Exam : STEP2 | Year : 1999 | Questions : Q1—Q8 | Total marks per question : 20
All questions are pure mathematics. Solutions and examiner commentary are included below.
Q Topic Difficulty Key Techniques 1 对数与大数比较 Logarithms and Large Number Comparison Challenging Stirling公式,对数换底,数量级分析 2 二次方程与判别式 Quadratic Equations and Discriminant Standard 判别式,参数分类讨论,不等式推导 3 高阶导数与数学归纳法 Higher Derivatives and Mathematical Induction Challenging 数学归纳法,乘积求导法则,多项式次数分析 4 组合恒等式 Combinatorial Identities Challenging 二项式定理,比较系数法,组合恒等式 5 三角方程 Trigonometric Equations Standard 辅助角公式,三角恒等式,反三角函数 6 积分与换元法 / Integration and Substitution Challenging 线性分式换元,对数性质化简,分部积分 7 曲线分析与作图 / Curve Analysis and Sketching Challenging 求导判驻点,参数分类讨论,渐近线分析,曲线作图 8 三角级数与求和 / Trigonometric Series and Summation Hard 差积化简,三角求和公式,小角近似,微分求和
Topic : 对数与大数比较 Logarithms and Large Number Comparison | Difficulty : Challenging | Marks : 20
1 Let x = 10 100 x = 10^{100} x = 1 0 100 , y = 10 x y = 10^x y = 1 0 x , z = 10 y z = 10^y z = 1 0 y , and let
a 1 = x ! , a 2 = x y , a 3 = y x , a 4 = z x , a 5 = e x y z , a 6 = z 1 / y , a 7 = y z / x . a_1 = x!, \quad a_2 = x^y, \quad a_3 = y^x, \quad a_4 = z^x, \quad a_5 = e^{xyz}, \quad a_6 = z^{1/y}, \quad a_7 = y^{z/x}. a 1 = x ! , a 2 = x y , a 3 = y x , a 4 = z x , a 5 = e x y z , a 6 = z 1/ y , a 7 = y z / x .
(i) Use Stirling’s approximation n ! ≈ 2 π n n + 1 2 e − n n! \approx \sqrt{2\pi} \, n^{n+\frac{1}{2}} e^{-n} n ! ≈ 2 π n n + 2 1 e − n , which is valid for large n n n , to show that log 10 ( log 10 a 1 ) ≈ 102 \log_{10} (\log_{10} a_1) \approx 102 log 10 ( log 10 a 1 ) ≈ 102 .
(ii) Arrange the seven numbers a 1 , … , a 7 a_1, \dots, a_7 a 1 , … , a 7 in ascending order of magnitude, justifying your result.
Model Solution
Part (i)
Using Stirling’s approximation with n = x = 10 100 n = x = 10^{100} n = x = 1 0 100 :
x ! ≈ 2 π x x + 1 2 e − x x! \approx \sqrt{2\pi} \, x^{x + \frac{1}{2}} e^{-x} x ! ≈ 2 π x x + 2 1 e − x
Taking log 10 \log_{10} log 10 of both sides:
log 10 ( x ! ) ≈ log 10 ( 2 π ) + ( x + 1 2 ) log 10 x − x log 10 e \log_{10}(x!) \approx \log_{10}\!\left(\sqrt{2\pi}\right) + \left(x + \tfrac{1}{2}\right)\log_{10} x - x \log_{10} e log 10 ( x !) ≈ log 10 ( 2 π ) + ( x + 2 1 ) log 10 x − x log 10 e
Since x = 10 100 x = 10^{100} x = 1 0 100 , we have log 10 x = 100 \log_{10} x = 100 log 10 x = 100 . The first term log 10 ( 2 π ) ≈ 0.40 \log_{10}(\sqrt{2\pi}) \approx 0.40 log 10 ( 2 π ) ≈ 0.40 and the third term x log 10 e ≈ 0.434 × 10 100 x \log_{10} e \approx 0.434 \times 10^{100} x log 10 e ≈ 0.434 × 1 0 100 are both negligible compared to the second term ( x + 1 2 ) ⋅ 100 ≈ 10 102 (x + \frac{1}{2}) \cdot 100 \approx 10^{102} ( x + 2 1 ) ⋅ 100 ≈ 1 0 102 . Thus:
log 10 ( x ! ) ≈ 100 x = 100 × 10 100 = 10 102 \log_{10}(x!) \approx 100x = 100 \times 10^{100} = 10^{102} log 10 ( x !) ≈ 100 x = 100 × 1 0 100 = 1 0 102
Taking log 10 \log_{10} log 10 again:
log 10 ( log 10 ( x ! ) ) ≈ log 10 ( 10 102 ) = 102 ■ \log_{10}\!\left(\log_{10}(x!)\right) \approx \log_{10}\!\left(10^{102}\right) = 102 \qquad \blacksquare log 10 ( log 10 ( x !) ) ≈ log 10 ( 1 0 102 ) = 102 ■
Part (ii)
For numbers this large, we compare via log 10 \log_{10} log 10 , and when those are still enormous, via log 10 ( log 10 ) \log_{10}(\log_{10}) log 10 ( log 10 ) . We use x = 10 100 x = 10^{100} x = 1 0 100 , y = 10 x y = 10^x y = 1 0 x , z = 10 y z = 10^y z = 1 0 y , so log 10 x = 100 \log_{10} x = 100 log 10 x = 100 , log 10 y = x = 10 100 \log_{10} y = x = 10^{100} log 10 y = x = 1 0 100 , log 10 z = y = 10 10 100 \log_{10} z = y = 10^{10^{100}} log 10 z = y = 1 0 1 0 100 .
