MATH 203 Calculus I • Concordia University, Montreal

Practice final with full solutions (MATH 203)

This is a practice final examination for MATH 203, Differential and Integral Calculus I, the first calculus course at Concordia University. Like the real final, it lasts three hours and covers the whole course, from the review of functions to applied optimization, Thomas' Calculus sections 1.2 to 4.6. Twelve questions, one hundred points, three parts: techniques, theorems and reasoning, then applications for about half of the marks. A scientific calculator is allowed, as on the exam: it evaluates, it never differentiates, never computes a limit and never draws a graph, so every derivative, every limit and every sign is still yours.

The thread of the paper is the thread of the course: the marks are lost in the ALGEBRA inside the calculus, not in the rules. Multiply both floors by the conjugate, remember that x2=∣x∣\sqrt{x^2} = |x| at −∞-\infty, factor the lowest power of a bracket before reading a sign, see a difference of squares under a root, rewrite a negative exponent first, reject the candidate that squaring created, and put every angle in radians before it meets a derivative. Every solution names the gesture on the line where it is used. Not one question repeats an exercise of the twenty chapter sets of this site or of the practice midterm: the gestures are those of every MATH 203 final, the functions and situations are new.

The traps named in the solutions: a conjugate applied to one floor only, x2=x\sqrt{x^2} = x for a negative xx, a horizontal tangent outside the domain, 1f′(b)\frac{1}{f'(b)} instead of 1f′(a)\frac{1}{f'(a)} for an inverse, a sign read from a value rounded too hard in a bisection table, a numerator that is not a multiple of hh, a critical point where f′f' does not exist, a calculator that refuses the cube root of a negative number, an angle that closes read as a positive rate, a relative error not multiplied by its exponent, a critical number taken for a maximum when it is a minimum, and 1∞1^\infty read as 11.

12 corrected exercises • 100 points • 180 minutes

Every MATH 203 chapter →

Self-checking set Type your answers under each question: the page tells you right or wrong before you open the solution. With an account, every answer right on the first try earns points.

Part A: techniques (/20)

Exercise 1: Four limits, and the algebra that opens each one

Evaluate each limit exactly, or say that it is infinite. L'Hôpital's Rule is NOT allowed in this question: the first line of each part names the algebraic move (factor, conjugate, match the angle, squeeze).

  • a) lim⁡x→7x2−5x−14x+2−3\displaystyle\lim_{x\to 7} \frac{x^2 - 5x - 14}{\sqrt{x + 2} - 3}
  • b) lim⁡x→0sin⁡(x2)1−cos⁡x\displaystyle\lim_{x\to 0} \frac{\sin(x^2)}{1 - \cos x}
  • c) lim⁡x→−∞(9x2+6x+3x)\displaystyle\lim_{x\to -\infty} \left(\sqrt{9x^2 + 6x} + 3x\right) and lim⁡x→∞(9x2+6x+3x)\displaystyle\lim_{x\to \infty} \left(\sqrt{9x^2 + 6x} + 3x\right)
  • d) lim⁡x→∞5x+2sin⁡xx2+1\displaystyle\lim_{x\to \infty} \frac{5x + 2\sin x}{\sqrt{x^2 + 1}}

Type your answers, the page tells you right or wrong 0/5

a)
b)
c)
d)
Show the solution

Answers

  • a) 5454
  • b) 22
  • c) −1-1 at −∞-\infty; +∞+\infty at +∞+\infty (no indeterminate form there)
  • d) 55: sin⁡xx→0\frac{\sin x}{x} \to 0 at ∞\infty, by the Sandwich Theorem

a) Substitution gives 49−35−143−3=00\frac{49 - 35 - 14}{3 - 3} = \frac{0}{0}: the top and the bottom both hide the factor x−7x - 7. Factor the trinomial, x2−5x−14=(x−7)(x+2)x^2 - 5x - 14 = (x - 7)(x + 2), and multiply top AND bottom by the conjugate x+2+3\sqrt{x + 2} + 3: the denominator becomes (x+2)−9=x−7(x + 2) - 9 = x - 7. For x≠7x \ne 7 the quotient is (x−7)(x+2)(x+2+3)x−7=(x+2)(x+2+3)\frac{(x - 7)(x + 2)\left(\sqrt{x + 2} + 3\right)}{x - 7} = (x + 2)\left(\sqrt{x + 2} + 3\right), continuous at 77, so the limit is 9⋅6=549 \cdot 6 = 54. Multiplying the denominator alone by the conjugate changes the value of the fraction: both floors, always.

b) Substitution gives sin⁡01−1=00\frac{\sin 0}{1 - 1} = \frac{0}{0}. Two known limits are hidden here, and the algebra must bring each one out. On top, the angle is x2x^2, so divide by x2x^2: sin⁡(x2)x2→1\frac{\sin(x^2)}{x^2} \to 1 as x2→0x^2 \to 0. Below, the half angle identity 1−cos⁡x=2sin⁡2x21 - \cos x = 2\sin^2\frac{x}{2} gives 1−cos⁡xx2=2sin⁡2(x/2)x2=12(sin⁡(x/2)x/2)2→12\frac{1 - \cos x}{x^2} = \frac{2\sin^2(x/2)}{x^2} = \frac{1}{2}\left(\frac{\sin(x/2)}{x/2}\right)^2 \to \frac{1}{2}. So sin⁡(x2)1−cos⁡x=sin⁡(x2)x2⋅x21−cos⁡x→1⋅2=2\frac{\sin(x^2)}{1 - \cos x} = \frac{\sin(x^2)}{x^2} \cdot \frac{x^2}{1 - \cos x} \to 1 \cdot 2 = 2. The factor x2x2\frac{x^2}{x^2} is exactly what was multiplied in to match each angle with its denominator; it must be written, not guessed.

c) At +∞+\infty both terms tend to +∞+\infty and so does their SUM: no indeterminate form, no algebra, the limit is +∞+\infty. At −∞-\infty, 9x2+6x→+∞\sqrt{9x^2 + 6x} \to +\infty while 3x→−∞3x \to -\infty: the form ∞−∞\infty - \infty. Multiply and divide by the conjugate 9x2+6x−3x\sqrt{9x^2 + 6x} - 3x: (9x2+6x)−9x29x2+6x−3x=6x9x2+6x−3x\frac{(9x^2 + 6x) - 9x^2}{\sqrt{9x^2 + 6x} - 3x} = \frac{6x}{\sqrt{9x^2 + 6x} - 3x}. Divide top and bottom by xx. For x<0x < 0, x2=∣x∣=−x\sqrt{x^2} = |x| = -x, so 9x2+6xx=−9+6x\frac{\sqrt{9x^2 + 6x}}{x} = -\sqrt{9 + \frac{6}{x}}, and the quotient is 6−9+6/x−3→6−6=−1\frac{6}{-\sqrt{9 + 6/x} - 3} \to \frac{6}{-6} = -1. Writing x2=x\sqrt{x^2} = x here gives 63−3\frac{6}{3 - 3}, a division by zero that should have rung the alarm.

d) Move: divide top and bottom by xx, the dominant power, with x2=∣x∣=x\sqrt{x^2} = |x| = x for x>0x > 0: the quotient becomes 5+2sin⁡xx1+1x2\frac{5 + \frac{2\sin x}{x}}{\sqrt{1 + \frac{1}{x^2}}}. The quotient law needs the limit of sin⁡xx\frac{\sin x}{x} at infinity, and sin⁡x\sin x itself has none: it oscillates forever. But it is bounded, and that is enough: for x>0x > 0, −1x≤sin⁡xx≤1x-\frac{1}{x} \le \frac{\sin x}{x} \le \frac{1}{x}, both bounds tend to 00, so by the Sandwich Theorem sin⁡xx→0\frac{\sin x}{x} \to 0. Not 11: the limit sin⁡θθ→1\frac{\sin\theta}{\theta} \to 1 is at 00, not at infinity. Now every piece has a limit and the quotient law applies: 5+01+0=5\frac{5 + 0}{\sqrt{1 + 0}} = 5.

Tick the exercises you have done or want to review: a free account, no password, keeps your ticks from one visit to the next and tells you which chapter to tackle next. Create your space, an email is enough.

Exercise 2: Three derivatives: every rule named, every result simplified

Name each rule as you use it (product, quotient, chain with its inner function written down), then simplify: the factored form is part of the answer.

  • a) Let f(x)=(x+1)32x+3f(x) = \dfrac{(x + 1)^3}{\sqrt{2x + 3}}. Give its domain, write f′(x)f'(x) as one factored fraction, compute f′(3)f'(3), and find every point where the tangent is horizontal.
  • b) Let g(x)=arcsin⁡(x−1x+1)g(x) = \arcsin\left(\dfrac{x - 1}{x + 1}\right). Find the domain of gg, show that g′(x)=1(x+1)xg'(x) = \dfrac{1}{(x + 1)\sqrt{x}} for x>0x > 0, and compute g′(1)g'(1) and g′(4)g'(4).
  • c) Let y=xe−2xy = xe^{-2x}. Find y′y' and y′′y'', factoring out e−2xe^{-2x} at each step, and show that y′′+4y′+4y=0y'' + 4y' + 4y = 0 for every xx. Give y′(0)y'(0) and y′′(0)y''(0).

