Time derivatives of vector expressions

Time derivatives of vector expressions describe how vector-valued functions change as time elapses. The key result: differentiating a vector requires accounting for changes in both its magnitude and its direction — and when expressed in a rotating frame, an extra term

ω \times r

appears. Applies throughout kinematics, dynamics, electromagnetism, and orbital mechanics. Formalized by Euler, Coriolis, and Gibbs in the 18th–19th centuries.

A vector expression is any quantity of the form $v (t) = v_{1} (t) \hat{e}_{1} + v_{2} (t) \hat{e}_{2} + v_{3} (t) \hat{e}_{3}$ , where both the scalar components and possibly the basis vectors depend on time. Its time derivative is the limit of the difference quotient $lim_{Δ t \to 0} [v (t + Δ t) - v (t)] /Δ t$ , evaluated componentwise in a fixed basis. This matters because nearly every law of motion is a statement about a time derivative of a vector — velocity, acceleration, momentum, force. The surprising fact: a vector of constant magnitude generally has a nonzero derivative, and that derivative is always perpendicular to the vector itself. This single observation underpins centripetal acceleration, gyroscopic precession, and the Coriolis effect.

The Core Insight

A vector changes in two independent ways: it can stretch or shrink (change in magnitude) and it can turn (change in direction). The time derivative captures both. When the basis vectors themselves rotate — as in a non-inertial frame — their derivatives are nonzero and contribute extra terms. The master tool linking the two pictures is the transport theorem: the rate of change measured in a fixed frame equals the rate of change measured in a rotating frame plus $ω \times (the vector)$ .

Formal Statement

Time derivative of a vector function

Let $v : R \to R^{3}$ be a vector-valued function of time $t$ . In a fixed orthonormal basis ${\hat{e}_{1}, \hat{e}_{2}, \hat{e}_{3}}$ , write $v (t) = \sum_{i} v_{i} (t) \hat{e}_{i}$ . The time derivative is

\frac{d v}{d t} = Δ t \to 0 lim \frac{v ( t + Δ t ) - v ( t )}{Δ t} = i = 1 \sum 3 \frac{d v _{i}}{d t} \hat{e}_{i} .

⟳Because

\hat{e}_{i}

is constant in a fixed frame, the limit acts on each scalar component separately; linearity of the limit pulls it through the sum.

Here $v_{i} (t)$ are the components (scalars) and $\hat{e}_{i}$ are the basis vectors (unit, mutually orthogonal). The result $d v / d t$ is again a vector.

The product rules generalize directly. For a scalar function $f (t)$ and vectors $u (t), v (t)$ :

\frac{d}{d t} (f u) = \frac{df}{d t} u + f \frac{d u}{d t} .

⟳Apply the ordinary product rule to each component

f u_{i}

and resum.

\frac{d}{d t} (u \cdot v) = \frac{d u}{d t} \cdot v + u \cdot \frac{d v}{d t} .

⟳Expand the dot product as

\sum_{i} u_{i} v_{i}

and apply the scalar product rule term by term.

\frac{d}{d t} (u \times v) = \frac{d u}{d t} \times v + u \times \frac{d v}{d t} .

⟳Write the cross product with the Levi-Civita symbol

(u \times v)_{k} = ϵ_{ij k} u_{i} v_{j}

; differentiate the product

u_{i} v_{j}

— order must be preserved since the cross product is non-commutative.

Order matters in the cross-product rule

Unlike the dot product, the cross product is anti-commutative, so the two terms in $[4]$ may not be swapped: $\frac{d u}{d t} \times v \neq = v \times \frac{d u}{d t}$ . Always keep each factor in its original left/right position.

Derivative of a unit vector / constant-magnitude vector

If $∣ v (t) ∣$ is constant, then $d v / d t$ is perpendicular to $v$ at every instant:

∣ v ∣ = const ⟹ v \cdot \frac{d v}{d t} = 0.