Step 1: compute log 10 a k \log_{10} a_k log 10 a k for each k k k .
a 6 = z 1 / y a_6 = z^{1/y} a 6 = z 1/ y :
log 10 a 6 = 1 y log 10 z = y y = 1. \log_{10} a_6 = \frac{1}{y}\log_{10} z = \frac{y}{y} = 1. log 10 a 6 = y 1 log 10 z = y y = 1.
a 1 = x ! a_1 = x! a 1 = x ! : From part (i), log 10 a 1 ≈ 10 102 \log_{10} a_1 \approx 10^{102} log 10 a 1 ≈ 1 0 102 .
a 3 = y x a_3 = y^x a 3 = y x :
log 10 a 3 = x log 10 y = x ⋅ x = x 2 = 10 200 . \log_{10} a_3 = x \log_{10} y = x \cdot x = x^2 = 10^{200}. log 10 a 3 = x log 10 y = x ⋅ x = x 2 = 1 0 200 .
a 5 = e x y z a_5 = e^{xyz} a 5 = e x y z :
log 10 a 5 = x y z ln 10 . \log_{10} a_5 = \frac{xyz}{\ln 10}. log 10 a 5 = l n 10 x y z .
Here x y z = 10 100 ⋅ 10 10 100 ⋅ 10 10 10 100 = 10 100 + 10 100 + 10 10 100 xyz = 10^{100} \cdot 10^{10^{100}} \cdot 10^{10^{10^{100}}} = 10^{100 + 10^{100} + 10^{10^{100}}} x y z = 1 0 100 ⋅ 1 0 1 0 100 ⋅ 1 0 1 0 1 0 100 = 1 0 100 + 1 0 100 + 1 0 1 0 100 , so
log 10 a 5 ≈ 10 10 10 100 . \log_{10} a_5 \approx 10^{10^{10^{100}}}. log 10 a 5 ≈ 1 0 1 0 1 0 100 .
a 2 = x y a_2 = x^y a 2 = x y :
log 10 a 2 = y log 10 x = 100 y = 100 × 10 10 100 ≈ 10 10 100 + 2 . \log_{10} a_2 = y \log_{10} x = 100y = 100 \times 10^{10^{100}} \approx 10^{10^{100} + 2}. log 10 a 2 = y log 10 x = 100 y = 100 × 1 0 1 0 100 ≈ 1 0 1 0 100 + 2 .
a 4 = z x a_4 = z^x a 4 = z x :
log 10 a 4 = x log 10 z = x y = 10 100 ⋅ 10 10 100 = 10 10 100 + 100 . \log_{10} a_4 = x \log_{10} z = xy = 10^{100} \cdot 10^{10^{100}} = 10^{10^{100} + 100}. log 10 a 4 = x log 10 z = x y = 1 0 100 ⋅ 1 0 1 0 100 = 1 0 1 0 100 + 100 .
a 7 = y z / x a_7 = y^{z/x} a 7 = y z / x :
Since z / x = 10 y / 10 100 = 10 y − 100 z/x = 10^y / 10^{100} = 10^{y - 100} z / x = 1 0 y /1 0 100 = 1 0 y − 100 :
log 10 a 7 = z x log 10 y = 10 y − 100 ⋅ 10 100 = 10 y = 10 10 10 100 . \log_{10} a_7 = \frac{z}{x}\log_{10} y = 10^{y - 100} \cdot 10^{100} = 10^{y} = 10^{10^{10^{100}}}. log 10 a 7 = x z log 10 y = 1 0 y − 100 ⋅ 1 0 100 = 1 0 y = 1 0 1 0 1 0 100 .
Step 2: compare the log 10 a k \log_{10} a_k log 10 a k values.
k k k log 10 a k \log_{10} a_k log 10 a k 6 1 1 1 1 ≈ 10 102 \approx 10^{102} ≈ 1 0 102 3 10 200 10^{200} 1 0 200 2 ≈ 10 10 100 + 2 \approx 10^{10^{100} + 2} ≈ 1 0 1 0 100 + 2 4 ≈ 10 10 100 + 100 \approx 10^{10^{100} + 100} ≈ 1 0 1 0 100 + 100 5 ≈ 10 10 10 100 \approx 10^{10^{10^{100}}} ≈ 1 0 1 0 1 0 100 7 = 10 10 10 100 = 10^{10^{10^{100}}} = 1 0 1 0 1 0 100
The first four are clearly ordered: 1 ≪ 10 102 ≪ 10 200 ≪ 10 10 100 + 2 1 \ll 10^{102} \ll 10^{200} \ll 10^{10^{100}+2} 1 ≪ 1 0 102 ≪ 1 0 200 ≪ 1 0 1 0 100 + 2 , giving a 6 < a 1 < a 3 < a 2 a_6 < a_1 < a_3 < a_2 a 6 < a 1 < a 3 < a 2 .
Next, a 2 < a 4 a_2 < a_4 a 2 < a 4 since 10 10 100 + 2 ≪ 10 10 100 + 100 10^{10^{100}+2} \ll 10^{10^{100}+100} 1 0 1 0 100 + 2 ≪ 1 0 1 0 100 + 100 .
For a 4 a_4 a 4 vs a 5 a_5 a 5 vs a 7 a_7 a 7 : we have 10 10 100 + 100 ≪ 10 10 10 100 10^{10^{100}+100} \ll 10^{10^{10^{100}}} 1 0 1 0 100 + 100 ≪ 1 0 1 0 1 0 100 , so a 4 a_4 a 4 is smaller than both a 5 a_5 a 5 and a 7 a_7 a 7 .