Type your answers, the page tells you right or wrong 0/8

a)
b)
Domain of gg ,
c)
Show the solution

Answers

  • a) Domain (−32,∞)\left(-\frac{3}{2}, \infty\right); f′(x)=(x+1)2(5x+8)(2x+3)3/2f'(x) = \frac{(x + 1)^2(5x + 8)}{(2x + 3)^{3/2}}; f′(3)=36827f'(3) = \frac{368}{27}; one horizontal tangent, at (−1,0)(-1, 0) (x=−85x = -\frac{8}{5} rejected)
  • b) Domain [0,∞)[0, \infty); g′(1)=12g'(1) = \frac{1}{2}, g′(4)=110g'(4) = \frac{1}{10}
  • c) y′=e−2x(1−2x)y' = e^{-2x}(1 - 2x), y′′=4e−2x(x−1)y'' = 4e^{-2x}(x - 1); y′(0)=1y'(0) = 1, y′′(0)=−4y''(0) = -4

a) The root needs 2x+3>02x + 3 > 0 (it is also a denominator): the domain is (−32,∞)\left(-\frac{3}{2}, \infty\right). Rewrite the root as a power before differentiating: f(x)=(x+1)3(2x+3)−1/2f(x) = (x + 1)^3(2x + 3)^{-1/2}. Product rule, with the chain rule on each factor (inner functions x+1x + 1 and 2x+32x + 3): f′(x)=3(x+1)2(2x+3)−1/2+(x+1)3⋅(−12)(2x+3)−3/2⋅2f'(x) = 3(x + 1)^2(2x + 3)^{-1/2} + (x + 1)^3 \cdot \left(-\frac{1}{2}\right)(2x + 3)^{-3/2} \cdot 2. Factor the LOWEST power of each bracket, (x+1)2(x + 1)^2 and (2x+3)−3/2(2x + 3)^{-3/2}: f′(x)=(x+1)2(2x+3)−3/2[3(2x+3)−(x+1)]=(x+1)2(5x+8)(2x+3)3/2f'(x) = (x + 1)^2(2x + 3)^{-3/2}\left[3(2x + 3) - (x + 1)\right] = \frac{(x + 1)^2(5x + 8)}{(2x + 3)^{3/2}}. The 2x+32x + 3 inside the bracket is what is left of (2x+3)−1/2(2x + 3)^{-1/2} after taking out (2x+3)−3/2(2x + 3)^{-3/2}: the exponents subtract, −12−(−32)=1-\frac{1}{2} - \left(-\frac{3}{2}\right) = 1.

At x=3x = 3: f′(3)=16⋅2393/2=36827≈13.6296f'(3) = \frac{16 \cdot 23}{9^{3/2}} = \frac{368}{27} \approx 13.6296, since 93/2=279^{3/2} = 27. Horizontal tangents: the numerator vanishes at x=−1x = -1 and at x=−85x = -\frac{8}{5}. But −85=−1.6<−1.5-\frac{8}{5} = -1.6 < -1.5 lies OUTSIDE the domain: rejected. There is exactly one horizontal tangent, at (−1,0)(-1, 0).

b) Domain first. arcsin⁡u\arcsin u needs −1≤u≤1-1 \le u \le 1, with u=x−1x+1u = \frac{x - 1}{x + 1}. For x>−1x > -1 the denominator is positive: u≤1u \le 1 reads x−1≤x+1x - 1 \le x + 1, always true, and u≥−1u \ge -1 reads x−1≥−(x+1)x - 1 \ge -(x + 1), that is x≥0x \ge 0. For x<−1x < -1, u=1+−2x+1>1u = 1 + \frac{-2}{x + 1} > 1. The domain is [0,∞)[0, \infty). Chain rule with the inner function uu: g′(x)=u′1−u2g'(x) = \frac{u'}{\sqrt{1 - u^2}}, and the quotient rule gives u′=(x+1)−(x−1)(x+1)2=2(x+1)2u' = \frac{(x + 1) - (x - 1)}{(x + 1)^2} = \frac{2}{(x + 1)^2}.

The algebra is under the root. Bring 1−u21 - u^2 to one fraction and use a DIFFERENCE OF SQUARES instead of expanding: 1−u2=(x+1)2−(x−1)2(x+1)2=[(x+1)−(x−1)][(x+1)+(x−1)](x+1)2=4x(x+1)21 - u^2 = \frac{(x + 1)^2 - (x - 1)^2}{(x + 1)^2} = \frac{\left[(x + 1) - (x - 1)\right]\left[(x + 1) + (x - 1)\right]}{(x + 1)^2} = \frac{4x}{(x + 1)^2}. Since x+1>0x + 1 > 0, 1−u2=2xx+1\sqrt{1 - u^2} = \frac{2\sqrt{x}}{x + 1}. Then g′(x)=2(x+1)2⋅x+12x=1(x+1)xg'(x) = \frac{2}{(x + 1)^2} \cdot \frac{x + 1}{2\sqrt{x}} = \frac{1}{(x + 1)\sqrt{x}} for x>0x > 0. So g′(1)=12g'(1) = \frac{1}{2} and g′(4)=15⋅2=110g'(4) = \frac{1}{5 \cdot 2} = \frac{1}{10}. At x=0x = 0, u=−1u = -1 and g′(x)→∞g'(x) \to \infty: the graph starts with a vertical tangent.

c) Product rule, with the chain rule on e−2xe^{-2x} (inner function −2x-2x, derivative −2-2): y′=e−2x+x(−2)e−2x=e−2x(1−2x)y' = e^{-2x} + x(-2)e^{-2x} = e^{-2x}(1 - 2x). Again, product rule: y′′=−2e−2x(1−2x)+e−2x(−2)=e−2x(−2+4x−2)=4e−2x(x−1)y'' = -2e^{-2x}(1 - 2x) + e^{-2x}(-2) = e^{-2x}(-2 + 4x - 2) = 4e^{-2x}(x - 1). Keeping e−2xe^{-2x} factored, never distributed, is what makes the check one line: y′′+4y′+4y=e−2x[(4x−4)+4(1−2x)+4x]=e−2x⋅0=0y'' + 4y' + 4y = e^{-2x}\left[(4x - 4) + 4(1 - 2x) + 4x\right] = e^{-2x} \cdot 0 = 0 for every xx. Values: y′(0)=1y'(0) = 1 and y′′(0)=−4y''(0) = -4.

Exercise 3: A moving exponent, and the slope of an inverse with no formula

Parts a) and b) are independent. Three points each.

  • a) Let y=(1+1x)xy = \left(1 + \dfrac{1}{x}\right)^x for x>0x > 0. Explain why neither the power rule nor the rule for axa^x applies, find y′y\frac{y'}{y} by logarithmic differentiation, expanding with the laws of logarithms first, and compute y′(1)y'(1).
  • b) Let f(x)=3x+sin⁡xf(x) = 3x + \sin x. Explain why ff is one-to-one. Find (f−1)′(0)(f^{-1})'(0) and the tangent line to y=f−1(x)y = f^{-1}(x) at the origin, then (f−1)′(3π2+1)(f^{-1})'\left(\frac{3\pi}{2} + 1\right).

Type your answers, the page tells you right or wrong 0/3

a)
b)
Show the solution

Answers

  • a) y′y=ln⁡x+1x−1x+1\frac{y'}{y} = \ln\frac{x + 1}{x} - \frac{1}{x + 1}; y′(1)=2ln⁡2−1≈0.3863y'(1) = 2\ln 2 - 1 \approx 0.3863
  • b) f′≥2>0f' \ge 2 > 0; (f−1)′(0)=14(f^{-1})'(0) = \frac{1}{4}, tangent y=x4y = \frac{x}{4}; (f−1)′(3π2+1)=13(f^{-1})'\left(\frac{3\pi}{2} + 1\right) = \frac{1}{3}

a) The variable sits in the base AND in the exponent: the power rule needs a constant exponent, the rule for axa^x a constant base. Take logarithms, y>0y > 0: ln⁡y=xln⁡(1+1x)\ln y = x\ln\left(1 + \frac{1}{x}\right). Before differentiating, write the inside as ONE fraction and split it with the quotient law: 1+1x=x+1x1 + \frac{1}{x} = \frac{x + 1}{x}, so ln⁡y=xln⁡(x+1)−xln⁡x\ln y = x\ln(x + 1) - x\ln x. Differentiate, the product rule on each term: y′y=ln⁡(x+1)+xx+1−ln⁡x−1=ln⁡x+1x−1x+1\frac{y'}{y} = \ln(x + 1) + \frac{x}{x + 1} - \ln x - 1 = \ln\frac{x + 1}{x} - \frac{1}{x + 1}, since xx+1−1=−1x+1\frac{x}{x + 1} - 1 = -\frac{1}{x + 1}. At x=1x = 1: y(1)=2y(1) = 2 and y′y=ln⁡2−12\frac{y'}{y} = \ln 2 - \frac{1}{2}, so y′(1)=2ln⁡2−1≈0.3863y'(1) = 2\ln 2 - 1 \approx 0.3863. Forgetting to multiply by y(1)=2y(1) = 2 gives half the answer.

b) f′(x)=3+cos⁡x≥3−1=2>0f'(x) = 3 + \cos x \ge 3 - 1 = 2 > 0 for every xx: ff is increasing on R\mathbb{R}, so it never takes the same value twice, and it is one-to-one. No formula for f−1f^{-1} exists, and none is needed: if f(a)=bf(a) = b, then (f−1)′(b)=1f′(a)(f^{-1})'(b) = \frac{1}{f'(a)}. For b=0b = 0: f(0)=0f(0) = 0, so a=0a = 0 and (f−1)′(0)=13+1=14(f^{-1})'(0) = \frac{1}{3 + 1} = \frac{1}{4}. The tangent to y=f−1(x)y = f^{-1}(x) at the origin is y=x4y = \frac{x}{4}, the reflection in y=xy = x of the tangent y=4xy = 4x of ff.

