Eulerian vs Lagrangian Descriptions
Intuition
Two ways to study traffic: stand on a bridge and count cars passing below (Eulerian — fixed point, watch the field), or ride inside one car and log its journey (Lagrangian — follow the particle). Fluid mechanics constantly switches between these views, and the dictionary between them is the material derivative.
Formal definitions
| Characteristic | Eulerian | Lagrangian |
|---|---|---|
| Focus | fixed spatial points | individual fluid particles |
| Tracks | properties at locations | trajectory of each particle |
| Variables | fields \(f(\mathbf{x}, t)\) | particle functions \(f_P(t)\) |
| Frame | fixed (watching flow) | moving (riding the flow) |
Eulerian: the velocity field \(\mathbf{v} = \mathbf{v}(\mathbf{x}, t)\). The local rate of change at a fixed point is \(\partial f/\partial t\). This is the natural language of measurement (anemometer on a mast) and of CFD grids.
Lagrangian: each particle's position \(\mathbf{X}(t)\), with velocity \(\mathbf{v}_L = d\mathbf{X}/dt\). This is the natural language of Newton's laws — force acts on matter, not on points of space.
Flow visualization: three kinds of lines
- Streamlines — curves tangent to \(\mathbf{v}\) everywhere at one instant: \(\frac{dx}{u} = \frac{dy}{v} = \frac{dz}{w}\). Snapshots; can't cross in steady flow.
- Pathlines — actual particle trajectories over time: integrate \(\frac{d\mathbf{r}}{dt} = \mathbf{v}(\mathbf{r}(t), t)\). Pure Lagrangian objects.
- Streaklines — the locus of all particles that ever passed a fixed point (what a dye injector shows). History-dependent.
In steady flow all three coincide; in unsteady flow they differ, and confusing them is a classic exam trap.
Four motions of a fluid element
A velocity field generically does four things to a small blob of fluid:
- Translation (\(\mathbf{v}\)) — carries it along
- Rotation (\(\boldsymbol{\omega} = \nabla\times\mathbf{v}\)) — spins it
- Linear strain (\(\varepsilon_{xx}\)…) — stretches it
- Shear strain (\(\varepsilon_{xy}\)…) — skews it
Potential flow is the special case with zero rotation; viscosity generates stresses from the strain rates.
Common mistakes
- Thinking \(\partial \mathbf{v}/\partial t = 0\) means particles don't accelerate. A steady converging nozzle has \(\partial \mathbf{v}/\partial t = 0\) everywhere, yet every particle speeds up as it moves — the acceleration is convective. See material derivative.
- Reading dye photos as streamlines. Dye shows streaklines; they only match streamlines when the flow is steady.
Related concepts
- Material derivative — the bridge between the two views
- Reynolds transport theorem — the same bridge for finite volumes
- Potential flow — irrotational special case
Knowledge graph position
Prerequisites: What is a fluid? Leads to: Material derivative, Reynolds transport theorem.
Quiz
Q1 (conceptual). A weather station reports temperature falling at 2 °C/h. A balloon drifting with the wind reports temperature rising at 1 °C/h. Are these contradictory?
Answer
No. The station measures the Eulerian rate \(\partial T/\partial t\) at a fixed point; the balloon measures the Lagrangian rate \(DT/Dt\). They differ by advection \(\mathbf{v}\cdot\nabla T\) — the wind is blowing the balloon into warmer air.
Q2 (multiple choice). In an unsteady flow, the line traced by continuous dye injection at a point is a:
- (a) streamline (b) pathline (c) streakline (d) isobar
Answer
(c). Streaklines connect all particles that passed the injection point.
Q3 (conceptual). Why do computational fluid dynamics codes usually favor the Eulerian description?
Answer
Fields on a fixed grid are easy to discretize and the mesh never tangles. Tracking billions of individual particles (Lagrangian) is costly and the particles cluster, leaving regions unresolved. (Particle methods do exist — e.g. SPH, and the PIC method in plasma physics is a hybrid.)