Comparing a 5 a_5 a 5 and a 7 a_7 a 7 : log 10 a 5 = x y z ln 10 \log_{10} a_5 = \frac{xyz}{\ln 10} log 10 a 5 = l n 10 x y z and log 10 a 7 = 10 y \log_{10} a_7 = 10^y log 10 a 7 = 1 0 y . Now
x y z ln 10 = 10 100 ⋅ 10 10 100 ⋅ 10 10 10 100 ln 10 = 10 10 10 100 ⋅ 10 10 100 + 100 ln 10 . \frac{xyz}{\ln 10} = \frac{10^{100} \cdot 10^{10^{100}} \cdot 10^{10^{10^{100}}}}{\ln 10} = \frac{10^{10^{10^{100}}} \cdot 10^{10^{100} + 100}}{\ln 10}. l n 10 x y z = l n 10 1 0 100 ⋅ 1 0 1 0 100 ⋅ 1 0 1 0 1 0 100 = l n 10 1 0 1 0 1 0 100 ⋅ 1 0 1 0 100 + 100 .
Since 10 10 100 + 100 ln 10 ≫ 1 \frac{10^{10^{100}+100}}{\ln 10} \gg 1 l n 10 1 0 1 0 100 + 100 ≫ 1 , we get log 10 a 5 ≫ log 10 a 7 = 10 10 10 100 \log_{10} a_5 \gg \log_{10} a_7 = 10^{10^{10^{100}}} log 10 a 5 ≫ log 10 a 7 = 1 0 1 0 1 0 100 , so a 5 > a 7 a_5 > a_7 a 5 > a 7 .
Ascending order:
a 6 < a 1 < a 3 < a 2 < a 4 < a 7 < a 5 \boxed{a_6 \;<\; a_1 \;<\; a_3 \;<\; a_2 \;<\; a_4 \;<\; a_7 \;<\; a_5} a 6 < a 1 < a 3 < a 2 < a 4 < a 7 < a 5
Topic : 二次方程与判别式 Quadratic Equations and Discriminant | Difficulty : Standard | Marks : 20
2 Consider the quadratic equation
n x 2 + 2 x ( p n 2 + q ) + r n + s = 0 , (*) nx^2 + 2x\sqrt{(pn^2 + q)} + rn + s = 0, \qquad \text{(*)} n x 2 + 2 x ( p n 2 + q ) + r n + s = 0 , (*)
where p > 0 p > 0 p > 0 , p ≠ r p \neq r p = r and n = 1 , 2 , 3 , … n = 1, 2, 3, \dots n = 1 , 2 , 3 , … .
(i) For the case where p = 3 p = 3 p = 3 , q = 50 q = 50 q = 50 , r = 2 r = 2 r = 2 , s = 15 s = 15 s = 15 , find the set of values of n n n for which equation (∗ * ∗ ) has no real roots.
(ii) Prove that if p < r p < r p < r and 4 q ( p − r ) > s 2 4q(p - r) > s^2 4 q ( p − r ) > s 2 , then (∗ * ∗ ) has no real roots for any value of n n n .
(iii) If n = 1 n = 1 n = 1 , p − r = 1 p - r = 1 p − r = 1 and q = s 2 / 8 q = s^2/8 q = s 2 /8 , show that (∗ * ∗ ) has real roots if, and only if, s ⩽ 4 − 2 2 s \leqslant 4 - 2\sqrt{2} s ⩽ 4 − 2 2 or s ⩾ 4 + 2 2 s \geqslant 4 + 2\sqrt{2} s ⩾ 4 + 2 2 .
Model Solution
Part (i)
With p = 3 p = 3 p = 3 , q = 50 q = 50 q = 50 , r = 2 r = 2 r = 2 , s = 15 s = 15 s = 15 , the equation becomes
n x 2 + 2 x 3 n 2 + 50 + 2 n + 15 = 0. nx^2 + 2x\sqrt{3n^2 + 50} + 2n + 15 = 0. n x 2 + 2 x 3 n 2 + 50 + 2 n + 15 = 0.
For no real roots, the discriminant must be negative:
D = 4 ( 3 n 2 + 50 ) − 4 n ( 2 n + 15 ) < 0 D = 4(3n^2 + 50) - 4n(2n + 15) < 0 D = 4 ( 3 n 2 + 50 ) − 4 n ( 2 n + 15 ) < 0
12 n 2 + 200 − 8 n 2 − 60 n < 0 12n^2 + 200 - 8n^2 - 60n < 0 12 n 2 + 200 − 8 n 2 − 60 n < 0
4 n 2 − 60 n + 200 < 0 4n^2 - 60n + 200 < 0 4 n 2 − 60 n + 200 < 0
n 2 − 15 n + 50 < 0 n^2 - 15n + 50 < 0 n 2 − 15 n + 50 < 0
( n − 5 ) ( n − 10 ) < 0 (n - 5)(n - 10) < 0 ( n − 5 ) ( n − 10 ) < 0
This holds when 5 < n < 10 5 < n < 10 5 < n < 10 . Since n n n is a positive integer, the set of values is n ∈ { 6 , 7 , 8 , 9 } n \in \{6, 7, 8, 9\} n ∈ { 6 , 7 , 8 , 9 } .
Part (ii)
The discriminant of n x 2 + 2 x p n 2 + q + r n + s = 0 nx^2 + 2x\sqrt{pn^2 + q} + rn + s = 0 n x 2 + 2 x p n 2 + q + r n + s = 0 is
D = 4 ( p n 2 + q ) − 4 n ( r n + s ) = 4 [ ( p − r ) n 2 − s n + q ] . D = 4(pn^2 + q) - 4n(rn + s) = 4\left[(p - r)n^2 - sn + q\right]. D = 4 ( p n 2 + q ) − 4 n ( r n + s ) = 4 [ ( p − r ) n 2 − s n + q ] .
We need to show D < 0 D < 0 D < 0 for all positive integers n n n when p < r p < r p < r and 4 q ( p − r ) > s 2 4q(p - r) > s^2 4 q ( p − r ) > s 2 .