⟳Differentiate

v \cdot v = ∣ v ∣^{2} = const

using

[3]

; the left side gives

2 v \cdot d v / d t

, the right gives

0

Transport theorem / rotating-frame rule

Let frame $S^{'}$ rotate with angular velocity $ω (t)$ relative to a fixed frame $S$ . For any vector $A$ ,

(\frac{d A}{d t})_{S} = (\frac{d A}{d t})_{S^{'}} + ω \times A .

⟳Write

A = \sum_{i} A_{i} \hat{e}_{i}^{'}

with rotating basis vectors. Differentiate; the component terms give

(d A / d t)_{S^{'}}

, while

d \hat{e}_{i}^{'} / d t = ω \times \hat{e}_{i}^{'}

generates the cross-product term.

Why It Holds

Componentwise rule and the perpendicularity theorem

Step 1 — Componentwise differentiation. In a fixed frame the basis vectors satisfy

d \hat{e}_{i} / d t = 0

. Apply the limit definition to

v = \sum_{i} v_{i} \hat{e}_{i}

\frac{v ( t + Δ t ) - v ( t )}{Δ t} = i \sum \frac{v _{i} ( t + Δ t ) - v _{i} ( t )}{Δ t} \hat{e}_{i} .

Taking

Δ t \to 0

and using continuity of the basis yields

[1]

. Step 2 — Product rule for the dot product. Write

u \cdot v = \sum_{i} u_{i} v_{i}

. By the scalar product rule,

\frac{d}{d t} i \sum u_{i} v_{i} = i \sum (\overset{u}{˙}_{i} v_{i} + u_{i} \overset{v}{˙}_{i}) = \dot{u} \cdot v + u \cdot \dot{v},

which is

[3]

. Step 3 — Perpendicularity. Set

u = v

[3]

\frac{d}{d t} (v \cdot v) = 2 v \cdot \dot{v} .

∣ v ∣

is constant, the left side is

\frac{d}{d t} ∣ v ∣^{2} = 0

, forcing

v \cdot \dot{v} = 0

. Since neither vector is zero (generically), they are orthogonal. This proves

[5]

■

Transport theorem

Step 1. Express

A

in the rotating basis:

A = \sum_{i} A_{i} \hat{e}_{i}^{'}

. Step 2. Differentiate as seen from the fixed frame

S

, applying the product rule

[2]

to each term:

(\frac{d A}{d t})_{S} = i \sum \frac{d A _{i}}{d t} \hat{e}_{i}^{'} + i \sum A_{i} \frac{d e ^ _{i}^{'}}{d t} .

Step 3. The first sum is, by definition, the rate of change observed in

S^{'}

(where the basis is held fixed):

(d A / d t)_{S^{'}} = \sum_{i} \dot{A}_{i} \hat{e}_{i}^{'}

. Step 4. Each rotating basis vector has constant magnitude, so by Euler's rotation kinematics

d \hat{e}_{i}^{'} / d t = ω \times \hat{e}_{i}^{'}

. Therefore

i \sum A_{i} \frac{d e ^ _{i}^{'}}{d t} = ω \times i \sum A_{i} \hat{e}_{i}^{'} = ω \times A .

Combining Steps 3 and 4 gives $[6]$ . $■$

Gaspard-Gustave de Coriolis | 1792–1843

A French engineer-mathematician, Coriolis published Du calcul de l'effet des machines (1829) and his decisive 1835 paper on the equations of relative motion in rotating systems. By carefully tracking the time derivatives of position vectors in a rotating frame, he isolated the term $2 ω \times v_{rel}$ that now bears his name. His work made the transport theorem $[6]$ an operational tool for engineering and meteorology, transforming abstract vector calculus into the language of weather systems and rotating machinery.

