CALCULUS 2 • PARAMETRIC, POLAR & VECTOR FUNCTIONS

Second Derivatives: Parametric — Second Derivatives of Parametric Equations

Extend the chain rule to find concavity and curvature of curves defined by parametric equations.

Historical Context & Motivation

The study of curves defined by parametric equations has a rich history stretching back to the earliest days of calculus. When a curve cannot be conveniently expressed as y = f(x)—think of a loop, a spiral, or the trajectory of a projectile—mathematicians found it far more natural to describe both coordinates as functions of an independent parameter, typically time. The first derivative dy/dx captures the slope of such a curve at each parameter value, but understanding the curve's concavity and the rate at which the tangent direction changes demands a second derivative. The challenge is that differentiating once more with respect to the parameter t does not directly yield d²y/dx²; an additional application of the chain rule is required, and this subtlety was not fully systematized until the eighteenth century.

1637
Descartes and Coordinate Geometry
René Descartes publishes La Géométrie, establishing analytic geometry. Curves are expressed algebraically, but many important curves resist explicit y = f(x) representation.
1686
Newton's Parametric Fluxions
Isaac Newton introduces fluxions (derivatives) and treats curves parametrically using time as the independent variable, computing ẏ/ẋ for tangent slopes of planetary orbits.
1696
Leibniz and the Chain Rule
Gottfried Wilhelm Leibniz formalizes the differential notation dy/dx and articulates the chain rule, providing the algebraic machinery needed to relate parametric and Cartesian derivatives.
1748
Euler's Systematic Treatment
Leonhard Euler's Introductio in analysin infinitorum treats parametric curves comprehensively, including higher-order derivatives and curvature, laying the groundwork for modern differential geometry.
1850s
Curvature and Differential Geometry
The second parametric derivative becomes central to the Frenet–Serret formulas and the formal definition of curvature κ, connecting calculus to the geometry of space curves studied by Gauss and Riemann.

The essential question that motivates this topic is deceptively simple: if a curve is given by x = f(t) and y = g(t), how do we compute d²y/dx² in terms of f and g and their derivatives with respect to t? The answer requires more than naive differentiation; it demands careful application of the chain rule a second time, treating the first derivative dy/dx itself as a function of t and then dividing by dx/dt once more. Mastering this technique is essential for analyzing concavity, identifying inflection points, computing curvature, and solving optimization problems along parametric paths.

Core Principles & Definitions

Before computing the second derivative, it is essential to revisit the foundational ideas that govern differentiation of parametric curves. The following principles constitute the logical chain from parametric representation to the second-order rate of change d²y/dx².

1

Parametric Representation

A curve C is described by x = f(t) and y = g(t) where t ranges over some interval. The parameter t is the independent variable; x and y are both dependent on t.
2

First Derivative via Chain Rule

The slope dy/dx is obtained as (dy/dt) ÷ (dx/dt), provided dx/dt ≠ 0. This eliminates the parameter t from the derivative without eliminating it from the curve itself.
3

Second Derivative Is Not d²y/dt² ÷ d²x/dt²

A common misconception is that d²y/dx² equals g″(t)/f″(t). This is incorrect. The second derivative requires differentiating dy/dx (as a function of t) with respect to x.
4

Chain Rule Applied Twice

To find d²y/dx², compute d/dt(dy/dx) and then divide by dx/dt. This is equivalent to (d/dx)(dy/dx) via the chain rule: d/dx = (1/(dx/dt)) · d/dt.
5

Concavity Interpretation

When d²y/dx² > 0 the curve is concave up; when d²y/dx² < 0 it is concave down. Points where d²y/dx² changes sign are inflection points of the parametric curve.
KEY TAKEAWAY
Think of the first parametric derivative dy/dx as a speedometer reading at each moment t. The second derivative d²y/dx² tells you how that speedometer reading changes per unit of horizontal distance—it measures the curve's bending. Just as an engineer studies how a beam's curvature varies along its length to predict failure, the second parametric derivative reveals where a trajectory bows upward, bows downward, or straightens out. The key procedural insight is that you must differentiate the first derivative with respect to x, which means dividing by dx/dt—not by d²x/dt².