Since p < r p < r p < r , we have p − r < 0 p - r < 0 p − r < 0 . Write p − r = − ( r − p ) p - r = -(r - p) p − r = − ( r − p ) where r − p > 0 r - p > 0 r − p > 0 . Then:
D = 4 [ − ( r − p ) n 2 − s n + q ] D = 4\left[-(r - p)n^2 - sn + q\right] D = 4 [ − ( r − p ) n 2 − s n + q ]
We want to show ( r − p ) n 2 + s n − q > 0 (r - p)n^2 + sn - q > 0 ( r − p ) n 2 + s n − q > 0 for all positive integers n n n . Consider the quadratic
g ( n ) = ( r − p ) n 2 + s n − q g(n) = (r - p)n^2 + sn - q g ( n ) = ( r − p ) n 2 + s n − q
as a function of real n n n . Its discriminant is
Δ g = s 2 + 4 q ( r − p ) . \Delta_g = s^2 + 4q(r - p). Δ g = s 2 + 4 q ( r − p ) .
The condition 4 q ( p − r ) > s 2 4q(p - r) > s^2 4 q ( p − r ) > s 2 rearranges to s 2 < 4 q ( p − r ) = − 4 q ( r − p ) s^2 < 4q(p - r) = -4q(r - p) s 2 < 4 q ( p − r ) = − 4 q ( r − p ) , so
s 2 + 4 q ( r − p ) < 0 , s^2 + 4q(r - p) < 0, s 2 + 4 q ( r − p ) < 0 ,
meaning Δ g < 0 \Delta_g < 0 Δ g < 0 . Since the leading coefficient r − p > 0 r - p > 0 r − p > 0 and the discriminant is negative, g ( n ) > 0 g(n) > 0 g ( n ) > 0 for all real n n n . In particular, g ( n ) > 0 g(n) > 0 g ( n ) > 0 for all positive integers n n n , so D < 0 D < 0 D < 0 and the equation has no real roots for any value of n n n . ■ \qquad \blacksquare ■
Part (iii)
With n = 1 n = 1 n = 1 and p − r = 1 p - r = 1 p − r = 1 (so p = r + 1 p = r + 1 p = r + 1 ), the equation becomes
x 2 + 2 x p + q + r + s = 0. x^2 + 2x\sqrt{p + q} + r + s = 0. x 2 + 2 x p + q + r + s = 0.
For real roots, the discriminant must satisfy:
D = 4 ( p + q ) − 4 ( r + s ) ⩾ 0 D = 4(p + q) - 4(r + s) \geqslant 0 D = 4 ( p + q ) − 4 ( r + s ) ⩾ 0
p + q − r − s ⩾ 0 p + q - r - s \geqslant 0 p + q − r − s ⩾ 0
Since p − r = 1 p - r = 1 p − r = 1 :
1 + q − s ⩾ 0 1 + q - s \geqslant 0 1 + q − s ⩾ 0
Substituting q = s 2 / 8 q = s^2/8 q = s 2 /8 :
1 + s 2 8 − s ⩾ 0 1 + \frac{s^2}{8} - s \geqslant 0 1 + 8 s 2 − s ⩾ 0
Multiplying through by 8:
s 2 − 8 s + 8 ⩾ 0 s^2 - 8s + 8 \geqslant 0 s 2 − 8 s + 8 ⩾ 0
Using the quadratic formula to find the roots of s 2 − 8 s + 8 = 0 s^2 - 8s + 8 = 0 s 2 − 8 s + 8 = 0 :
s = 8 ± 64 − 32 2 = 8 ± 32 2 = 8 ± 4 2 2 = 4 ± 2 2 s = \frac{8 \pm \sqrt{64 - 32}}{2} = \frac{8 \pm \sqrt{32}}{2} = \frac{8 \pm 4\sqrt{2}}{2} = 4 \pm 2\sqrt{2} s = 2 8 ± 64 − 32 = 2 8 ± 32 = 2 8 ± 4 2 = 4 ± 2 2
Since the quadratic s 2 − 8 s + 8 s^2 - 8s + 8 s 2 − 8 s + 8 opens upward, it is non-negative outside the roots:
s 2 − 8 s + 8 ⩾ 0 ⟺ s ⩽ 4 − 2 2 or s ⩾ 4 + 2 2 s^2 - 8s + 8 \geqslant 0 \iff s \leqslant 4 - 2\sqrt{2} \quad \text{or} \quad s \geqslant 4 + 2\sqrt{2} s 2 − 8 s + 8 ⩾ 0 ⟺ s ⩽ 4 − 2 2 or s ⩾ 4 + 2 2
Therefore the equation has real roots if and only if s ⩽ 4 − 2 2 s \leqslant 4 - 2\sqrt{2} s ⩽ 4 − 2 2 or s ⩾ 4 + 2 2 s \geqslant 4 + 2\sqrt{2} s ⩾ 4 + 2 2 . ■ \qquad \blacksquare ■
Topic : 高阶导数与数学归纳法 Higher Derivatives and Mathematical Induction | Difficulty : Challenging | Marks : 20
3 Let
S n ( x ) = e x 3 d n d x n ( e − x 3 ) . S_n(x) = e^{x^3} \frac{d^n}{dx^n} \left( e^{-x^3} \right). S n ( x ) = e x 3 d x n d n ( e − x 3 ) .
Show that S 2 ( x ) = 9 x 4 − 6 x S_2(x) = 9x^4 - 6x S 2 ( x ) = 9 x 4 − 6 x and find S 3 ( x ) S_3(x) S 3 ( x ) .
Prove by induction on n n n that S n ( x ) S_n(x) S n ( x ) is a polynomial. By means of your induction argument, determine the order of this polynomial and the coefficient of the highest power of x x x .
Show also that if d S n d x = 0 \frac{dS_n}{dx} = 0 d x d S n = 0 for some value a a a of x x x , then S n ( a ) S n + 1 ( a ) ⩽ 0 S_n(a)S_{n+1}(a) \leqslant 0 S n ( a ) S n + 1 ( a ) ⩽ 0 .