Building Intuition

Picture a vector as an arrow you are holding. There are exactly two things you can do to change it: stretch it (change its length) or rotate it (change its direction). The derivative $d v / d t$ is the velocity of the arrow's tip. Decompose that tip-velocity into a part along the arrow (stretching) and a part perpendicular to it (turning). Theorem $[5]$ says: if you forbid stretching (constant length), the tip can only move sideways — so the derivative is purely perpendicular. This is exactly why a planet in a circular orbit, moving at constant speed, still accelerates: its velocity vector keeps the same length but constantly turns, and that turning is the centripetal acceleration pointing inward.

A constant-magnitude vector and its perpendicular derivative

Worked Examples

Velocity and acceleration in uniform circular motion

Given: A particle moves on a circle of radius

R = 2.0 m

with constant angular speed

ω = 3.0 rad/s

. Position vector:

r (t) = R cos (ω t) \hat{e}_{x} + R sin (ω t) \hat{e}_{y} .

Step 1 — Velocity via

[1]

v = \frac{d r}{d t} = - R ω sin (ω t) \hat{e}_{x} + R ω cos (ω t) \hat{e}_{y} .

Speed:

∣ v ∣ = R ω = 2.0 \times 3.0 = 6.0 m/s

(constant). Step 2 — Check perpendicularity via

[5]

r \cdot v = - R^{2} ω cos sin + R^{2} ω sin cos = 0. ✓

Step 3 — Acceleration:

a = \frac{d v}{d t} = - R ω^{2} cos (ω t) \hat{e}_{x} - R ω^{2} sin (ω t) \hat{e}_{y} = - ω^{2} r .

Magnitude:

∣ a ∣ = R ω^{2} = 2.0 \times (3.0)^{2} = 18 m/s^{2}

, directed inward.

∣ v ∣ = 6.0 m/s, ∣ a ∣ = 18 m/s^{2} (centripetal)

Applying the transport theorem

Given: A bead's position in a frame rotating about the

z

-axis with

ω = 2.0 \hat{e}_{z} rad/s

is fixed in that frame at

A = 3.0 \hat{e}_{x}^{'} m

(so

(d A / d t)_{S^{'}} = 0

). Find its velocity in the fixed frame. Step 1 — Transport theorem $[6]$ :

(\frac{d A}{d t})_{S} = 0 + ω \times A .

Step 2 — Compute the cross product:

ω \times A = (2.0 \hat{e}_{z}) \times (3.0 \hat{e}_{x}^{'}) = 6.0 \hat{e}_{y}^{'} m/s,

using

\hat{e}_{z} \times \hat{e}_{x}^{'} = \hat{e}_{y}^{'}

at the instant the frames align. Step 3 — Interpretation: Although the bead is stationary in the rotating frame, it moves at

6.0 m/s

tangentially in the fixed frame — pure rotational carry-along.

(\frac{d A}{d t})_{S} = 6.0 \hat{e}_{y}^{'} m/s

Common Misconceptions

Constant magnitude ≠ constant vector

A common error is assuming $∣ v ∣ = const \Rightarrow d v / d t = 0$ . This is false: only the length is fixed; the direction may still change, giving a nonzero perpendicular derivative (Theorem $[5]$ ). The classic trap is concluding that uniform circular motion has zero acceleration "because the speed doesn't change."

A unit vector with nonzero derivative

Take $\hat{u} (t) = cos t \hat{e}_{x} + sin t \hat{e}_{y}$ . Its magnitude is $∣ \hat{u} ∣ = cos^{2} t + sin^{2} t = 1$ for all $t$ — perfectly constant. Yet

\frac{d u ^}{d t} = - sin t \hat{e}_{x} + cos t \hat{e}_{y} \neq = 0,

with

∣ d \hat{u} / d t ∣ = 1

and

\hat{u} \cdot d \hat{u} / d t = - cos t sin t + sin t cos t = 0

. A vector can have a constant length and a derivative of equal length — perpendicular to it.