Visual Explanation

The following diagram illustrates the geometric meaning of the second parametric derivative. Consider a parametric curve traced out as t increases. At a given point, the first derivative dy/dx gives the slope of the tangent line. As the parameter advances, the tangent direction rotates. The second derivative d²y/dx² quantifies how rapidly that slope changes per unit horizontal displacement. Where d²y/dx² is positive, the curve bends upward (concave up); where it is negative, the curve bends downward (concave down). An inflection point occurs where the second derivative transitions through zero and changes sign.

A parametric curve traced from t = 0 to t = 4. The violet curve transitions from concave down (d²y/dx² < 0) to concave up (d²y/dx² > 0) at the amber inflection point. Tangent lines are shown at two locations to illustrate the changing slope.

Notice in the diagram that the tangent line at t₁ slopes downward and the curve is bending to the right and downward—classic concave-down behavior. By the time the curve reaches t₃, the tangent slopes upward and the curve is bending upward—concave-up behavior. The transition between these two regimes is the inflection point, where d²y/dx² = 0 and the sign changes. Identifying these features on a parametric curve follows the same logical steps as for explicit functions, but the computation involves the parametric second-derivative formula rather than a simple d²y/dx² = f″(x).

Mathematical Framework

We now derive the formula for d²y/dx² in terms of the parameter t. Let x = f(t) and y = g(t) be differentiable functions of t, and assume f′(t) ≠ 0 on the interval of interest so that x is a monotonic function of t and can serve as the effective independent variable.

FIRST PARAMETRIC DERIVATIVE
dy/dx = (dy/dt) / (dx/dt) = g′(t) / f′(t)
This follows directly from the chain rule: dy/dx = (dy/dt) · (dt/dx) = g′(t) · (1/f′(t)), valid when f′(t) ≠ 0.

To find the second derivative, we differentiate dy/dx with respect to x. However, dy/dx is expressed as a function of t, so we apply the chain rule once more. Let w(t) = dy/dx = g′(t)/f′(t). Then:

SECOND PARAMETRIC DERIVATIVE
d²y/dx² = (d/dx)(dy/dx) = (dw/dt) / (dx/dt) = [d/dt(dy/dx)] / (dx/dt)
Here w = dy/dx is treated as a function of t. We differentiate w with respect to t and divide by dx/dt. This is the central formula of this lesson.

Expanding dw/dt using the quotient rule on w = g′(t)/f′(t) yields a more explicit form.

EXPANDED FORM
d²y/dx² = [f′(t) · g″(t) − g′(t) · f″(t)] / [f′(t)]³
Derivation: dw/dt = [f′(t)g″(t) − g′(t)f″(t)] / [f′(t)]². Dividing by dx/dt = f′(t) yields the cube in the denominator. This formula is valid whenever f′(t) ≠ 0.
⚠️ Common Mistake
Students frequently attempt to compute d²y/dx² as g″(t)/f″(t) or as d²y/dt² ÷ d²x/dt². Neither of these is correct. The second derivative with respect to x requires the chain rule: you must first differentiate dy/dx = g′(t)/f′(t) with respect to t, and then divide by dx/dt. Forgetting this final division is the most common source of error.
COMPACT NOTATION
d²y/dx² = (ẋÿ − ẏẍ) / ẋ³ [Newton's dot notation]
Here ẋ = dx/dt, ẍ = d²x/dt², ẏ = dy/dt, ÿ = d²y/dt². Some texts use primes: x′, x″, y′, y″. The formula is d²y/dx² = (x′y″ − y′x″) / (x′)³.

Step-by-Step Derivation & Classification of Concavity

Let us walk through the derivation in full detail to ensure every algebraic step is transparent. We also classify the concavity conditions and connect d²y/dx² to curvature, which provides a parametrization-independent measure of bending.

The four-step algorithmic flow from parametric functions to d²y/dx². Start with the parametric equations (violet), compute first derivatives (cyan), form dy/dx (pink), differentiate w = dy/dx with respect to t using the quotient rule (orange), and finally divide by dx/dt to obtain the second parametric derivative (emerald).
Classification of concavity for parametric curves
ConditionConcavityGeometric Meaning
d²y/dx² > 0Concave UpCurve opens upward; tangent line lies below the curve locally.
d²y/dx² < 0Concave DownCurve opens downward; tangent line lies above the curve locally.
d²y/dx² = 0 (sign change)Inflection PointTransition between concave up and concave down; curvature passes through zero.
d²y/dx² = 0 (no sign change)Higher-order flatnessNot an inflection point; analogous to f″(x) = 0 at x⁴ type behavior.