Model Solution
We write f ( x ) = e − x 3 f(x) = e^{-x^3} f ( x ) = e − x 3 , so S n ( x ) = e x 3 f ( n ) ( x ) S_n(x) = e^{x^3} f^{(n)}(x) S n ( x ) = e x 3 f ( n ) ( x ) .
Computing S 2 ( x ) S_2(x) S 2 ( x ) . Differentiate f ( x ) = e − x 3 f(x) = e^{-x^3} f ( x ) = e − x 3 :
f ′ ( x ) = − 3 x 2 e − x 3 . f'(x) = -3x^2 \, e^{-x^3}. f ′ ( x ) = − 3 x 2 e − x 3 .
Differentiate again using the product rule:
f ′ ′ ( x ) = ( − 6 x + 9 x 4 ) e − x 3 . f''(x) = \bigl(-6x + 9x^4\bigr) e^{-x^3}. f ′′ ( x ) = ( − 6 x + 9 x 4 ) e − x 3 .
Therefore
S 2 ( x ) = e x 3 f ′ ′ ( x ) = 9 x 4 − 6 x . ✓ S_2(x) = e^{x^3} f''(x) = 9x^4 - 6x. \qquad \checkmark S 2 ( x ) = e x 3 f ′′ ( x ) = 9 x 4 − 6 x . ✓
Computing S 3 ( x ) S_3(x) S 3 ( x ) . Differentiate f ′ ′ ( x ) = ( − 6 x + 9 x 4 ) e − x 3 f''(x) = (-6x + 9x^4)e^{-x^3} f ′′ ( x ) = ( − 6 x + 9 x 4 ) e − x 3 using the product rule:
f ′ ′ ′ ( x ) = ( − 6 + 36 x 3 ) e − x 3 + ( − 6 x + 9 x 4 ) ( − 3 x 2 ) e − x 3 . f'''(x) = \bigl(-6 + 36x^3\bigr) e^{-x^3} + \bigl(-6x + 9x^4\bigr)\bigl(-3x^2\bigr) e^{-x^3}. f ′′′ ( x ) = ( − 6 + 36 x 3 ) e − x 3 + ( − 6 x + 9 x 4 ) ( − 3 x 2 ) e − x 3 .
Expanding the second product:
( − 6 x + 9 x 4 ) ( − 3 x 2 ) = 18 x 3 − 27 x 6 . (-6x + 9x^4)(-3x^2) = 18x^3 - 27x^6. ( − 6 x + 9 x 4 ) ( − 3 x 2 ) = 18 x 3 − 27 x 6 .
Combining:
f ′ ′ ′ ( x ) = ( − 6 + 36 x 3 + 18 x 3 − 27 x 6 ) e − x 3 = ( − 27 x 6 + 54 x 3 − 6 ) e − x 3 . f'''(x) = \bigl(-6 + 36x^3 + 18x^3 - 27x^6\bigr) e^{-x^3} = \bigl(-27x^6 + 54x^3 - 6\bigr) e^{-x^3}. f ′′′ ( x ) = ( − 6 + 36 x 3 + 18 x 3 − 27 x 6 ) e − x 3 = ( − 27 x 6 + 54 x 3 − 6 ) e − x 3 .
Therefore
S 3 ( x ) = e x 3 f ′ ′ ′ ( x ) = − 27 x 6 + 54 x 3 − 6. S_3(x) = e^{x^3} f'''(x) = -27x^6 + 54x^3 - 6. S 3 ( x ) = e x 3 f ′′′ ( x ) = − 27 x 6 + 54 x 3 − 6.
Recurrence relation. From S n ( x ) = e x 3 f ( n ) ( x ) S_n(x) = e^{x^3} f^{(n)}(x) S n ( x ) = e x 3 f ( n ) ( x ) , differentiating with the product rule:
S n ′ ( x ) = 3 x 2 e x 3 f ( n ) ( x ) + e x 3 f ( n + 1 ) ( x ) = 3 x 2 S n ( x ) + S n + 1 ( x ) . S_n'(x) = 3x^2 \, e^{x^3} f^{(n)}(x) + e^{x^3} f^{(n+1)}(x) = 3x^2 S_n(x) + S_{n+1}(x). S n ′ ( x ) = 3 x 2 e x 3 f ( n ) ( x ) + e x 3 f ( n + 1 ) ( x ) = 3 x 2 S n ( x ) + S n + 1 ( x ) .
Rearranging:
S n + 1 ( x ) = S n ′ ( x ) − 3 x 2 S n ( x ) . ( † ) S_{n+1}(x) = S_n'(x) - 3x^2 S_n(x). \qquad (\dagger) S n + 1 ( x ) = S n ′ ( x ) − 3 x 2 S n ( x ) . ( † )
Induction proof. We prove by induction that S n ( x ) S_n(x) S n ( x ) is a polynomial of degree 2 n 2n 2 n with leading coefficient ( − 3 ) n (-3)^n ( − 3 ) n .
Base case. S 0 ( x ) = e x 3 ⋅ e − x 3 = 1 S_0(x) = e^{x^3} \cdot e^{-x^3} = 1 S 0 ( x ) = e x 3 ⋅ e − x 3 = 1 , a polynomial of degree 0 = 2 × 0 0 = 2 \times 0 0 = 2 × 0 with leading coefficient 1 = ( − 3 ) 0 1 = (-3)^0 1 = ( − 3 ) 0 .
Inductive step. Suppose S n ( x ) S_n(x) S n ( x ) is a polynomial of degree 2 n 2n 2 n with leading coefficient ( − 3 ) n (-3)^n ( − 3 ) n , so
S n ( x ) = ( − 3 ) n x 2 n + lower-order terms . S_n(x) = (-3)^n x^{2n} + \text{lower-order terms}. S n ( x ) = ( − 3 ) n x 2 n + lower-order terms .
