flux. That's why the first derivative is enough — a first derivative is exactly the right tool for "how steep is this curve, right here."

Why conservation needs a second derivative

Conservation asks a fundamentally different question. It doesn't care what the flux is at one point — it cares about the difference between the flux entering one face of a control volume and the flux leaving the opposite face.

Picture a thin slab from $x$ to $x + \Delta x$. The net amount of stuff gained by that slab, per unit time, is:

$$\text{net gain} = j_x\big|_x - j_x\big|_{x+\Delta x}$$

Divide by $\Delta x$ and shrink it to a point, and you get:

$$\text{net gain per unit volume} = -\frac{\partial j_x}{\partial x}$$

That minus sign isn't arbitrary — it falls straight out of the bookkeeping. If more flux is leaving than entering, the quantity inside the slab is dropping, so accumulation has to be the negative of the net outflow.

Now substitute in the constitutive law. If $j_x = -\Gamma \dfrac{\partial \phi}{\partial x}$, then:

$$-\frac{\partial j_x}{\partial x} = -\frac{\partial}{\partial x}\left(-\Gamma\frac{\partial \phi}{\partial x}\right) = \Gamma \frac{\partial^2 \phi}{\partial x^2}$$

There it is. The second derivative isn't some special "diffusion operator" you have to memorize — it's just what happens automatically when you take the spatial derivative (conservation) of something that was already a derivative (flux). One local statement, fed into one global bookkeeping statement, and you get a second derivative for free.

This is also why diffusion flips from a minus sign to a plus sign. Fourier's and Fick's laws have a minus sign baked in. Conservation puts another minus sign in front of the flux's derivative. Two negatives, one positive — which is exactly why you see $+k\nabla^2 T$ and $+D_{AB}\nabla^2 C_A$ in the governing equations even though the underlying physics is "stuff moves downhill."

A physical reading of the second derivative

This part is worth sitting with, because it's where the abstract math turns back into something you can picture.

The first derivative, $\dfrac{\partial T}{\partial x}$, tells you the local heat flux — how hard heat is being pushed at that exact spot.

The second derivative, $\dfrac{\partial^2 T}{\partial x^2}$, tells you whether the flux is converging into a region or diverging out of it.

This is also why a linear temperature profile is the signature of 1-D steady-state conduction with no generation: the flux is constant everywhere, so there's nothing for conservation to "report." Nothing's accumulating because nothing's imbalanced.

The conservation equation, dissected

Once you see flux and conservation as separate steps, the general transport equation stops looking like a wall of symbols and starts looking like a sentence:

$$\frac{\partial (\rho \phi)}{\partial t} = -\nabla \cdot \vec{j} + \dot{S}$$

Or in plain English, the version every transport course eventually writes on the board:

$$\text{Accumulation} = \text{In} - \text{Out} + \text{Generation}$$

Swap in $T$ for heat, $C_A$ for species, $\rho \vec{v}$ for momentum, and you get the heat, mass, and momentum conservation equations respectively. The "stuff" changes. The bookkeeping logic does not.

Why this matters more than memorizing the analogy table

It's tempting to just memorize that $k \leftrightarrow D_{AB} \leftrightarrow \mu$ and call it a day. But if you understand why the structure repeats — local constitutive law, then global conservation, then automatic second derivative — you stop needing to memorize the governing equations at all. You can rebuild them from scratch under exam pressure, because you know the two-step recipe:

  1. Write the local flux law (first derivative, minus sign).
  2. Apply conservation to a control volume (another derivative, another minus sign, generation term if needed).

That's the whole machine. Everything in CHEG 3220 — boundary layers, the Biot number, transient conduction, diffusion through a stagnant film — is this same two-step process applied to a slightly different geometry or a slightly different boundary condition.