The connection to curvature κ is immediate. The signed curvature of a parametric curve is given by κ = (x′y″ − y′x″) / (x′² + y′²)3/2. The numerator of κ is identical to the numerator of d²y/dx², but the denominators differ: d²y/dx² uses (x′)³ while κ uses (x′² + y′²)3/2, which accounts for the full arc-length speed. As a result, d²y/dx² and κ need not share the same zeros or nonzero values: a curve can have d²y/dx² ≠ 0 yet κ = 0 (if the numerator x′y″ − y′x″ is zero but x′ is not), and more generally, the magnitude of each quantity depends on the respective denominator. What the shared numerator does guarantee is that the two quantities vanish simultaneously—so d²y/dx² = 0 if and only if κ = 0 (when x′ ≠ 0)—confirming that the second parametric derivative correctly identifies the curve's inflection points.

Worked Example

Let us compute d²y/dx² for the parametric curve x = t² − 1, y = t³ − 3t at t = 2. We will also determine the concavity of the curve at that point.

Finding d²y/dx² for x = t² − 1, y = t³ − 3t
1
Step 1 — Compute First Derivatives with Respect to tDifferentiate x and y with respect to t: dx/dt = 2t and dy/dt = 3t² − 3.
dx/dt = 2t, dy/dt = 3t² − 3
2
Step 2 — Form the First Derivative dy/dxDivide dy/dt by dx/dt: dy/dx = (3t² − 3) / (2t) = (3(t² − 1)) / (2t). This is valid for t ≠ 0.
dy/dx = (3t² − 3) / (2t)
3
Step 3 — Differentiate dy/dx with Respect to tLet w(t) = (3t² − 3) / (2t). Apply the quotient rule: dw/dt = [(2t)(6t) − (3t² − 3)(2)] / (2t)². Simplify the numerator: 12t² − 6t² + 6 = 6t² + 6. So dw/dt = (6t² + 6) / (4t²) = (6(t² + 1)) / (4t²) = (3(t² + 1)) / (2t²).
d/dt(dy/dx) = 3(t² + 1) / (2t²)
4
Step 4 — Divide by dx/dt to Get d²y/dx²d²y/dx² = [d/dt(dy/dx)] / (dx/dt) = [3(t² + 1) / (2t²)] / (2t) = 3(t² + 1) / (4t³).
d²y/dx² = 3(t² + 1) / (4t³)
5
Step 5 — Evaluate at t = 2Substitute t = 2: d²y/dx² = 3(4 + 1) / (4 × 8) = 15/32 ≈ 0.469. Since d²y/dx² > 0, the curve is concave up at t = 2.
d²y/dx² at t = 2 equals 15/32 (concave up)
6
Step 6 — Verify Using the Expanded FormulaCross-check with d²y/dx² = [f′g″ − g′f″] / (f′)³. Here f′ = 2t, f″ = 2, g′ = 3t² − 3, g″ = 6t. At t = 2: numerator = (4)(12) − (9)(2) = 48 − 18 = 30; denominator = 4³ = 64. So d²y/dx² = 30/64 = 15/32. ✓ Matches Step 5.
Confirmed: 15/32 ✓

Common Pitfalls & Method Comparisons

Students and practitioners working with parametric second derivatives often encounter predictable difficulties. The table below contrasts the correct approach with common errors and compares the parametric method to the standard Cartesian approach.

Comparison of correct techniques and common errors
Issue / MethodCorrect ApproachCommon Error / Limitation
Computing d²y/dx²Differentiate dy/dx with respect to t, then divide by dx/dt.Taking g″(t)/f″(t) or (d²y/dt²)/(d²x/dt²)—both are wrong.
Handling dx/dt = 0Recognize vertical tangent; d²y/dx² is undefined there. Analyze the curve's behavior separately.Plugging in values blindly, producing division by zero without comment.
Quotient rule in Step 3Apply quotient rule carefully to g′(t)/f′(t), simplify, then divide by f′(t).Algebraic errors in the quotient rule, especially sign mistakes in the numerator.
Parametric vs. CartesianParametric: use the formula. Cartesian y = f(x): simply compute f″(x).Eliminating the parameter t first may introduce complications (e.g., multi-valued functions).
Inflection point detectionSet d²y/dx² = 0, verify sign change in d²y/dx² across that t value.Assuming d²y/dx² = 0 is sufficient without checking sign change.
💡 WHY NOT ELIMINATE THE PARAMETER?
For some curves—like x = t², y = t³—you could write y = x3/2 and compute d²y/dx² directly. But many parametric curves (ellipses, cycloids, Lissajous figures) do not yield a single-valued Cartesian equation. The parametric formula is universal: it works regardless of whether the curve can be expressed as y = f(x). Think of it as analyzing the steering of a car by reading the instruments on the dashboard (the parameter t) rather than pulling over, getting out, and measuring the tire tracks on the road.