Then S n ′ ( x ) = 2 n ( − 3 ) n x 2 n − 1 + ⋯ S_n'(x) = 2n(-3)^n x^{2n-1} + \cdots S n ′ ( x ) = 2 n ( − 3 ) n x 2 n − 1 + ⋯ is a polynomial of degree 2 n − 1 2n - 1 2 n − 1 .
Also 3 x 2 S n ( x ) = 3 ( − 3 ) n x 2 n + 2 + ⋯ 3x^2 S_n(x) = 3(-3)^n x^{2n+2} + \cdots 3 x 2 S n ( x ) = 3 ( − 3 ) n x 2 n + 2 + ⋯ , a polynomial of degree 2 n + 2 2n + 2 2 n + 2 .
Since 3 ( − 3 ) n = 3 ⋅ ( − 1 ) n ⋅ 3 n = ( − 1 ) n ⋅ 3 n + 1 3(-3)^n = 3 \cdot (-1)^n \cdot 3^n = (-1)^n \cdot 3^{n+1} 3 ( − 3 ) n = 3 ⋅ ( − 1 ) n ⋅ 3 n = ( − 1 ) n ⋅ 3 n + 1 and ( − 3 ) n + 1 = ( − 1 ) n + 1 ⋅ 3 n + 1 (-3)^{n+1} = (-1)^{n+1} \cdot 3^{n+1} ( − 3 ) n + 1 = ( − 1 ) n + 1 ⋅ 3 n + 1 , we have
3 ( − 3 ) n = − ( − 3 ) n + 1 . 3(-3)^n = -(-3)^{n+1}. 3 ( − 3 ) n = − ( − 3 ) n + 1 .
So 3 x 2 S n ( x ) 3x^2 S_n(x) 3 x 2 S n ( x ) has leading term − ( − 3 ) n + 1 x 2 n + 2 -(-3)^{n+1} x^{2n+2} − ( − 3 ) n + 1 x 2 n + 2 .
From ( † ) (\dagger) ( † ) : the term S n ′ ( x ) S_n'(x) S n ′ ( x ) has degree 2 n − 1 2n - 1 2 n − 1 , which is less than 2 n + 2 2n + 2 2 n + 2 , so the leading term of S n + 1 ( x ) S_{n+1}(x) S n + 1 ( x ) comes entirely from − 3 x 2 S n ( x ) -3x^2 S_n(x) − 3 x 2 S n ( x ) :
S n + 1 ( x ) = S n ′ ( x ) ⏟ degree 2 n − 1 − 3 x 2 S n ( x ) ⏟ degree 2 n + 2 = ( − 3 ) n + 1 x 2 n + 2 + ⋯ S_{n+1}(x) = \underbrace{S_n'(x)}_{\text{degree } 2n-1} - \underbrace{3x^2 S_n(x)}_{\text{degree } 2n+2} = (-3)^{n+1} x^{2n+2} + \cdots S n + 1 ( x ) = degree 2 n − 1 S n ′ ( x ) − degree 2 n + 2 3 x 2 S n ( x ) = ( − 3 ) n + 1 x 2 n + 2 + ⋯
Hence S n + 1 ( x ) S_{n+1}(x) S n + 1 ( x ) is a polynomial of degree 2 ( n + 1 ) 2(n+1) 2 ( n + 1 ) with leading coefficient ( − 3 ) n + 1 (-3)^{n+1} ( − 3 ) n + 1 . This completes the induction.
Conclusion. S n ( x ) S_n(x) S n ( x ) is a polynomial of degree 2 n 2n 2 n with leading coefficient ( − 3 ) n (-3)^n ( − 3 ) n .
Sign property. Suppose S n ′ ( a ) = 0 S_n'(a) = 0 S n ′ ( a ) = 0 for some value a a a . Substituting into ( † ) (\dagger) ( † ) :
S n + 1 ( a ) = S n ′ ( a ) − 3 a 2 S n ( a ) = 0 − 3 a 2 S n ( a ) = − 3 a 2 S n ( a ) . S_{n+1}(a) = S_n'(a) - 3a^2 S_n(a) = 0 - 3a^2 S_n(a) = -3a^2 S_n(a). S n + 1 ( a ) = S n ′ ( a ) − 3 a 2 S n ( a ) = 0 − 3 a 2 S n ( a ) = − 3 a 2 S n ( a ) .
Therefore
S n ( a ) S n + 1 ( a ) = S n ( a ) ⋅ ( − 3 a 2 S n ( a ) ) = − 3 a 2 [ S n ( a ) ] 2 . S_n(a) \, S_{n+1}(a) = S_n(a) \cdot \bigl(-3a^2 S_n(a)\bigr) = -3a^2 \bigl[S_n(a)\bigr]^2. S n ( a ) S n + 1 ( a ) = S n ( a ) ⋅ ( − 3 a 2 S n ( a ) ) = − 3 a 2 [ S n ( a ) ] 2 .
Since a 2 ⩾ 0 a^2 \geqslant 0 a 2 ⩾ 0 and [ S n ( a ) ] 2 ⩾ 0 [S_n(a)]^2 \geqslant 0 [ S n ( a ) ] 2 ⩾ 0 , we conclude
S n ( a ) S n + 1 ( a ) = − 3 a 2 [ S n ( a ) ] 2 ⩽ 0. ■ S_n(a) \, S_{n+1}(a) = -3a^2 [S_n(a)]^2 \leqslant 0. \qquad \blacksquare S n ( a ) S n + 1 ( a ) = − 3 a 2 [ S n ( a ) ] 2 ⩽ 0. ■
Topic : 组合恒等式 Combinatorial Identities | Difficulty : Challenging | Marks : 20
4 By considering the expansions in powers of x x x of both sides of the identity
( 1 + x ) n ( 1 + x ) n ≡ ( 1 + x ) 2 n , (1 + x)^n(1 + x)^n \equiv (1 + x)^{2n}, ( 1 + x ) n ( 1 + x ) n ≡ ( 1 + x ) 2 n ,
show that
∑ s = 0 n ( n s ) 2 = ( 2 n n ) , \sum_{s=0}^{n} \binom{n}{s}^2 = \binom{2n}{n}, ∑ s = 0 n ( s n ) 2 = ( n 2 n ) ,
where ( n s ) = n ! s ! ( n − s ) ! \binom{n}{s} = \frac{n!}{s!(n-s)!} ( s n ) = s ! ( n − s )! n ! .