For b=3π2+1b = \frac{3\pi}{2} + 1, find aa FIRST: f(π2)=3π2+1f\left(\frac{\pi}{2}\right) = \frac{3\pi}{2} + 1, so a=π2a = \frac{\pi}{2} and (f−1)′(b)=13+cos⁡π2=13(f^{-1})'(b) = \frac{1}{3 + \cos\frac{\pi}{2}} = \frac{1}{3}. Evaluating f′f' at bb itself gives 13+cos⁡(3π2+1)=13+sin⁡1≈0.26\frac{1}{3 + \cos\left(\frac{3\pi}{2} + 1\right)} = \frac{1}{3 + \sin 1} \approx 0.26: the slope of the wrong point.

Part B: theorems and reasoning (/32)

Exercise 4: Two seams to close, then a root located by bisection

Parts a) and b) are independent. In a), aa and bb are constants and h(x)=sin⁡3xtan⁡xh(x) = \dfrac{\sin 3x}{\tan x} for −π2<x<0-\frac{\pi}{2} < x < 0, h(x)=ax+bh(x) = ax + b for 0≤x≤10 \le x \le 1, and h(x)=x−x−1x−1h(x) = \dfrac{x - x^{-1}}{x - 1} for x>1x > 1. The figure belongs to b), where a calculator is allowed; give its values to four decimals.

0.511.522.533.544.5-2-1.5-1-0.50.511.522.53y = ln xy = 3 − x
  • a) Find aa and bb such that hh is continuous on (−π2,∞)\left(-\frac{\pi}{2}, \infty\right). Each one-sided limit needs one algebraic move: name it.
  • b) Prove that the equation ln⁡x=3−x\ln x = 3 - x has exactly one solution, and that it lies in (2,3)(2, 3), naming the function, the interval and each hypothesis of the Intermediate Value Theorem. Then do four steps of bisection on [2,3][2, 3] with your calculator, in a table, and give the final interval, of length 116\frac{1}{16}.

Type your answers, the page tells you right or wrong 0/5

a)
b)
Show the solution

Answers

  • a) Left limit at 00: 33, so b=3b = 3; right limit at 11: 22, so a+b=2a + b = 2 and a=−1a = -1
  • b) g(x)=ln⁡x+x−3g(x) = \ln x + x - 3: g(2)<0<g(3)g(2) < 0 < g(3), g′>0g' > 0; the root lies in [2.1875,2.25][2.1875, 2.25] (about 2.20792.2079)

a) On each open piece hh is continuous: sin⁡3xtan⁡x\frac{\sin 3x}{\tan x} is a quotient of continuous functions whose denominator does not vanish on (−π2,0)\left(-\frac{\pi}{2}, 0\right), the middle piece is a polynomial, and x−x−1x−1\frac{x - x^{-1}}{x - 1} has a nonzero denominator for x>1x > 1. Only the seams remain, one equation each. At 00, the middle formula owns the point: h(0)=bh(0) = b, and the right-hand limit is bb. Left-hand limit, form 00\frac{0}{0}; move: write tan⁡x=sin⁡xcos⁡x\tan x = \frac{\sin x}{\cos x} and match each sine with its angle, sin⁡3xtan⁡x=sin⁡3xcos⁡xsin⁡x=3⋅sin⁡3x3x⋅xsin⁡x⋅cos⁡x→3⋅1⋅1⋅1=3\frac{\sin 3x}{\tan x} = \frac{\sin 3x\cos x}{\sin x} = 3 \cdot \frac{\sin 3x}{3x} \cdot \frac{x}{\sin x} \cdot \cos x \to 3 \cdot 1 \cdot 1 \cdot 1 = 3. Continuity at 00 requires b=3b = 3.

At 11: h(1)=a+bh(1) = a + b, which is also the left-hand limit. Right-hand limit, form 00\frac{0}{0}; move: rewrite the NEGATIVE exponent first, x−x−1=x−1x=x2−1xx - x^{-1} = x - \frac{1}{x} = \frac{x^2 - 1}{x}, then factor the difference of squares: for x>1x > 1, h(x)=(x−1)(x+1)x(x−1)=x+1x→2h(x) = \frac{(x - 1)(x + 1)}{x(x - 1)} = \frac{x + 1}{x} \to 2. Continuity at 11 requires a+b=2a + b = 2, so a=−1a = -1. With a=−1a = -1 and b=3b = 3, the middle piece 3−x3 - x joins 33 on the left and 22 on the right.

b) Let g(x)=ln⁡x+x−3g(x) = \ln x + x - 3, so that the equation reads g(x)=0g(x) = 0. Existence: gg is continuous on [2,3][2, 3] (a logarithm, continuous for x>0x > 0, plus a polynomial); g(2)=ln⁡2−1≈−0.3069<0g(2) = \ln 2 - 1 \approx -0.3069 < 0 and g(3)=ln⁡3≈1.0986>0g(3) = \ln 3 \approx 1.0986 > 0. By the Intermediate Value Theorem, g(c)=0g(c) = 0 for some cc in (2,3)(2, 3). Uniqueness: g′(x)=1x+1>0g'(x) = \frac{1}{x} + 1 > 0 for x>0x > 0, so gg is increasing on (0,∞)(0, \infty) and takes the value 00 only once. The figure shows it: ln⁡x\ln x rises, 3−x3 - x falls, they cross once.

Bisection, keeping at each step the half where gg changes sign: g(2.5)≈0.4163>0g(2.5) \approx 0.4163 > 0, keep [2,2.5][2, 2.5]; g(2.25)≈0.0609>0g(2.25) \approx 0.0609 > 0, keep [2,2.25][2, 2.25]; g(2.125)≈−0.1212<0g(2.125) \approx -0.1212 < 0, keep [2.125,2.25][2.125, 2.25]; g(2.1875)≈−0.0297<0g(2.1875) \approx -0.0297 < 0, keep [2.1875,2.25][2.1875, 2.25]. The final interval [2.1875,2.25][2.1875, 2.25] has length 116\frac{1}{16} and contains the solution, which is about 2.20792.2079. Each step needs the SIGN only, but a sign read from a value rounded too hard, g(2.25)≈0.1g(2.25) \approx 0.1 or 0.00.0, is where the table goes wrong.

Exercise 5: The derivative from its definition, then read backwards

Let f(x)=3x2+2f(x) = \dfrac{3}{x^2 + 2}. In a), the derivative must come from the definition f′(x)=lim⁡h→0f(x+h)−f(x)hf'(x) = \lim_{h\to 0}\frac{f(x + h) - f(x)}{h}; the differentiation rules may only check the answer.

  • a) Find f′(x)f'(x) from the definition, then the tangent line to the graph of ff at x=1x = 1.
  • b) Each limit is a derivative F′(c)F'(c): name FF and cc, then evaluate. (i) lim⁡h→0esin⁡h−1h\displaystyle\lim_{h\to 0}\frac{e^{\sin h} - 1}{h}; (ii) lim⁡x→π/62sin⁡x−16x−π\displaystyle\lim_{x\to \pi/6}\frac{2\sin x - 1}{6x - \pi}.
  • c) With the function ff of a), find lim⁡h→0f(1+2h)−f(1−h)h\displaystyle\lim_{h\to 0}\frac{f(1 + 2h) - f(1 - h)}{h} without computing any new difference quotient.

Type your answers, the page tells you right or wrong 0/5

a)
b)
c)
Show the solution

Answers

  • a) f′(x)=−6x(x2+2)2f'(x) = -\frac{6x}{(x^2 + 2)^2}; tangent y=1−23(x−1)y = 1 - \frac{2}{3}(x - 1)
  • b) (i) F(x)=esin⁡xF(x) = e^{\sin x}, c=0c = 0: 11; (ii) F=sin⁡F = \sin, c=π6c = \frac{\pi}{6}, times 13\frac{1}{3}: 36\frac{\sqrt{3}}{6}
  • c) 3f′(1)=−23f'(1) = -2

a) Numerator of the difference quotient over ONE common denominator: f(x+h)−f(x)=3(x2+2)−3((x+h)2+2)((x+h)2+2)(x2+2)=3(x2−(x+h)2)((x+h)2+2)(x2+2)f(x + h) - f(x) = \frac{3(x^2 + 2) - 3\left((x + h)^2 + 2\right)}{\left((x + h)^2 + 2\right)(x^2 + 2)} = \frac{3\left(x^2 - (x + h)^2\right)}{\left((x + h)^2 + 2\right)(x^2 + 2)}. The difference of squares x2−(x+h)2=−h(2x+h)x^2 - (x + h)^2 = -h(2x + h) shows the factor hh, as it must: a numerator that is not a multiple of hh signals an algebra error. Dividing by hh cancels it: f(x+h)−f(x)h=−3(2x+h)((x+h)2+2)(x2+2)→−6x(x2+2)2\frac{f(x + h) - f(x)}{h} = \frac{-3(2x + h)}{\left((x + h)^2 + 2\right)(x^2 + 2)} \to \frac{-6x}{(x^2 + 2)^2}. At x=1x = 1: f(1)=1f(1) = 1 and f′(1)=−69=−23f'(1) = -\frac{6}{9} = -\frac{2}{3}, so the tangent is y=1−23(x−1)y = 1 - \frac{2}{3}(x - 1). Check with the rules: f=3(x2+2)−1f = 3(x^2 + 2)^{-1} gives f′=−3(x2+2)−2⋅2xf' = -3(x^2 + 2)^{-2} \cdot 2x.