Connection to Curvature & Higher Derivatives

The second parametric derivative is not an isolated technique—it opens the door to deeper analysis of curve geometry. The most important generalization is curvature, which measures bending in a way that is independent of how fast the curve is traversed. Beyond curvature, the same differentiation scheme extends to third and higher parametric derivatives, which appear in the study of torsion for space curves and in jerk analysis in physics.

Connections between the second parametric derivative and advanced curve-geometry concepts
ConceptFormulaRelationship to d²y/dx²
Curvature κκ = |x′y″ − y′x″| / (x′² + y′²)^(3/2)Numerator matches d²y/dx² numerator; denominator uses arc-length speed instead of (dx/dt)³.
Radius of curvature RR = 1/κLarge |d²y/dx²| implies large κ and small R (tight bending).
Third parametric derivative d³y/dx³d³y/dx³ = [d/dt(d²y/dx²)] / (dx/dt)Same chain-rule pattern: differentiate the previous result with respect to t, divide by dx/dt.
Arc length element dsds/dt = √(x′² + y′²)Curvature re-expresses d²y/dx² in terms of arc-length parametrization.

Notice the recursive pattern: the n-th derivative of y with respect to x, expressed parametrically, is always obtained by differentiating the (n − 1)-th derivative with respect to t and dividing by dx/dt. This elegant structure underpins the Faà di Bruno formula for higher-order chain-rule computations, which generalizes the pattern we have seen here. In differential geometry, these ideas extend naturally to three-dimensional space curves via the Frenet–Serret frame, where the second derivative contributes to the curvature vector and the third derivative introduces torsion, measuring how the curve departs from planarity.

Practice Problems

PROBLEM 1CONCEPTUAL
Explain why d²y/dx² for a parametric curve is not equal to (d²y/dt²) / (d²x/dt²). What is the correct relationship, and why does the chain rule require an extra step?
PROBLEM 2BASIC CALCULATION
Given x = 3t, y = t² + 1, find d²y/dx² and determine the concavity of the curve for all t.
PROBLEM 3INTERMEDIATE
For the curve x = cos(t), y = sin(t) (the unit circle), compute d²y/dx² as a function of t. At what values of t is d²y/dx² undefined, and what happens geometrically at those points?
PROBLEM 4APPLIED
A projectile follows the path x = 20t, y = 30t − 4.9t² (distance in meters, t in seconds). Find d²y/dx² and interpret it physically. At what time t does the projectile reach an inflection point, if any?
PROBLEM 5CRITICAL THINKING
Consider the parametric curve x = t³, y = t⁶. (a) Compute d²y/dx² using the parametric formula. (b) Eliminate the parameter to write y as a function of x, then compute d²y/dx² directly. (c) Reconcile any apparent discrepancies and discuss what happens at t = 0.

Lesson Summary

The second parametric derivative d²y/dx² for a curve defined by x = f(t) and y = g(t) is computed by first forming the first derivative dy/dx = g′(t)/f′(t), then differentiating this quotient with respect to t, and finally dividing by dx/dt. The resulting formula is d²y/dx² = [f′(t)g″(t) − g′(t)f″(t)] / [f′(t)]³, valid whenever f′(t) ≠ 0. This formula is not the same as g″(t)/f″(t); the chain rule must be applied carefully through the quotient rule and an additional division by dx/dt.

The sign of d²y/dx² determines concavity: positive for concave up, negative for concave down, with inflection points occurring where d²y/dx² changes sign. This technique connects directly to curvature κ = |x′y″ − y′x″|/(x′² + y′²)^(3/2), and the recursive differentiation pattern (differentiate by t, divide by dx/dt) extends to higher-order derivatives, underpinning the analysis of space curves and motion in differential geometry and physics.

Varsity Tutors • Calculus 2 • Second Derivatives of Parametric Equations