By considering similar identities, or otherwise, show also that:
(i) if n n n is an even integer, then
∑ s = 0 n ( − 1 ) s ( n s ) 2 = ( − 1 ) n / 2 ( n n / 2 ) ; \sum_{s=0}^{n} (-1)^s \binom{n}{s}^2 = (-1)^{n/2} \binom{n}{n/2}; ∑ s = 0 n ( − 1 ) s ( s n ) 2 = ( − 1 ) n /2 ( n /2 n ) ;
(ii) ∑ t = 1 n 2 t ( n t ) 2 = n ( 2 n n ) \sum_{t=1}^{n} 2t \binom{n}{t}^2 = n \binom{2n}{n} ∑ t = 1 n 2 t ( t n ) 2 = n ( n 2 n ) .
Model Solution
Main identity. Expand both sides of ( 1 + x ) n ( 1 + x ) n = ( 1 + x ) 2 n (1+x)^n(1+x)^n = (1+x)^{2n} ( 1 + x ) n ( 1 + x ) n = ( 1 + x ) 2 n .
The left side is
( 1 + x ) n ⋅ ( 1 + x ) n = ( ∑ s = 0 n ( n s ) x s ) ( ∑ t = 0 n ( n t ) x t ) . (1+x)^n \cdot (1+x)^n = \left(\sum_{s=0}^{n} \binom{n}{s} x^s\right)\left(\sum_{t=0}^{n} \binom{n}{t} x^t\right). ( 1 + x ) n ⋅ ( 1 + x ) n = ( ∑ s = 0 n ( s n ) x s ) ( ∑ t = 0 n ( t n ) x t ) .
The coefficient of x n x^n x n in this product is obtained by pairing terms where s + t = n s + t = n s + t = n , i.e. t = n − s t = n - s t = n − s :
[ x n ] : ∑ s = 0 n ( n s ) ( n n − s ) = ∑ s = 0 n ( n s ) 2 , [x^n]: \sum_{s=0}^{n} \binom{n}{s} \binom{n}{n-s} = \sum_{s=0}^{n} \binom{n}{s}^2, [ x n ] : ∑ s = 0 n ( s n ) ( n − s n ) = ∑ s = 0 n ( s n ) 2 ,
using ( n n − s ) = ( n s ) \binom{n}{n-s} = \binom{n}{s} ( n − s n ) = ( s n ) .
The right side gives ( 1 + x ) 2 n (1+x)^{2n} ( 1 + x ) 2 n , whose coefficient of x n x^n x n is ( 2 n n ) \binom{2n}{n} ( n 2 n ) .
Equating coefficients of x n x^n x n :
∑ s = 0 n ( n s ) 2 = ( 2 n n ) . ■ \sum_{s=0}^{n} \binom{n}{s}^2 = \binom{2n}{n}. \qquad \blacksquare ∑ s = 0 n ( s n ) 2 = ( n 2 n ) . ■
Part (i). Consider the identity
( 1 + x ) n ( 1 − x ) n = ( 1 − x 2 ) n . (1+x)^n(1-x)^n = (1-x^2)^n. ( 1 + x ) n ( 1 − x ) n = ( 1 − x 2 ) n .
The left side expands as
( ∑ s = 0 n ( n s ) x s ) ( ∑ t = 0 n ( n t ) ( − x ) t ) = ( ∑ s = 0 n ( n s ) x s ) ( ∑ t = 0 n ( − 1 ) t ( n t ) x t ) . \left(\sum_{s=0}^{n} \binom{n}{s} x^s\right)\left(\sum_{t=0}^{n} \binom{n}{t} (-x)^t\right) = \left(\sum_{s=0}^{n} \binom{n}{s} x^s\right)\left(\sum_{t=0}^{n} (-1)^t \binom{n}{t} x^t\right). ( ∑ s = 0 n ( s n ) x s ) ( ∑ t = 0 n ( t n ) ( − x ) t ) = ( ∑ s = 0 n ( s n ) x s ) ( ∑ t = 0 n ( − 1 ) t ( t n ) x t ) .
The coefficient of x n x^n x n on the left side is
∑ s = 0 n ( n s ) ⋅ ( − 1 ) n − s ( n n − s ) = ( − 1 ) n ∑ s = 0 n ( − 1 ) − s ( n s ) 2 = ( − 1 ) n ∑ s = 0 n ( − 1 ) − s ( n s ) 2 . \sum_{s=0}^{n} \binom{n}{s} \cdot (-1)^{n-s} \binom{n}{n-s} = (-1)^n \sum_{s=0}^{n} (-1)^{-s} \binom{n}{s}^2 = (-1)^n \sum_{s=0}^{n} (-1)^{-s} \binom{n}{s}^2. ∑ s = 0 n ( s n ) ⋅ ( − 1 ) n − s ( n − s n ) = ( − 1 ) n ∑ s = 0 n ( − 1 ) − s ( s n ) 2 = ( − 1 ) n ∑ s = 0 n ( − 1 ) − s ( s n ) 2 .