b) (i) The shape is F(c+h)−F(c)h\frac{F(c + h) - F(c)}{h} with F(x)=esin⁡xF(x) = e^{\sin x} and c=0c = 0: the subtracted number 11 is exactly F(0)=esin⁡0=e0F(0) = e^{\sin 0} = e^0. The limit is F′(0)F'(0), and the chain rule with the inner function sin⁡x\sin x gives F′(x)=esin⁡xcos⁡xF'(x) = e^{\sin x}\cos x, so the limit is F′(0)=1F'(0) = 1. No algebra opens esin⁡h−1e^{\sin h} - 1; recognizing the derivative is the way. (ii) Not yet of the shape F(x)−F(c)x−c\frac{F(x) - F(c)}{x - c}: factor the constants first. 2sin⁡x−1=2(sin⁡x−12)2\sin x - 1 = 2\left(\sin x - \frac{1}{2}\right), 6x−π=6(x−π6)6x - \pi = 6\left(x - \frac{\pi}{6}\right), and sin⁡π6=12\sin\frac{\pi}{6} = \frac{1}{2}. So the quotient is 26⋅sin⁡x−sin⁡π6x−π6→13cos⁡π6=36≈0.2887\frac{2}{6} \cdot \frac{\sin x - \sin\frac{\pi}{6}}{x - \frac{\pi}{6}} \to \frac{1}{3}\cos\frac{\pi}{6} = \frac{\sqrt{3}}{6} \approx 0.2887.

c) Add and subtract f(1)f(1) in the numerator, then split: f(1+2h)−f(1−h)h=2⋅f(1+2h)−f(1)2h+f(1−h)−f(1)−h\frac{f(1 + 2h) - f(1 - h)}{h} = 2 \cdot \frac{f(1 + 2h) - f(1)}{2h} + \frac{f(1 - h) - f(1)}{-h}. As h→0h \to 0, both 2h2h and −h-h tend to 00, so each quotient tends to f′(1)f'(1): the limit is 2f′(1)+f′(1)=3f′(1)=−22f'(1) + f'(1) = 3f'(1) = -2. The increment must be the SAME in the numerator and the denominator of each quotient: that is why the factor 22 and the sign are pulled out first.

Exercise 6: A curve with an exponential in it: tangents, symmetry, and a ceiling

The curve CC of equation x2ey+y=1x^2e^{y} + y = 1 is shown in the figure, with the points A(0,1)A(0, 1), B(1,0)B(1, 0) and B′(−1,0)B'(-1, 0), and the tangent at BB, dashed.

-4-3-2-11234-2-1.5-1-0.50.511.52ABB'tangent at B
  • a) Check that AA, BB and B′B' lie on CC, and explain why CC is symmetric in the yy-axis. Find dydx\frac{dy}{dx} by implicit differentiation, then use the equation of CC to show that, for x≠0x \ne 0, dydx=−2(1−y)x(2−y)\frac{dy}{dx} = -\frac{2(1 - y)}{x(2 - y)}, with no exponential left.
  • b) Find the tangent lines at AA, at BB and at B′B'. Where does the tangent at BB meet the yy-axis?
  • c) Find d2ydx2\frac{d^2y}{dx^2} at AA. Near AA, CC is the graph of a function y(x)y(x): does it have a local extremum at x=0x = 0? Show that no point of CC lies above the line y=1y = 1.

Type your answers, the page tells you right or wrong 0/6

a)
b)
c)
Show the solution

Answers

  • a) dydx=−2xeyx2ey+1=−2(1−y)x(2−y)\frac{dy}{dx} = -\frac{2xe^{y}}{x^2e^{y} + 1} = -\frac{2(1 - y)}{x(2 - y)} for x≠0x \ne 0
  • b) At AA: y=1y = 1; at BB: y=1−xy = 1 - x, which passes through A(0,1)A(0, 1); at B′B': y=1+xy = 1 + x
  • c) y′′=−2e≈−5.4366y'' = -2e \approx -5.4366 at AA: a local maximum; y=1−x2ey≤1y = 1 - x^2e^{y} \le 1 everywhere

a) At AA: 0+1=10 + 1 = 1; at BB: 1⋅e0+0=11 \cdot e^{0} + 0 = 1; at B′B': (−1)2e0+0=1(-1)^2e^{0} + 0 = 1. Replacing xx by −x-x does not change x2x^2, so (x,y)(x, y) is on CC exactly when (−x,y)(-x, y) is: CC is symmetric in the yy-axis. Differentiate both sides with respect to xx, yy being a function of xx. The term x2eyx^2e^{y} is a product, and eye^{y} a composition with inner function yy: ddx(x2ey)=2xey+x2eyy′\frac{d}{dx}\left(x^2e^{y}\right) = 2xe^{y} + x^2e^{y}y'. So 2xey+x2eyy′+y′=02xe^{y} + x^2e^{y}y' + y' = 0; collect and factor, y′(x2ey+1)=−2xeyy'\left(x^2e^{y} + 1\right) = -2xe^{y}, and dydx=−2xeyx2ey+1\frac{dy}{dx} = -\frac{2xe^{y}}{x^2e^{y} + 1}.

Now the equation of CC does the algebra. On CC, x2ey=1−yx^2e^{y} = 1 - y, so the denominator is (1−y)+1=2−y(1 - y) + 1 = 2 - y. For x≠0x \ne 0, the same equation gives ey=1−yx2e^{y} = \frac{1 - y}{x^2}, so the numerator is 2x⋅1−yx2=2(1−y)x2x \cdot \frac{1 - y}{x^2} = \frac{2(1 - y)}{x}. Hence dydx=−2(1−y)x(2−y)\frac{dy}{dx} = -\frac{2(1 - y)}{x(2 - y)}. The xx cancels once, not twice: xx2=1x\frac{x}{x^2} = \frac{1}{x}.

b) At AA: dydx=−00+1=0\frac{dy}{dx} = -\frac{0}{0 + 1} = 0, horizontal tangent y=1y = 1. At BB: dydx=−2(1−0)1(2−0)=−1\frac{dy}{dx} = -\frac{2(1 - 0)}{1(2 - 0)} = -1, tangent y=−(x−1)=1−xy = -(x - 1) = 1 - x. At B′B': −2(−1)(2)=1-\frac{2}{(-1)(2)} = 1, tangent y=1+xy = 1 + x, the mirror image of the tangent at BB, as the symmetry requires: a reflection in the yy-axis changes the sign of every slope. The tangent at BB meets the yy-axis at (0,1)(0, 1), which is the point AA itself: both tangents at BB and B′B' pass through AA.

c) Differentiate 2xey+x2eyy′+y′=02xe^{y} + x^2e^{y}y' + y' = 0 once more. The first term gives 2ey+2xeyy′2e^{y} + 2xe^{y}y'; the second, a product of three factors, gives 2xeyy′+x2ey(y′)2+x2eyy′′2xe^{y}y' + x^2e^{y}(y')^2 + x^2e^{y}y''; the last gives y′′y''. At AA, where x=0x = 0, y=1y = 1 and y′=0y' = 0, everything vanishes except 2e+y′′=02e + y'' = 0: y′′=−2e≈−5.4366y'' = -2e \approx -5.4366. With y′(0)=0y'(0) = 0 and y′′(0)<0y''(0) < 0, the Second Derivative Test gives a local MAXIMUM of y(x)y(x) at x=0x = 0. In fact the maximum is global: y=1−x2eyy = 1 - x^2e^{y} and x2ey≥0x^2e^{y} \ge 0, so y≤1y \le 1 at every point of CC, with equality only at x=0x = 0. The curve never rises above its tangent at AA.

Exercise 7: The closed interval method with a cube root of a negative number

Let f(x)=xx+43=x(x+4)1/3f(x) = x\sqrt[3]{x + 4} = x(x + 4)^{1/3} on the closed interval [−5,4][-5, 4]. A calculator is allowed, but several models answer ERROR to (−1)1/3(-1)^{1/3}: every value asked here is exact.

  • a) Write f′(x)f'(x) as one factored fraction and list the critical points of ff in (−5,4)(-5, 4), saying for each whether f′f' is zero or undefined there.
  • b) Find the absolute maximum and minimum values of ff on [−5,4][-5, 4], and where they occur.
  • c) Does ff have a local extremum at x=−4x = -4? Does ff have an absolute maximum on the OPEN interval (−5,4)(-5, 4)? Does this contradict the Extreme Value Theorem?

Type your answers, the page tells you right or wrong 0/7

a)
b)
c)
Show the solution

Answers

  • a) f′(x)=4(x+3)3(x+4)2/3f'(x) = \frac{4(x + 3)}{3(x + 4)^{2/3}}; critical points −3-3 (f′=0f' = 0) and −4-4 (f′f' undefined)
  • b) f(−5)=5f(-5) = 5, f(−4)=0f(-4) = 0, f(−3)=−3f(-3) = -3, f(4)=8f(4) = 8: maximum 88 at x=4x = 4, minimum −3-3 at x=−3x = -3
  • c) No extremum at −4-4 (f′<0f' < 0 on both sides); no absolute maximum on (−5,4)(-5, 4); no contradiction, the interval is not closed

a) Product rule, and the chain rule on (x+4)1/3(x + 4)^{1/3} (inner function x+4x + 4): f′(x)=(x+4)1/3+x3(x+4)−2/3f'(x) = (x + 4)^{1/3} + \frac{x}{3}(x + 4)^{-2/3}. Factor the LOWEST power, (x+4)−2/3(x + 4)^{-2/3}, and send it to the denominator: f′(x)=(x+4)−2/33[3(x+4)+x]=4x+123(x+4)2/3=4(x+3)3(x+4)2/3f'(x) = \frac{(x + 4)^{-2/3}}{3}\left[3(x + 4) + x\right] = \frac{4x + 12}{3(x + 4)^{2/3}} = \frac{4(x + 3)}{3(x + 4)^{2/3}}. Critical points, interior points where f′f' is zero or undefined: x=−3x = -3, where the numerator vanishes, and x=−4x = -4, where the denominator vanishes while f(−4)=0f(-4) = 0 exists (a vertical tangent). Solving only f′(x)=0f'(x) = 0 loses −4-4.

