2.1.5 · D1Analytical Mechanics

Foundations — Derivation of Euler-Lagrange equations from D'Alembert's principle

2,294 words10 min readBack to topic

Before you can read a single line of the parent derivation, you must be able to read its alphabet. This page builds every symbol from zero, in an order where each one only uses symbols already defined. Nothing here is assumed.


0. The stage: particles and positions

The picture. Plant an origin. Every particle is a dot; the arrow from the origin to that dot is . As the particle moves, the tip of the arrow traces its path.

Why the topic needs it. Newton's law talks about where things are and how that changes. is the raw "where." Everything else is built by watching how changes.

The picture. If is the position arrow, then is a small arrow tangent to the path showing the direction of travel; two dots () is the arrow showing how that travel-arrow itself is turning or stretching.


1. The reading-frame for changes: ,

Why the topic needs it. The Euler–Lagrange equation literally is : one wrapped around one , minus another . If you cannot feel the difference between these two operators, the equation is unreadable.


2. Combining vectors: the dot product

The picture. Shine a light straight down onto ; the shadow casts along , times the length of , is the dot product. Key fact we will lean on constantly:

  • If the two vectors are perpendicular (), then , so .

Why the topic needs it. The whole magic of D'Alembert's principle is that constraint forces point perpendicular to the allowed motion. Perpendicular means the dot product is zero, so those forces vanish from the equation. The dot product is the eraser.


3. Force, momentum, and Newton

Why the topic needs it. The entire derivation is a machine for making disappear so only (and inertia) remain. See Generalized forces and potentials.


4. Constraints and generalized coordinates

The picture. A bead on a bent wire lives in 3D, but you only need one number — how far along the wire it has slid — to pin it down. That number is . The wire is the constraint; is the freedom that survives.


5. Virtual displacement — the star of the show

The picture. Pause the movie. Ask: "In this frozen frame, which tiny nudges could I give the particle without breaking any constraint?" Those permitted nudges are the virtual displacements. For a bead on a wire, the only allowed points along the wire — never off it.

Why the topic needs it. Combine sections 2 and 5: constraint force allowed motion . Sum over particles and the constraint forces are gone. That single fact is D'Alembert's principle.


6. Generalized force and the potential


7. Putting the alphabet together

Here is how every symbol feeds the final Euler–Lagrange equation. Read top to bottom.

position r_k

velocity r_k dot

constraints and generalized coords q_i

virtual displacement delta r_k

dot product

constraint forces do zero virtual work

Newton F = p dot

DAlembert principle

kinetic energy T

split force part and inertia part

generalized force Q_i

product rule trick with d dt and partial q

Euler Lagrange equation

potential V

Every arrow is a dependency: you cannot understand a node until you understand its parents. This is exactly the order the parent derivation walks. Onward to Hamilton's principle (least action) and Noether's theorem once this is solid.


Equipment checklist

Cover the right side and see if you can answer each before revealing.

What does the arrow on tell you?
That is a vector — it carries a direction as well as a length, not just one number.
What does a single dot mean, and a double dot ?
Single dot = velocity (time derivative of position); double dot = acceleration (time derivative of velocity).
In words, what is ?
The change in when you wiggle only and freeze every other variable.
How is different from ?
counts every path through which real time affects the quantity (chain rule over all its variables); wiggles only the explicit time slot.
When is a dot product equal to zero?
When the two vectors are perpendicular, because .
Why does give the length squared?
The angle is , so and .
What is the difference between and ?
= applied forces you know and care about (gravity, springs); = constraint forces a rod/rail/wire silently supplies.
What makes a constraint holonomic?
You can write it as an equation relating positions (and possibly time), like .
Why use generalized coordinates instead of ?
They are a minimal set that automatically satisfies the constraints, so you can never describe an illegal configuration and constraint forces drop out.
What are the two defining properties of a virtual displacement ?
It is imagined and infinitesimal, taken with time frozen (), yet still consistent with the constraints at that instant.
Why freeze time in a virtual displacement?
So the shift always points along the instantaneously allowed direction, keeping it perpendicular to even when the constraint itself moves.
What does the generalized force collect?
Every applied force projected onto the direction that changing moves the system: .
For a conservative force, how does relate to ?
.