Chain Rule | Tony's Intertwingled Memex

Core Definition

The Chain Rule computes the derivative of a composite function. If $y = f(g(x))$ , then: Prime Notation:

\frac{d}{dx}[f(g(x))] = f'(g(x)) \cdot g'(x)

Leibniz Notation: Letting

u = g(x)

and

y = f(u)

\frac{dy}{dx} = \frac{dy}{du} \cdot \frac{du}{dx}

For a three-layer composite function

y = f(g(h(x)))

with

u = g(v)

and

v = h(x)

\frac{dy}{dx} = \frac{dy}{du} \cdot \frac{du}{dv} \cdot \frac{dv}{dx}

Concept Explanation

The Chain Rule describes how rates of change propagate through nested dependent systems. If variable $y$ changes $A$ times faster than $u$ , and $u$ changes $B$ times faster than $x$ , then the total sensitivity of $y$ relative to $x$ is the multiplicative product $A \times B$ .

The "Peeling the Onion" Mental Model

When evaluating a composite derivative, differentiate from the outside inward:

Differentiate the outer shell, preserving the interior argument completely intact. Never differentiate the inside function on this first step.
Multiply by the derivative of the inner argument.
For multi-layer systems ( $f(g(h(x)))$ ), repeat the process step-by-step until the innermost variable ( $x$ ) is reached.

Leibniz Notation vs. Prime Notation

Operational Feature	Leibniz Notation ( $\frac{dy}{du}\cdot\frac{du}{dx}$ )	Prime Notation ( $f'(g(x))\cdot g'(x)$ )
Primary Strength	Explicitly tracks rates with respect to physical units and intermediate variables.	Fast, compact shorthand for algebraic calculations on paper.
Variable Tracking	Introduces named intermediate variables ( $u, v, z$ ).	Tracks nested positions directly without auxiliary variables.
Engineering Context	Preferred in physics, thermodynamics, and multi-stage control loops.	Preferred in single-variable algebraic transformations.

Engineering Applications

Electrical Engineering (Thermal Drift in Timing Circuits): When capacitor voltage decays over time $V(t) = V_0 e^{-t/\tau}$ while clock speed varies with temperature $t(\theta)$ , the temperature sensitivity is governed by $\frac{dV}{d\theta} = \frac{dV}{dt} \cdot \frac{dt}{d\theta}$ (see RC Circuit Transient Surge (Electrical Engineering)).
Aerospace Engineering (Atmospheric Drag Force): Aerodynamic drag force $F_d$ depends on atmospheric air density $\rho$ , density depends on altitude $z$ , and altitude depends on burn-time $t$ . The time rate of change of drag force requires a three-layer chain: $\frac{dF_d}{dt} = \frac{dF_d}{d\rho} \cdot \frac{d\rho}{dz} \cdot \frac{dz}{dt}$ (see Atmospheric Density and Aerodynamic Drag (Aerospace Engineering)).
Robotics & Actuator Kinematics: Converting linear Cartesian actuator positions to rotational joint angles requires chain-differentiating inverse trigonometric functions (see Robotic Kinematics & Angular Velocity Mapping (Mechanical Engineering)).

Example Problems

Example 1: Two-Layer Leibniz Notation

Find the derivative of $y = \frac{1}{\sqrt{1 + x^5}}$ with respect to $x$ using Leibniz notation.

Solution

Step 1: Pre-process with negative and rational exponents Rewrite the function in power form:

y = (1 + x^5)^{-\frac{1}{2}}

Step 2: Define the intermediate variable $u$ Let $u = 1 + x^5$ , so $y = u^{-\frac{1}{2}}$ .

Step 3: Differentiate both components Using the Power Rule:

\frac{dy}{du} = -\frac{1}{2}u^{-\frac{3}{2}}, \qquad \frac{du}{dx} = 5x^4

Step 4: Combine via the Chain Rule

\frac{dy}{dx} = \frac{dy}{du} \cdot \frac{du}{dx} = \left(-\frac{1}{2}u^{-\frac{3}{2}}\right)(5x^4) = -\frac{5x^4}{2}u^{-\frac{3}{2}}

Step 5: Back-substitute $u$ and post-process into radical form

\frac{dy}{dx} = -\frac{5x^4}{2(1 + x^5)^{\frac{3}{2}}} = -\frac{5x^4}{2\sqrt{(1 + x^5)^3}}

Example 2: Three-Layer Composite Function (Prime Notation)

Find the derivative of $h(x) = \cos^4(3x^2 - 1)$ with respect to $x$ .

Solution

Step 1: Identify the nested structural layers

Outer layer: $[u]^4$
Middle layer: $\cos(v)$
Inner layer: $v = 3x^2 - 1$

Step 2: Differentiate the outer power layer By the Power Rule, bring down the exponent 4, reduce the power to 3, and keep the entire interior untouched:

h'(x) = 4\cos^3(3x^2 - 1) \cdot \frac{d}{dx}[\cos(3x^2 - 1)]

Step 3: Differentiate the middle trigonometric layer The derivative of $\cos(v)$ is $-\sin(v)$ , leaving the innermost argument untouched:

h'(x) = 4\cos^3(3x^2 - 1) \cdot \left(-\sin(3x^2 - 1)\right) \cdot \frac{d}{dx}[3x^2 - 1]

Step 4: Differentiate the inner polynomial layer

\frac{d}{dx}[3x^2 - 1] = 6x

Step 5: Multiply all factors and simplify coefficients

h'(x) = 4 \cdot (-\sin(3x^2 - 1)) \cdot 6x \cdot \cos^3(3x^2 - 1) = -24x \cos^3(3x^2 - 1)\sin(3x^2 - 1)

Connections & References

Parent Concepts: Derivatives
Prerequisites: Power Rule, Negative and Rational Exponents, Logarithmic and Exponential Properties
Transcendental Implementations: Derivatives of Exponential Functions, Derivatives of Logarithmic Functions, Derivatives of Inverse Trigonometric Functions, Derivatives of Trigonometric Functions
Advanced Extensions: Implicit Differentiation, Related Rates
Course Sessions: Calc-1 Session 08, Calc-1 Session 09