b) Closed interval method. ff is continuous on [−5,4][-5, 4]: a polynomial times a cube root, which is defined for every real number. Table of values of ff at the critical points and the endpoints: f(−5)=−5−13=(−5)(−1)=5f(-5) = -5\sqrt[3]{-1} = (-5)(-1) = 5, f(−4)=0f(-4) = 0, f(−3)=−313=−3f(-3) = -3\sqrt[3]{1} = -3, f(4)=483=8f(4) = 4\sqrt[3]{8} = 8. The absolute maximum is 88, at x=4x = 4; the absolute minimum is −3-3, at x=−3x = -3. A calculator that refuses (−1)1/3(-1)^{1/3} tempts one to drop the endpoint −5-5: the cube root of a negative number exists, −13=−1\sqrt[3]{-1} = -1, and f(−5)=5f(-5) = 5 is a real candidate, the second largest value.

c) The denominator 3(x+4)2/33(x + 4)^{2/3} is positive on both sides of −4-4, and the numerator 4(x+3)4(x + 3) is negative there: f′<0f' < 0 on (−5,−4)(-5, -4) and on (−4,−3)(-4, -3). So ff keeps decreasing through −4-4: no local extremum, only a vertical tangent. A critical point is a candidate, not a verdict. On the open interval (−5,4)(-5, 4), ff increases on [−3,4)[-3, 4) and takes values as close to 88 as we like, but never 88, since x=4x = 4 is excluded; and f(x)<5<8f(x) < 5 < 8 on (−5,−3](-5, -3]. So ff has NO absolute maximum there; the minimum −3-3 at x=−3x = -3 survives. No contradiction: the Extreme Value Theorem requires a CLOSED interval, and that hypothesis fails.

Exercise 8: A complete study: a root, a line, and two slant asymptotes

Let f(x)=x2+4−x2f(x) = \sqrt{x^2 + 4} - \dfrac{x}{2}. The figure shows its graph, its lowest point and two dashed lines. Every feature must be justified by a computation.

-4-3-2-11234567-11234567y = x/2y = −3x/2y = f(x)
  • a) Give the domain and the yy-intercept of ff. Show that y=x2y = \frac{x}{2} is a slant asymptote as x→∞x \to \infty and y=−3x2y = -\frac{3x}{2} one as x→−∞x \to -\infty, and on which side of each the graph lies.
  • b) Find f′(x)f'(x) and the critical number, checking every candidate that squaring produces. Give the intervals of increase and decrease and the absolute minimum value, exactly.
  • c) Find f′′(x)f''(x) by factoring the lowest power, and the concavity. Does the graph have a point of inflection? Give the range of ff.
  • d) Solve f(x)=2f(x) = 2 exactly.

Type your answers, the page tells you right or wrong 0/9

a)
b)
c)
ff is concave up on ,
d)
Show the solution

Answers

  • a) Domain R\mathbb{R}, f(0)=2f(0) = 2; f(x)−x2=4x2+4+x→0+f(x) - \frac{x}{2} = \frac{4}{\sqrt{x^2 + 4} + x} \to 0^+ and f(x)+3x2=4x2+4−x→0+f(x) + \frac{3x}{2} = \frac{4}{\sqrt{x^2 + 4} - x} \to 0^+: the graph is above both
  • b) Critical number 23\frac{2}{\sqrt{3}} (−23-\frac{2}{\sqrt{3}} rejected); decreasing on (−∞,23]\left(-\infty, \frac{2}{\sqrt{3}}\right], increasing after; minimum 3\sqrt{3}
  • c) f′′(x)=4(x2+4)3/2>0f''(x) = \frac{4}{(x^2 + 4)^{3/2}} > 0: concave up everywhere, no inflection; range [3,∞)[\sqrt{3}, \infty)
  • d) x=0x = 0 or x=83x = \frac{8}{3}

a) x2+4>0x^2 + 4 > 0 for every xx: the domain is R\mathbb{R}, ff is continuous, and there is no vertical asymptote. f(0)=4=2f(0) = \sqrt{4} = 2. As x→∞x \to \infty: f(x)−x2=x2+4−xf(x) - \frac{x}{2} = \sqrt{x^2 + 4} - x, of the form ∞−∞\infty - \infty. Conjugate: x2+4−x=(x2+4)−x2x2+4+x=4x2+4+x→0\sqrt{x^2 + 4} - x = \frac{(x^2 + 4) - x^2}{\sqrt{x^2 + 4} + x} = \frac{4}{\sqrt{x^2 + 4} + x} \to 0, so y=x2y = \frac{x}{2} is a slant asymptote on the right. As x→−∞x \to -\infty: f(x)−(−3x2)=x2+4+x=4x2+4−x→0f(x) - \left(-\frac{3x}{2}\right) = \sqrt{x^2 + 4} + x = \frac{4}{\sqrt{x^2 + 4} - x} \to 0, the denominator tending to +∞+\infty because −x>0-x > 0; so y=−3x2y = -\frac{3x}{2} is a slant asymptote on the left.

Two different slopes, because x2+4\sqrt{x^2 + 4} behaves like x2=∣x∣\sqrt{x^2} = |x|, which is xx on the right and −x-x on the left: ∣x∣−x2|x| - \frac{x}{2} is x2\frac{x}{2} on one side and −3x2-\frac{3x}{2} on the other. Position: x2+4>x2=∣x∣\sqrt{x^2 + 4} > \sqrt{x^2} = |x|, so both differences x2+4−x\sqrt{x^2 + 4} - x and x2+4+x\sqrt{x^2 + 4} + x are positive for EVERY xx: the graph lies above both lines everywhere.

b) f′(x)=xx2+4−12=2x−x2+42x2+4f'(x) = \frac{x}{\sqrt{x^2 + 4}} - \frac{1}{2} = \frac{2x - \sqrt{x^2 + 4}}{2\sqrt{x^2 + 4}}, defined everywhere. f′(x)=0f'(x) = 0 means 2x=x2+42x = \sqrt{x^2 + 4}, which forces x≥0x \ge 0, a root never being negative. Squaring: 4x2=x2+44x^2 = x^2 + 4, x2=43x^2 = \frac{4}{3}, x=±23x = \pm\frac{2}{\sqrt{3}}. Check each: at x=−23x = -\frac{2}{\sqrt{3}}, 2x<0<x2+42x < 0 < \sqrt{x^2 + 4}, so it is NOT a solution: squaring created it, rejected. The only critical number is 23=233≈1.1547\frac{2}{\sqrt{3}} = \frac{2\sqrt{3}}{3} \approx 1.1547.

Sign of the numerator 2x−x2+42x - \sqrt{x^2 + 4}: for x≤0x \le 0 it is negative; for x>0x > 0 both terms are positive and it has the sign of the difference of their squares, 4x2−(x2+4)=3x2−44x^2 - (x^2 + 4) = 3x^2 - 4. So f′<0f' < 0 on (−∞,23)\left(-\infty, \frac{2}{\sqrt{3}}\right) and f′>0f' > 0 on (23,∞)\left(\frac{2}{\sqrt{3}}, \infty\right): ff decreases, then increases. The minimum is ABSOLUTE, since ff decreases on everything to its left and increases on everything to its right. Its value: f(23)=43+4−13=43−13=33=3f\left(\frac{2}{\sqrt{3}}\right) = \sqrt{\frac{4}{3} + 4} - \frac{1}{\sqrt{3}} = \frac{4}{\sqrt{3}} - \frac{1}{\sqrt{3}} = \frac{3}{\sqrt{3}} = \sqrt{3}.

c) Write f′(x)=x(x2+4)−1/2−12f'(x) = x(x^2 + 4)^{-1/2} - \frac{1}{2} and use the product rule: f′′(x)=(x2+4)−1/2−x2(x2+4)−3/2f''(x) = (x^2 + 4)^{-1/2} - x^2(x^2 + 4)^{-3/2}. Factor the lowest power (x2+4)−3/2(x^2 + 4)^{-3/2}: f′′(x)=(x2+4)−3/2[(x2+4)−x2]=4(x2+4)3/2>0f''(x) = (x^2 + 4)^{-3/2}\left[(x^2 + 4) - x^2\right] = \frac{4}{(x^2 + 4)^{3/2}} > 0. The graph is concave up on R\mathbb{R} and has NO point of inflection. With b), the range is [3,∞)[\sqrt{3}, \infty): the minimum is reached, and f(x)→∞f(x) \to \infty at both ends, the graph staying above two lines that rise.

d) x2+4=2+x2\sqrt{x^2 + 4} = 2 + \frac{x}{2} requires 2+x2≥02 + \frac{x}{2} \ge 0, that is x≥−4x \ge -4. Squaring: x2+4=4+2x+x24x^2 + 4 = 4 + 2x + \frac{x^2}{4}, so 3x24−2x=0\frac{3x^2}{4} - 2x = 0 and x(3x4−2)=0x\left(\frac{3x}{4} - 2\right) = 0: x=0x = 0 or x=83x = \frac{8}{3}. Both satisfy x≥−4x \ge -4, and both check: f(0)=2f(0) = 2 and f(83)=649+4−43=103−43=2f\left(\frac{8}{3}\right) = \sqrt{\frac{64}{9} + 4} - \frac{4}{3} = \frac{10}{3} - \frac{4}{3} = 2. Two solutions, one on each side of the minimum, as 2>32 > \sqrt{3} predicts. Dividing 3x24=2x\frac{3x^2}{4} = 2x by xx loses the solution 00.

