Lagrangian vs Eulerian: Fluid Kinematics Explained

Solid mechanics is easy to describe because a solid body keeps its identity: a crane hook is the same crane hook at every instant, so you can attach a coordinate system to it and follow it around. A fluid does not cooperate. The drop of water entering a pipe is not the same drop leaving it microseconds later — fluid deforms, mixes, and splits, so there is no convenient 'body' to track. Faced with that, kinematics offers two ways to describe the motion, and the choice between them is the first real decision in fluid mechanics.

The Lagrangian description names each particle by where it started — a label a⃗, usually its initial position — and follows that particle forever. Its position is a function of the label and time, r⃗ = r⃗(a⃗, t). Because you are riding along with one particle, its ordinary velocity and acceleration are just simple time derivatives. This is exactly how Newton's laws were written for particles, so it feels natural. The weather-balloon swarm in the hook is Lagrangian.

The Eulerian description gives up on particle identity entirely. Instead it plants a grid of points fixed in space and asks, at each point and each instant, what is the velocity of whatever fluid is passing through here? The answer is the velocity field, V⃗(x, y, z, t) — a vector attached to every point in space, varying with time. The weather-station row is Eulerian. So is almost every instrument an engineer builds: a pressure tap, a thermocouple, a flow meter all sit still and let the fluid flow past them.

An Eulerian velocity field is read like a weather map of arrows. At each point (x, y, z) the field assigns a vector with components u, v, w along x, y, z. To find the velocity at a particular spot you simply evaluate the field there; to find the speed you take the magnitude |V⃗| = √(u² + v² + w²). A field that contains no time t describes steady flow — the reading at every fixed point never changes, even though the fluid itself keeps moving through. A field with explicit time dependence is unsteady.

Here is the subtle point that the rest of this module builds on. In the Lagrangian view, a particle's acceleration is just dV⃗/dt. In the Eulerian view the obvious-looking derivative ∂V⃗/∂t is NOT the particle acceleration in general — it is only the rate of change observed at a fixed point (the local part). A particle can also accelerate simply by drifting into a region where the field is faster, even when the field itself never changes with time. Capturing that second, convective, contribution requires the material derivative, which is the subject of Lesson 3. For now, hold the warning: steady does not mean 'no acceleration'.