Menu
Sign up Sign in

AI HL · AHL 2

Further Derivatives

AHL 2.8

Classifying stationary points and inflection

AHL 2.8

Classifying stationary points and inflection

Target: I can find the stationary points of a function by solving f′(x) = 0 and give their coordinates. (+5 more)

By the end of this section

Key points

Stationary points

  • A stationary point is where f′(x) = 0. Solve f′(x) = 0 for x, then substitute into f(x) for the y-coordinate.
  • Give each stationary point as a pair of coordinates, with its type.
  • Classify by the sign of the gradient either side: + then − is a local maximum, − then + is a local minimum.
  • For fractional powers, ln and e^x, check that each x you find is allowed: ln x needs x > 0.

The second derivative test (HL)

  • If f′(a) = 0 and f″(a) < 0, there is a local maximum at x = a (concave-down).
  • If f′(a) = 0 and f″(a) > 0, there is a local minimum at x = a (concave-up).
  • If f″(a) = 0 the test does not decide. The point may be a maximum, a minimum or a point of inflection. Check the sign of f′ either side.
  • A cubic or quartic with a repeated factor in f′, such as x² in 4x²(x − 3), has f″ = 0 at that root.

Points of inflection (HL)

  • A point of inflection is where the concavity changes: f″ changes sign. At one, f″(x) = 0.
  • f″(x) = 0 alone is not enough. Check that f″ changes sign either side.
  • A point of inflection need not be stationary: y = x³ + x has f′(0) = 1 and a point of inflection at the origin.
  • At a point of inflection the gradient has a local maximum or minimum: in a model, the rate of change is greatest or least there.

Using derivatives to find constants (HL)

  • Differentiate with the letters in place, then substitute the given x into f′ or f″ to make an equation.
  • Two unknown constants need two equations: solve them as a pair.
  • The second derivative of a product uses the product rule twice.

Worked examples

Problems like the ones in class. Pick one, try it, then open the working.
Stationary points of a cubic

Find the stationary points of \(y=x^3+3x^2-24x\) and say whether each is a local maximum or a local minimum.

Working
  1. Gradient zero

    \(\frac{dy}{dx}=3x^2+6x-24=3(x+4)(x-2)=0\)

    A stationary point has gradient 0. Factorise to solve: x = −4 or x = 2.

  2. y-values

    \(x=-4:\ y=80\); \(x=2:\ y=-28\)

    Put each x back into the original curve, not into the gradient.

  3. Gradient either side

    x = −5: +; x = 0: −; x = 3: +

    Test a value either side of each point in dy/dx. Gradient + to − is a maximum. Gradient − to + is a minimum.

  4. Conclusion

    (−4, 80) local maximum; (2, −28) local minimum

    Give the coordinates and the type.

Video walkthrough coming
When f″ is zero

Find and classify the stationary points of \(y=x^4-4x^3\).

Working
  1. Gradient zero

    \(\frac{dy}{dx}=4x^3-12x^2=4x^2(x-3)=0\), so x = 0 or x = 3

    Factorise: the repeated factor gives x = 0.

  2. Second derivative

    \(\frac{d^2y}{dx^2}=12x^2-24x\)

    At x = 3 it is 36 > 0: a local minimum, (3, −27).

  3. x = 0

    \(\frac{d^2y}{dx^2}=0\)

    The test does not decide. It could be a maximum, a minimum or a point of inflection. Check the gradient either side.

  4. Gradient either side of 0

    \(y'(-1)=-16\); \(y'(1)=-8\)

    The gradient is negative on both sides, so it does not change sign. (0, 0) is a stationary point of inflection: the curve is not a maximum or a minimum.

Video walkthrough coming
A stationary point with a fractional power

Find the stationary point of \(y=x+4x^{-1/2}\) for \(x>0\), and classify it. Give coordinates to 3 significant figures.

Working
  1. Gradient zero

    \(y'=1-2x^{-3/2}=0\), so \(x^{3/2}=2\)

    Differentiate each term: 4x^(−1/2) gives −2x^(−3/2).