The derivative of a basis vector is not always zero

In a fixed Cartesian frame, $d \hat{e}_{i} / d t = 0$ , which justifies componentwise differentiation in $[1]$ . But in curvilinear (polar, cylindrical, spherical) or rotating frames, the basis vectors change direction with the particle's motion, so $d \hat{e}_{r} / d t \neq = 0$ . Forgetting these extra terms is the single most common source of error in deriving acceleration in polar coordinates.

Connections

Time derivatives of vectors are the computational backbone of Kinematics (velocity and acceleration as derivatives of position) and Newtonian Mechanics ( $F = d p / d t$ ). The perpendicularity theorem $[5]$ explains the Frenet–Serret Formulas, where the derivatives of the unit tangent, normal, and binormal vectors are mutually orthogonal. The transport theorem $[6]$ generalizes into Rotating Reference Frames, producing the Coriolis Effect and centrifugal terms. The product rules $[2]$ – $[4]$ are special cases of differentiation in Vector Calculus, and the perpendicular component of $d v / d t$ connects to Angular Velocity via $\dot{v} = ω \times v$ for rigid rotation.

Real-World Applications

Orbital mechanics (NASA/ESA navigation): Spacecraft trajectories are integrated from $\ddot{r} = - μ r / r^{3}$ , a second time derivative of the position vector.
Meteorology — the Coriolis effect: Large-scale wind and ocean currents (e.g., the Gulf Stream, cyclone rotation) are governed by the $2 ω \times v$ term arising from $[6]$ in Earth's rotating frame.
Gyroscopes and inertial navigation systems (INS): Aircraft, submarines, and missiles compute orientation by integrating $d L / d t = τ$ , the time derivative of angular momentum.
Robotics and manipulator dynamics: Each joint frame rotates relative to the next; end-effector velocity is built by chaining transport-theorem terms across links.
Electromagnetism — the Lorentz force: A charge's motion obeys $m d v / d t = q (E + v \times B)$ , requiring time differentiation of $v$ .
Computer graphics / game physics: Rigid-body engines (Havok, PhysX) integrate orientation quaternions and angular velocity using discretized versions of $[6]$ .
Satellite attitude control: Reaction wheels and gyrostabilizers exploit $d L / d t$ to reorient spacecraft without expelling propellant.

Historical Development

1687 — Isaac Newton formulates the laws of motion in Principia, implicitly treating force as the time rate of change of momentum, though using geometric rather than vector language.
1749–1765 — Leonhard Euler develops rigid-body rotation kinematics, establishing $d \hat{e} / d t = ω \times \hat{e}$ , the seed of the transport theorem.
1788 — Joseph-Louis Lagrange's Mécanique analytique recasts dynamics analytically, foregrounding derivatives of coordinate functions.
1835 — Gaspard-Gustave de Coriolis publishes his analysis of relative motion in rotating frames, isolating the Coriolis term from differentiating position vectors.
1843 — William Rowan Hamilton invents quaternions, providing the first algebra that handles vector rotation and its derivatives natively.
1880s — Josiah Willard Gibbs and Oliver Heaviside develop modern vector analysis, formalizing the dot- and cross-product derivative rules $[3]$ – $[4]$ .
1901 — Gibbs and E. B. Wilson's textbook Vector Analysis standardizes the notation $d v / d t$ used today.
1910s–1920s — Vector and tensor time-derivative methods become standard in dynamics texts and feed into relativity and continuum mechanics.

Subtleties & Limits

The componentwise rule $[1]$ presumes a fixed basis; the deep subtlety is that "the time derivative of a vector" is frame-dependent — the transport theorem $[6]$ shows two observers in relative rotation disagree about $d A / d t$ even for the same physical vector. In curved spacetime (general relativity) and on curved manifolds, ordinary differentiation fails entirely because there is no canonical way to compare vectors at different points; one must introduce the covariant derivative and parallel transport, with the rate of change depending on the chosen connection. The practical open question for numerical work is how to integrate rotational kinematics ( $[6]$ ) over long times without accumulating drift — addressed by quaternion-based and Lie-group integrators that preserve the constant-magnitude constraint of $[5]$ exactly.