Part C: applications (/48)

Exercise 9: Related rates on a clock face: the tips of the hands

The minute hand of a wall clock is 66 cm long and its hour hand 44 cm, both measured from the centre OO of the dial. Let θ\theta be the angle between the hands, in radians, and DD the distance between their tips, in cm, both functions of the time tt in hours. The figure shows the hands a few minutes after 3:00. Calculator allowed; give decimals to two places.

123696 cm4 cmDθO
  • a) Give the angular speed of each hand in rad/h, and explain why dθdt=−11π6\frac{d\theta}{dt} = -\frac{11\pi}{6} rad/h from 3:00 until the hands overlap. Use the law of cosines to express D2D^2 in terms of θ\theta, and give DD at 3:00.
  • b) How fast is the distance between the tips changing at 3:00?
  • c) How fast is it changing at 4:00, when θ=2π3\theta = \frac{2\pi}{3} and the angle is still closing?
  • d) How many minutes after 3:00 do the hands first overlap? Give DD and dDdt\frac{dD}{dt} at that instant.
  • e) Let AA be the area of the triangle formed by OO and the two tips. How fast is AA changing at 3:00, and at 4:00?

Type your answers, the page tells you right or wrong 0/9

a)
b)
c)
d)
e)
Show the solution

Answers

  • a) 2π2\pi and π6\frac{\pi}{6} rad/h; D2=52−48cos⁡θD^2 = 52 - 48\cos\theta; D=213≈7.21D = 2\sqrt{13} \approx 7.21 cm at 3:00
  • b) dDdt=−22π13≈−19.17\frac{dD}{dt} = -\frac{22\pi}{\sqrt{13}} \approx -19.17 cm/h: the tips get closer
  • c) dDdt=−113 π19≈−13.73\frac{dD}{dt} = -\frac{11\sqrt{3}\,\pi}{\sqrt{19}} \approx -13.73 cm/h
  • d) 18011≈16.36\frac{180}{11} \approx 16.36 min after 3:00; D=2D = 2 cm, dDdt=0\frac{dD}{dt} = 0
  • e) dAdt=0\frac{dA}{dt} = 0 at 3:00; 11π≈34.5611\pi \approx 34.56 cm2^2/h at 4:00

a) The minute hand turns once, 2π2\pi rad, per hour; the hour hand once per 1212 hours, 2π12=π6\frac{2\pi}{12} = \frac{\pi}{6} rad/h. Both turn clockwise. After 3:00 the minute hand, behind the hour hand, closes the gap at the DIFFERENCE of the two speeds, so θ\theta decreases: dθdt=π6−2π=−11π6\frac{d\theta}{dt} = \frac{\pi}{6} - 2\pi = -\frac{11\pi}{6} rad/h, a negative rate. In the triangle with sides 66 and 44 around the angle θ\theta, the law of cosines gives, at every instant, D2=62+42−2(6)(4)cos⁡θ=52−48cos⁡θD^2 = 6^2 + 4^2 - 2(6)(4)\cos\theta = 52 - 48\cos\theta. At 3:00, θ=π2\theta = \frac{\pi}{2} and cos⁡θ=0\cos\theta = 0: D=52=213≈7.21D = \sqrt{52} = 2\sqrt{13} \approx 7.21 cm.

b) Differentiate the relation with respect to tt, the chain rule on both sides: 2DdDdt=48sin⁡θ dθdt2D\frac{dD}{dt} = 48\sin\theta\,\frac{d\theta}{dt}, so dDdt=24sin⁡θD dθdt\frac{dD}{dt} = \frac{24\sin\theta}{D}\,\frac{d\theta}{dt}. Only now substitute the instant: at 3:00, dDdt=24⋅1213⋅(−11π6)=−22π13≈−19.17\frac{dD}{dt} = \frac{24 \cdot 1}{2\sqrt{13}} \cdot \left(-\frac{11\pi}{6}\right) = -\frac{22\pi}{\sqrt{13}} \approx -19.17 cm/h. The tips are getting closer, at about 1919 cm per hour. With dθdt\frac{d\theta}{dt} in degrees per hour the answer would be 5757 times too large: ddθcos⁡θ=−sin⁡θ\frac{d}{d\theta}\cos\theta = -\sin\theta holds in radians only.

c) At 4:00, cos⁡2π3=−12\cos\frac{2\pi}{3} = -\frac{1}{2} and sin⁡2π3=32\sin\frac{2\pi}{3} = \frac{\sqrt{3}}{2}: D2=52+24=76D^2 = 52 + 24 = 76, D=219≈8.72D = 2\sqrt{19} \approx 8.72 cm. Then dDdt=24⋅32219⋅(−11π6)=−113 π19≈−13.73\frac{dD}{dt} = \frac{24 \cdot \frac{\sqrt{3}}{2}}{2\sqrt{19}} \cdot \left(-\frac{11\pi}{6}\right) = -\frac{11\sqrt{3}\,\pi}{\sqrt{19}} \approx -13.73 cm/h. The angle closes at the same rate as at 3:00, yet the distance shrinks more slowly: the factor 24sin⁡θD\frac{24\sin\theta}{D} is smaller, sin⁡θ\sin\theta being smaller and DD larger.

d) Measure positions clockwise from 12, in radians, tt hours after 3:00: the minute hand is at 2πt2\pi t, the hour hand at π2+π6t\frac{\pi}{2} + \frac{\pi}{6}t. They overlap when 2πt=π2+π6t2\pi t = \frac{\pi}{2} + \frac{\pi}{6}t, that is 11π6t=π2\frac{11\pi}{6}t = \frac{\pi}{2}, t=311t = \frac{3}{11} h =18011≈16.36= \frac{180}{11} \approx 16.36 min, at about 3:16:22. Then θ=0\theta = 0, D2=52−48=4D^2 = 52 - 48 = 4, D=2D = 2 cm, the difference of the two lengths, and dDdt=24sin⁡02⋅(−11π6)=0\frac{dD}{dt} = \frac{24\sin 0}{2}\cdot\left(-\frac{11\pi}{6}\right) = 0: the distance reaches its minimum and stops decreasing.

e) A=12(6)(4)sin⁡θ=12sin⁡θA = \frac{1}{2}(6)(4)\sin\theta = 12\sin\theta, so dAdt=12cos⁡θ dθdt\frac{dA}{dt} = 12\cos\theta\,\frac{d\theta}{dt}. At 3:00, cos⁡π2=0\cos\frac{\pi}{2} = 0: dAdt=0\frac{dA}{dt} = 0, the area is momentarily stationary, at its largest value 1212 cm2^2 with the hands at a right angle. At 4:00: dAdt=12(−12)(−11π6)=11π≈34.56\frac{dA}{dt} = 12\left(-\frac{1}{2}\right)\left(-\frac{11\pi}{6}\right) = 11\pi \approx 34.56 cm2^2/h. Two negative factors give a positive rate: the area GROWS while θ\theta closes from 2π3\frac{2\pi}{3} toward π2\frac{\pi}{2}.

Exercise 10: Linearization and error: the power of a wind turbine

The power delivered by a small wind turbine is modelled by P(v)=0.5v3P(v) = 0.5v^3 kilowatts when the wind blows at vv metres per second, for 4≤v≤144 \le v \le 14. The figure shows PP and its tangent at v=10v = 10. Calculator allowed; give values to two decimals.

34567891011121314200400600800100012001400P = 0.5v³tangent at v = 10v (m/s)P (kW)
  • a) Find the linearization L(v)L(v) of PP at v=10v = 10. Use it to estimate the power at v=10.4v = 10.4 m/s and at v=9.6v = 9.6 m/s, compare with the exact values, and explain the side of both errors with P′′P''.
  • b) The anemometer reads v=10v = 10 m/s with a possible error of 0.20.2 m/s. Use a differential to estimate the largest error in the computed power. Show that dPP=3dvv\frac{dP}{P} = 3\frac{dv}{v} and give the percentage error.
  • c) The operator needs the wind speed that gives a required power. Solve P=0.5v3P = 0.5v^3 for vv, linearize v(P)v(P) at P=500P = 500 kW, and estimate the wind speed that gives 530530 kW. Compare with the calculator. How is the slope of this linearization related to P′(10)P'(10)?
  • d) How precisely must the wind speed be measured at 1010 m/s for the power to be known within 1.5%1.5\%?

