2.4, 3.4 - Differentiating parametric curves

2.4, 3.4 - Differentiating parametric curves

A curve can be described by giving both coordinates in terms of a third variable. You will follow such a curve, turn its two parameter derivatives into a gradient, and find tangents and normals without first finding an equation of the form y=f(x)y=f(x).

A curve given by a parameter

How can one curve contain both (4,2)(4,2) and (4,−2)(4,-2), yet still be easy to describe? Let x=t2x=t^2 and y=ty=t. The inputs t=2t=2 and t=−2t=-2 give those two points. We have described both branches using one moving parameter.

Parametric equations

Parametric equations give the coordinates of a point using a common parameter, such as x=f(t)x=f(t), y=g(t)y=g(t). Each permitted value of tt supplies one ordered pair (x,y)(x,y). The parameter need not be time.

Use the same parameter value in both equations. Changing tt traces the curve, and the permitted parameter interval determines which part is traced.

Worked example 1: following a sideways parabola

Question. For x=t2x=t^2, y=ty=t, find the points corresponding to t=−2,0,2t=-2,0,2, and find a Cartesian equation.

Plan. Substitute each value into both coordinates, then eliminate tt using t=yt=y.

t−202x404y−202\begin{array}{c|rrr} t&-2&0&2\\\hline x&4&0&4\\ y&-2&0&2 \end{array}

Since t=yt=y, the equation x=t2x=t^2 becomes x=y2x=y^2.

Answer. The points are (4,−2)(4,-2), (0,0)(0,0), (4,2)(4,2); the curve is x=y2x=y^2, with x≥0x\ge0.

Check. Each point satisfies x=y2x=y^2. As tt increases through these values, the point moves towards the vertex along the lower branch, then away along the upper branch.

Shift only the yy-equation up by 11. The parameter substitution is unchanged, but elimination now uses t=y−1t=y-1.

Worked example 2: a shifted curve with a restricted parameter

Question. For x=t2x=t^2, y=t+1y=t+1, where −2≤t≤2-2\le t\le2, find the endpoints and a Cartesian equation with the restriction that selects the traced arc.

Plan. Substitute the parameter endpoints, then express tt in terms of yy.

t=−2⟹(x,y)=(4,−1)t=2⟹(x,y)=(4,3)t=y−1⟹x=(y−1)2\begin{aligned} t=-2&\Longrightarrow(x,y)=(4,-1)\\ t=2&\Longrightarrow(x,y)=(4,3)\\ t=y-1&\Longrightarrow x=(y-1)^2 \end{aligned}

The parameter restriction becomes −1≤y≤3-1\le y\le3.

Answer. Endpoints (4,−1)(4,-1), (4,3)(4,3); x=(y−1)2x=(y-1)^2, with −1≤y≤3-1\le y\le3.

Check. At t=0t=0, the point is (0,1)(0,1), the shifted vertex. Both endpoints satisfy the Cartesian equation and the stated restriction.

The parameter interval matters: the Cartesian equation alone would include the rest of the parabola too.

[DIAGRAM: asset_name: A parameter follows a curve; asset_slug: parametric-parabola; caption: Increasing t traces both branches, and t equals zero gives a vertical tangent; description: Coordinate plot of x equals t squared and y equals t, from t equals minus two to two. Points are labelled t equals minus two at four comma minus two, t equals zero at the origin, t equals one at one comma one, and t equals two at four comma two. Arrows show increasing t. A vertical tangent at the origin and the tangent of gradient one half at one comma one are drawn.]
Increasing t traces both branches, and t equals zero gives a vertical tangent

Watch out

x=t2x=t^2 does not imply t=xt=\sqrt x for every real tt. Negative tt gives the other branch. Use the other coordinate or the parameter interval before choosing a square-root sign.

Follow a new upward shift and interval.

The parametric gradient

A gradient measures vertical change per horizontal change, rather than either change per unit parameter. If a small parameter change gives Δx\Delta x and Δy\Delta y, then Δy/Δx=(Δy/Δt)/(Δx/Δt)\Delta y/\Delta x=(\Delta y/\Delta t)/(\Delta x/\Delta t). Taking limits gives the parametric gradient when the horizontal rate is nonzero.

The parametric gradient

For differentiable x=f(t)x=f(t), y=g(t)y=g(t), at a parameter value where dx/dt≠0\mathrm dx/\mathrm dt\ne0,

dydx=dy/dtdx/dt.\frac{\mathrm dy}{\mathrm dx} =\frac{\mathrm dy/\mathrm dt}{\mathrm dx/\mathrm dt}.

In the formula booklet.

Differentiate both coordinates with respect to tt, then divide in the order “vertical rate over horizontal rate”. The result will usually still contain tt.

Worked example 3: a gradient without eliminating the parameter

Question. Find dy/dx\mathrm dy/\mathrm dx for x=t2x=t^2, y=t+1y=t+1, and evaluate it at t=1t=1.

Plan. Find the two parameter derivatives, then take their ratio.

dxdt=2t,dydt=1dydx=12t(t≠0)\begin{aligned} \frac{\mathrm dx}{\mathrm dt}&=2t,&\frac{\mathrm dy}{\mathrm dt}&=1\\ \frac{\mathrm dy}{\mathrm dx}&=\frac1{2t}&& (t\ne0) \end{aligned}

Answer. dy/dx=1/(2t)\mathrm dy/\mathrm dx=1/(2t), and the gradient at t=1t=1 is 1/21/2.

Check. Near t=1t=1, the curve is y=1+xy=1+\sqrt x, whose derivative at x=1x=1 is 1/(21)=1/21/(2\sqrt1)=1/2.

Keep x=t2x=t^2, but replace the linear yy-coordinate with an exponential. The numerator is now a changing rate too.

Continue this lesson.

You’re previewing NovaLearn without an account. Create a free account to read the rest.

Create free account