  2. Solve for x

    \(x=2^{2/3}=1.59\)

    Raise both sides to the power 2/3.

  3. y-value

    \(y=4.76\)

    Substitute into the original curve.

  4. Second derivative

    \(y''=3x^{-5/2}\), at x = 1.59: \(y''=0.945>0\)

    Positive, so a local minimum at (1.59, 4.76).

Video walkthrough coming
A stationary point with e

Find the stationary point of \(y=e^{2x}-6x\) and classify it. Give coordinates to 3 significant figures.

Working
  1. Gradient zero

    \(y'=2e^{2x}-6=0\), so \(e^{2x}=3\)

    Divide by 2 first.

  2. Take ln

    \(2x=\ln 3\), \(x=\tfrac12\ln 3=0.549\)

    ln undoes e. Then divide by 2.

  3. y-value

    \(y=3-3\ln 3=−0.296\)

    e^(2x) = 3 at this point, and 6x = 3 ln 3.

  4. Second derivative

    \(y''=4e^{2x}\) = 12 > 0

    Positive, so a local minimum at (0.549, −0.296).

Video walkthrough coming
A stationary point with ln

Find the stationary point of \(y=\ln x-\tfrac12x^2\) for \(x>0\), and classify it.

Working
  1. Gradient zero

    \(y'=\frac1x-x=0\)

    Multiply by x: 1 = x², so x = 1. Reject x = −1: ln x needs x > 0.

  2. y-value

    \(y=\ln 1-\tfrac12=-0.5\)

    Substitute x = 1 into the original curve.

  3. Second derivative

    \(y''=-\frac{1}{x^2}-1\), at x = 1: \(y''=-2<0\)

    Negative, so a local maximum.

  4. Conclusion

    (1, −0.5) local maximum

    State the coordinates and the type.

Video walkthrough coming
Finding two constants

The function \(f(x)=2x^3+ax^2+bx-3\) has \(f''(-1)=-4\) and \(f'(2)=34\). Find the values of a and b.

Working
  1. Derivatives

    \(f'(x)=6x^2+2ax+b\), \(f''(x)=12x+2a\)

    Differentiate twice. a and b stay as letters.

  2. Use f″(−1)

    \(-12+2a=-4\), so \(a=4\)

    This equation has only a.

  3. Use f′(2)

    \(24+4a+b=34\), so \(24+16+b=34\)

    Put in a = 4.

  4. Solve

    \(b=-6\)

    Check: 24 + 16 − 6 = 34.

Video walkthrough coming
Second derivative with the product rule

Find \(\frac{d^2y}{dx^2}\) for \(y=xe^{2x}\), and evaluate it at x = 0.

Working
  1. First derivative

    \(u=x,\ v=e^{2x}\), \(y'=1\times e^{2x}+x\times 2e^{2x}=(1+2x)e^{2x}\)

    Product rule: u′v + uv′.

  2. Second derivative

    \(y''=2e^{2x}+(1+2x)\times 2e^{2x}=(4+4x)e^{2x}\)

    Product rule again, with u = 1 + 2x.

  3. Evaluate

    \(y''(0)=4\)

    Put x = 0: e⁰ = 1.

Video walkthrough coming
Sketching from sign conditions

Sketch a curve for which, for all x, \(f(x)>0\), \(f'(x)<0\) and \(f''(x)>0\).

−2−11232468xyy = e^(-x)
Working
  1. f(x) > 0

    Every point is above the x-axis.

    No x-intercepts.

  2. f′(x) < 0

    The curve falls from left to right everywhere.

    Gradient negative: decreasing.

  3. f″(x) > 0

    The gradient is increasing, towards 0, so the curve is concave-up.

    It falls more and more slowly.

  4. A curve that fits

    \(y=e^{-x}\)

    It is above the axis, always falling, and flattens out as x grows. Any curve with these three properties is correct.

Video walkthrough coming