Type your answers, the page tells you right or wrong 0/8

a)
b)
c)
d)
Show the solution

Answers

  • a) L(v)=500+150(v−10)L(v) = 500 + 150(v - 10); 560560 and 440440 kW against 562.43562.43 and 442.37442.37 kW: both too small, P′′=3v>0P'' = 3v > 0
  • b) dP=30dP = 30 kW; dPP=3⋅2%=6%\frac{dP}{P} = 3 \cdot 2\% = 6\%
  • c) v=(2P)1/3v = (2P)^{1/3}, slope 1150=1P′(10)\frac{1}{150} = \frac{1}{P'(10)}; estimate 10.210.2 m/s against 10.2010.20 m/s (10601/3≈10.1961060^{1/3} \approx 10.196)
  • d) dvv=0.5%\frac{dv}{v} = 0.5\%: within 0.050.05 m/s

a) P(10)=0.5⋅1000=500P(10) = 0.5 \cdot 1000 = 500 kW and P′(v)=1.5v2P'(v) = 1.5v^2, so P′(10)=150P'(10) = 150 kW per m/s: L(v)=500+150(v−10)L(v) = 500 + 150(v - 10). At 10.410.4: L=500+150(0.4)=560L = 500 + 150(0.4) = 560 kW, while P(10.4)=0.5⋅10.43=562.432P(10.4) = 0.5 \cdot 10.4^3 = 562.432, about 562.43562.43 kW. At 9.69.6: L=500−60=440L = 500 - 60 = 440 kW, while P(9.6)=0.5⋅884.736=442.368P(9.6) = 0.5 \cdot 884.736 = 442.368, about 442.37442.37 kW. Both estimates are too small: P′′(v)=3v>0P''(v) = 3v > 0, so PP is concave up and its tangent lies BELOW the curve on both sides of 1010, as the figure shows. The sign of v−10v - 10 does not decide the side; the concavity does.

b) dP=P′(10) dv=150⋅0.2=30dP = P'(10)\,dv = 150 \cdot 0.2 = 30 kW: the computed 500500 kW may be off by about 3030 kW. In relative terms, dPP=1.5v2 dv0.5v3=3dvv\frac{dP}{P} = \frac{1.5v^2\,dv}{0.5v^3} = 3\frac{dv}{v}: the constant 0.50.5 cancels and the exponent 33 comes down as a factor. Here dvv=0.210=2%\frac{dv}{v} = \frac{0.2}{10} = 2\%, so dPP=6%\frac{dP}{P} = 6\%. A power of 33 triples a relative error: the anemometer is three times more precise than the power it predicts.

c) v3=2Pv^3 = 2P, so v(P)=(2P)1/3v(P) = (2P)^{1/3}. Chain rule, inner function 2P2P: v′(P)=13(2P)−2/3⋅2=23(2P)−2/3v'(P) = \frac{1}{3}(2P)^{-2/3} \cdot 2 = \frac{2}{3}(2P)^{-2/3}. At P=500P = 500: (1000)−2/3=1(10003)2=1100(1000)^{-2/3} = \frac{1}{\left(\sqrt[3]{1000}\right)^2} = \frac{1}{100}, so v′(500)=2300=1150v'(500) = \frac{2}{300} = \frac{1}{150} m/s per kW. The negative fractional exponent is read in two steps, the cube root first, then the square and the reciprocal. The linearization is v≈10+P−500150v \approx 10 + \frac{P - 500}{150}, and for 530530 kW: v≈10+30150=10.2v \approx 10 + \frac{30}{150} = 10.2 m/s. The calculator gives 10601/3≈10.1961060^{1/3} \approx 10.196 m/s: this time the estimate is slightly too LARGE, because v(P)v(P), a cube root, is concave down.

The slope 1150\frac{1}{150} is exactly 1P′(10)\frac{1}{P'(10)}: v(P)v(P) is the inverse function of P(v)P(v), and the derivative of an inverse at b=P(a)b = P(a) is 1P′(a)\frac{1}{P'(a)}. Graphically, the graph of v(P)v(P) is the graph of P(v)P(v) with the two axes exchanged, and exchanging the axes inverts every slope: 150150 kW per m/s becomes 1150\frac{1}{150} m/s per kW.

d) We need ∣dPP∣=3∣dvv∣≤1.5%\left|\frac{dP}{P}\right| = 3\left|\frac{dv}{v}\right| \le 1.5\%, that is ∣dvv∣≤0.5%\left|\frac{dv}{v}\right| \le 0.5\%. At 1010 m/s: ∣dv∣≤0.005⋅10=0.05|dv| \le 0.005 \cdot 10 = 0.05 m/s. Reading the relation in this direction is the whole point of the exponent: to gain a factor on the power, the wind must be measured three times better.

Exercise 11: Applied optimization: where to cut the wire, and where not to

A wire 6060 cm long is cut into two pieces. The piece of length xx is bent into a square, the other, of length 60−x60 - x, into an equilateral triangle, as in the figure. The values x=0x = 0 (all triangle) and x=60x = 60 (all square) are allowed: the wire is then not cut at all. Give lengths and areas to two decimals.

x60 − xx/4(60 − x)/3
  • a) Show that the total area enclosed is A(x)=x216+336(60−x)2A(x) = \dfrac{x^2}{16} + \dfrac{\sqrt{3}}{36}(60 - x)^2 cm2^2, give its domain, and compute A(0)A(0).
  • b) Find the critical number of AA exactly, with a rationalized denominator, then to two decimals.
  • c) Where should the wire be cut to make the total area as small as possible? As large as possible? Give both areas, and justify that they are absolute.
  • d) Now the second piece is bent into a CIRCLE instead of a triangle. Find where to cut for the smallest total area, and show that the largest total area is obtained without cutting, with the circle alone. Rank the three shapes by the area one length of wire encloses.

Type your answers, the page tells you right or wrong 0/7

a)
b)
c)
d)
Show the solution

Answers

  • a) Square of side x4\frac{x}{4}, triangle of side 60−x3\frac{60 - x}{3}; domain [0,60][0, 60]; A(0)=1003≈173.21A(0) = 100\sqrt{3} \approx 173.21 cm2^2
  • b) x=24039+43=7203−96011≈26.10x = \frac{240\sqrt{3}}{9 + 4\sqrt{3}} = \frac{720\sqrt{3} - 960}{11} \approx 26.10 cm
  • c) Minimum ≈97.87\approx 97.87 cm2^2 at x≈26.10x \approx 26.10; maximum 225225 cm2^2 at x=60x = 60: no cut, all square
  • d) Minimum at x=240π+4≈33.61x = \frac{240}{\pi + 4} \approx 33.61 cm; maximum 900π≈286.48\frac{900}{\pi} \approx 286.48 cm2^2 at x=0x = 0, all circle; circle, then square, then triangle

a) The square has perimeter xx, so its side is x4\frac{x}{4} and its area x216\frac{x^2}{16}. The triangle has perimeter 60−x60 - x, so its side is s=60−x3s = \frac{60 - x}{3}, and an equilateral triangle of side ss has height 32s\frac{\sqrt{3}}{2}s and area 34s2=34⋅(60−x)29=336(60−x)2\frac{\sqrt{3}}{4}s^2 = \frac{\sqrt{3}}{4} \cdot \frac{(60 - x)^2}{9} = \frac{\sqrt{3}}{36}(60 - x)^2. The domain is the closed interval [0,60][0, 60], both ends included by the statement, and AA is continuous there. A(0)=336⋅3600=1003≈173.21A(0) = \frac{\sqrt{3}}{36} \cdot 3600 = 100\sqrt{3} \approx 173.21 cm2^2.

b) Chain rule on (60−x)2(60 - x)^2, inner function 60−x60 - x of derivative −1-1: A′(x)=x8−318(60−x)A'(x) = \frac{x}{8} - \frac{\sqrt{3}}{18}(60 - x). Set it to 00 and collect the xx terms: x(18+318)=60318x\left(\frac{1}{8} + \frac{\sqrt{3}}{18}\right) = \frac{60\sqrt{3}}{18}. Multiply both sides by 7272, the common denominator, to clear the fractions: x(9+43)=2403x(9 + 4\sqrt{3}) = 240\sqrt{3}, so x=24039+43x = \frac{240\sqrt{3}}{9 + 4\sqrt{3}}. Rationalize with the conjugate 9−439 - 4\sqrt{3}: the denominator becomes 81−48=3381 - 48 = 33, and x=2403(9−43)33=21603−288033=7203−96011≈26.10x = \frac{240\sqrt{3}(9 - 4\sqrt{3})}{33} = \frac{2160\sqrt{3} - 2880}{33} = \frac{720\sqrt{3} - 960}{11} \approx 26.10 cm. It lies in (0,60)(0, 60).

c) Closed Interval Method, AA being continuous on [0,60][0, 60]: A(0)≈173.21A(0) \approx 173.21, A(26.10)≈97.87A(26.10) \approx 97.87 and A(60)=360016=225A(60) = \frac{3600}{16} = 225 cm2^2. The smallest area, about 97.8797.87 cm2^2, comes from cutting at x≈26.10x \approx 26.10 cm (the exact value is 90039+43\frac{900\sqrt{3}}{9 + 4\sqrt{3}}). The largest, 225225 cm2^2, comes from NOT cutting: all the wire into the square. Also, A′′(x)=18+318>0A''(x) = \frac{1}{8} + \frac{\sqrt{3}}{18} > 0: AA is concave up, so its only critical number is a minimum and the maximum must be at an endpoint. Stopping at A′(x)=0A'(x) = 0 gives the cut that encloses the LEAST area, the opposite of what one wanted.

d) With a circle of circumference 60−x60 - x, the radius is 60−x2π\frac{60 - x}{2\pi} and the area π(60−x2π)2=(60−x)24π\pi\left(\frac{60 - x}{2\pi}\right)^2 = \frac{(60 - x)^2}{4\pi}. So A(x)=x216+(60−x)24πA(x) = \frac{x^2}{16} + \frac{(60 - x)^2}{4\pi} on [0,60][0, 60], and A′(x)=x8−60−x2π=0A'(x) = \frac{x}{8} - \frac{60 - x}{2\pi} = 0 gives 2πx=8(60−x)2\pi x = 8(60 - x), x=4802π+8=240π+4≈33.61x = \frac{480}{2\pi + 8} = \frac{240}{\pi + 4} \approx 33.61 cm, the cut for the smallest area. Endpoints: A(0)=36004π=900π≈286.48A(0) = \frac{3600}{4\pi} = \frac{900}{\pi} \approx 286.48 cm2^2 and A(60)=225A(60) = 225 cm2^2. The largest area is 900π\frac{900}{\pi}, with no cut and the circle alone. For 6060 cm of wire: circle 286.48286.48, square 225225, equilateral triangle 173.21173.21 cm2^2. The rounder the shape, the more it encloses.

Exercise 12: L'Hôpital's Rule in economics: from a CES function to Cobb-Douglas

A firm uses K=16K = 16 units of capital and L=81L = 81 units of labour. Economists model its output with a CES production function (constant elasticity of substitution): Q(ρ)=(0.25⋅16ρ+0.75⋅81ρ)1/ρQ(\rho) = \left(0.25 \cdot 16^{\rho} + 0.75 \cdot 81^{\rho}\right)^{1/\rho}, where the parameter ρ≠0\rho \ne 0 measures how easily capital and labour replace each other. The figure shows QQ as a function of ρ\rho, the dashed lines Q=16Q = 16 and Q=81Q = 81, and a hole at ρ=0\rho = 0. Calculator allowed; decimals to two places.

-6-5-4-3-2-1123456102030405060708090Q = 16Q = 81Q(ρ)ρQ
  • a) Compute Q(1)Q(1), Q(−1)Q(-1) and Q(2)Q(2).
  • b) Name the form of Q(ρ)Q(\rho) as ρ→0\rho \to 0. Find lim⁡ρ→0Q(ρ)\lim_{\rho\to 0} Q(\rho) with L'Hôpital's Rule, and show that it equals 161/4⋅813/416^{1/4} \cdot 81^{3/4}, the output of the Cobb-Douglas function K1/4L3/4K^{1/4}L^{3/4}.
  • c) Find lim⁡ρ→∞Q(ρ)\lim_{\rho\to\infty} Q(\rho) and lim⁡ρ→−∞Q(ρ)\lim_{\rho\to -\infty} Q(\rho), naming each form first and then factoring out the dominant power. Interpret the second limit.
  • d) Prove that for all K>0K > 0, L>0L > 0 and 0<a<10 < a < 1, lim⁡ρ→0(aKρ+(1−a)Lρ)1/ρ=KaL1−a\displaystyle\lim_{\rho\to 0}\left(aK^{\rho} + (1 - a)L^{\rho}\right)^{1/\rho} = K^aL^{1 - a}.
  • e) Compute Q(0.01)Q(0.01) and Q(−0.01)Q(-0.01), and say how the hole of the figure should be filled.

Type your answers, the page tells you right or wrong 0/8

a)
b)
c)
e)
Show the solution

Answers

  • a) Q(1)=64.75Q(1) = 64.75, Q(−1)≈40.19Q(-1) \approx 40.19, Q(2)≈70.60Q(2) \approx 70.60
  • b) 1±∞1^{\pm\infty}; ln⁡Q→0.25ln⁡16+0.75ln⁡81=ln⁡54\ln Q \to 0.25\ln 16 + 0.75\ln 81 = \ln 54: the limit is 54=161/4⋅813/454 = 16^{1/4} \cdot 81^{3/4}
  • c) ∞0\infty^0 at +∞+\infty: limit 8181; 000^0 at −∞-\infty: limit 1616, the scarcer input
  • d) ln⁡\ln of the expression →aln⁡K+(1−a)ln⁡L=ln⁡(KaL1−a)\to a\ln K + (1 - a)\ln L = \ln\left(K^aL^{1 - a}\right)
  • e) Q(0.01)≈54.13Q(0.01) \approx 54.13, Q(−0.01)≈53.87Q(-0.01) \approx 53.87; define Q(0)=54Q(0) = 54, the continuous extension

a) Q(1)=0.25⋅16+0.75⋅81=4+60.75=64.75Q(1) = 0.25 \cdot 16 + 0.75 \cdot 81 = 4 + 60.75 = 64.75, the weighted average of the inputs. Q(−1)=(0.2516+0.7581)−1=10.015625+0.009259…≈40.19Q(-1) = \left(\frac{0.25}{16} + \frac{0.75}{81}\right)^{-1} = \frac{1}{0.015625 + 0.009259\ldots} \approx 40.19: the exponent −1-1 inverts the WHOLE sum, not each term. Q(2)=0.25⋅256+0.75⋅6561=4984.75≈70.60Q(2) = \sqrt{0.25 \cdot 256 + 0.75 \cdot 6561} = \sqrt{4984.75} \approx 70.60.

b) As ρ→0\rho \to 0 the base tends to 0.25+0.75=10.25 + 0.75 = 1 and the exponent 1ρ\frac{1}{\rho} to ±∞\pm\infty: the form 1±∞1^{\pm\infty}, indeterminate, and the answer is NOT 11. The base is positive, so take the logarithm: ln⁡Q(ρ)=ln⁡(0.25⋅16ρ+0.75⋅81ρ)ρ\ln Q(\rho) = \frac{\ln\left(0.25 \cdot 16^{\rho} + 0.75 \cdot 81^{\rho}\right)}{\rho}, of the form ln⁡10=00\frac{\ln 1}{0} = \frac{0}{0}. L'Hôpital's Rule, the variable being ρ\rho and the bases constant, with ddρaρ=aρln⁡a\frac{d}{d\rho}a^{\rho} = a^{\rho}\ln a: the derivative of the numerator is 0.25⋅16ρln⁡16+0.75⋅81ρln⁡810.25⋅16ρ+0.75⋅81ρ\frac{0.25 \cdot 16^{\rho}\ln 16 + 0.75 \cdot 81^{\rho}\ln 81}{0.25 \cdot 16^{\rho} + 0.75 \cdot 81^{\rho}}, that of the denominator is 11.

This quotient is continuous at ρ=0\rho = 0, where it equals 0.25ln⁡16+0.75ln⁡810.25\ln 16 + 0.75\ln 81. Laws of logarithms: 0.25ln⁡16=ln⁡161/4=ln⁡20.25\ln 16 = \ln 16^{1/4} = \ln 2 and 0.75ln⁡81=ln⁡813/4=ln⁡270.75\ln 81 = \ln 81^{3/4} = \ln 27, so ln⁡Q(ρ)→ln⁡2+ln⁡27=ln⁡54\ln Q(\rho) \to \ln 2 + \ln 27 = \ln 54. The exponential being continuous, Q(ρ)=eln⁡Q(ρ)→54=161/4⋅813/4Q(\rho) = e^{\ln Q(\rho)} \to 54 = 16^{1/4} \cdot 81^{3/4}, the Cobb-Douglas output.

c) As ρ→∞\rho \to \infty the base tends to ∞\infty and the exponent to 00: the form ∞0\infty^0. Factor the dominant power 81ρ81^{\rho} out of the base, using (ab)1/ρ=a1/ρb1/ρ(ab)^{1/\rho} = a^{1/\rho}b^{1/\rho}: Q(ρ)=(81ρ)1/ρ(0.75+0.25(1681)ρ)1/ρ=81(0.75+0.25(1681)ρ)1/ρQ(\rho) = \left(81^{\rho}\right)^{1/\rho}\left(0.75 + 0.25\left(\frac{16}{81}\right)^{\rho}\right)^{1/\rho} = 81\left(0.75 + 0.25\left(\frac{16}{81}\right)^{\rho}\right)^{1/\rho}. Now (1681)ρ→0\left(\frac{16}{81}\right)^{\rho} \to 0, the base tends to 0.750.75 and the exponent to 00: 0.750=10.75^0 = 1 is a DETERMINATE form, and the limit is 8181.

As ρ→−∞\rho \to -\infty, 16ρ16^{\rho} and 81ρ81^{\rho} both tend to 00: the base tends to 0+0^+ and the exponent to 0−0^-, the form 000^0. The dominant term is now 16ρ=116∣ρ∣16^{\rho} = \frac{1}{16^{|\rho|}}, which vanishes more slowly than 81ρ81^{\rho}: Q(ρ)=16(0.25+0.75(8116)ρ)1/ρ→16⋅0.250=16Q(\rho) = 16\left(0.25 + 0.75\left(\frac{81}{16}\right)^{\rho}\right)^{1/\rho} \to 16 \cdot 0.25^0 = 16. When capital and labour cannot replace each other at all, output is limited by the scarcer input, the 1616 units of capital. As ρ\rho runs over R\mathbb{R}, QQ goes from 1616 to 8181, as the figure shows.

d) Same three steps. The base aKρ+(1−a)LρaK^{\rho} + (1 - a)L^{\rho} tends to a+(1−a)=1a + (1 - a) = 1 and the exponent to ±∞\pm\infty: the form 1±∞1^{\pm\infty}. Its logarithm, ln⁡(aKρ+(1−a)Lρ)ρ\frac{\ln\left(aK^{\rho} + (1 - a)L^{\rho}\right)}{\rho}, has the form 00\frac{0}{0}, and one round of the rule gives aKρln⁡K+(1−a)Lρln⁡LaKρ+(1−a)Lρ→aln⁡K+(1−a)ln⁡L=ln⁡(KaL1−a)\frac{aK^{\rho}\ln K + (1 - a)L^{\rho}\ln L}{aK^{\rho} + (1 - a)L^{\rho}} \to a\ln K + (1 - a)\ln L = \ln\left(K^aL^{1 - a}\right). The exponential, continuous, gives the limit KaL1−aK^aL^{1 - a}.

e) Q(0.01)≈54.13Q(0.01) \approx 54.13 and Q(−0.01)≈53.87Q(-0.01) \approx 53.87, one on each side of 5454: the values close in on the limit of b). QQ is not defined at 00, but the limit exists and is finite, so setting Q(0)=54Q(0) = 54 gives the continuous extension of QQ: the hole is filled by the Cobb-Douglas output. Typing ρ=0\rho = 0 into the formula gives 11/01^{1/0}, an error, which is why the limit is needed at all.

© Ahmed Squalli Houssaini. Exercise set published at www.letuteurscientifique.ca/en/exercices/math203-final-exam. Free for personal and classroom use; republishing it elsewhere requires written permission (legal notice).

See also

Preparing for the MATH 203 final?

I tutor first-year calculus at Concordia and McGill, in English or in French, in Montreal or online. Get in touch for a first session.

Site by Studio Squalli