Since ( − 1 ) − s = ( − 1 ) s (-1)^{-s} = (-1)^s ( − 1 ) − s = ( − 1 ) s (as ( − 1 ) − 1 = − 1 (-1)^{-1} = -1 ( − 1 ) − 1 = − 1 ), this becomes
( − 1 ) n ∑ s = 0 n ( − 1 ) s ( n s ) 2 . (-1)^n \sum_{s=0}^{n} (-1)^s \binom{n}{s}^2. ( − 1 ) n ∑ s = 0 n ( − 1 ) s ( s n ) 2 .
The right side is ( 1 − x 2 ) n = ∑ k = 0 n ( n k ) ( − 1 ) k x 2 k (1-x^2)^n = \sum_{k=0}^{n} \binom{n}{k} (-1)^k x^{2k} ( 1 − x 2 ) n = ∑ k = 0 n ( k n ) ( − 1 ) k x 2 k .
The coefficient of x n x^n x n on the right side is 0 0 0 when n n n is odd (since 2 k 2k 2 k is always even). When n n n is even, the term x n x^n x n arises from 2 k = n 2k = n 2 k = n , i.e. k = n / 2 k = n/2 k = n /2 :
[ x n ] right = { 0 if n is odd , ( − 1 ) n / 2 ( n n / 2 ) if n is even . [x^n]_{\text{right}} = \begin{cases} 0 & \text{if } n \text{ is odd}, \\ (-1)^{n/2} \binom{n}{n/2} & \text{if } n \text{ is even}. \end{cases} [ x n ] right = { 0 ( − 1 ) n /2 ( n /2 n ) if n is odd , if n is even .
For even n n n , equating coefficients of x n x^n x n :
( − 1 ) n ∑ s = 0 n ( − 1 ) s ( n s ) 2 = ( − 1 ) n / 2 ( n n / 2 ) . (-1)^n \sum_{s=0}^{n} (-1)^s \binom{n}{s}^2 = (-1)^{n/2} \binom{n}{n/2}. ( − 1 ) n ∑ s = 0 n ( − 1 ) s ( s n ) 2 = ( − 1 ) n /2 ( n /2 n ) .
Since n n n is even, ( − 1 ) n = 1 (-1)^n = 1 ( − 1 ) n = 1 , so
∑ s = 0 n ( − 1 ) s ( n s ) 2 = ( − 1 ) n / 2 ( n n / 2 ) . ■ \sum_{s=0}^{n} (-1)^s \binom{n}{s}^2 = (-1)^{n/2} \binom{n}{n/2}. \qquad \blacksquare ∑ s = 0 n ( − 1 ) s ( s n ) 2 = ( − 1 ) n /2 ( n /2 n ) . ■
Part (ii). We use the product ( 1 + x ) n ⋅ ( 1 + x ) n = ( 1 + x ) 2 n (1+x)^n \cdot (1+x)^n = (1+x)^{2n} ( 1 + x ) n ⋅ ( 1 + x ) n = ( 1 + x ) 2 n and differentiate both sides with respect to x x x .
Differentiating: 2 n ( 1 + x ) n ( 1 + x ) n − 1 = 2 n ( 1 + x ) 2 n − 1 2n(1+x)^n(1+x)^{n-1} = 2n(1+x)^{2n-1} 2 n ( 1 + x ) n ( 1 + x ) n − 1 = 2 n ( 1 + x ) 2 n − 1 .
Equivalently, writing f ( x ) = ( 1 + x ) n f(x) = (1+x)^n f ( x ) = ( 1 + x ) n :
f ′ ( x ) ⋅ f ( x ) + f ( x ) ⋅ f ′ ( x ) = 2 n ( 1 + x ) 2 n − 1 , f'(x) \cdot f(x) + f(x) \cdot f'(x) = 2n(1+x)^{2n-1}, f ′ ( x ) ⋅ f ( x ) + f ( x ) ⋅ f ′ ( x ) = 2 n ( 1 + x ) 2 n − 1 ,
which simplifies to 2 ⋅ n ( 1 + x ) n − 1 ( 1 + x ) n = 2 n ( 1 + x ) 2 n − 1 2 \cdot n(1+x)^{n-1}(1+x)^n = 2n(1+x)^{2n-1} 2 ⋅ n ( 1 + x ) n − 1 ( 1 + x ) n = 2 n ( 1 + x ) 2 n − 1 .
Now extract the coefficient of x n − 1 x^{n-1} x n − 1 from each side.
Left side. We have
2 ( ∑ s = 0 n s ( n s ) x s − 1 ) ( ∑ t = 0 n ( n t ) x t ) . 2 \left(\sum_{s=0}^{n} s \binom{n}{s} x^{s-1}\right)\left(\sum_{t=0}^{n} \binom{n}{t} x^t\right). 2 ( ∑ s = 0 n s ( s n ) x s − 1 ) ( ∑ t = 0 n ( t n ) x t ) .
The coefficient of x n − 1 x^{n-1} x n − 1 is obtained when ( s − 1 ) + t = n − 1 (s-1) + t = n-1 ( s − 1 ) + t = n − 1 , i.e. t = n − s t = n - s t = n − s :
2 ∑ s = 0 n s ( n s ) ( n n − s ) = 2 ∑ s = 0 n s ( n s ) 2 . 2 \sum_{s=0}^{n} s \binom{n}{s} \binom{n}{n-s} = 2 \sum_{s=0}^{n} s \binom{n}{s}^2. 2 ∑ s = 0 n s ( s n ) ( n − s n ) = 2 ∑ s = 0 n s ( s n ) 2 .
Right side. The coefficient of x n − 1 x^{n-1} x n − 1 in 2 n ( 1 + x ) 2 n − 1 2n(1+x)^{2n-1} 2 n ( 1 + x ) 2 n − 1 is
2 n ( 2 n − 1 n − 1 ) . 2n \binom{2n-1}{n-1}. 2 n ( n − 1 2 n − 1 ) .
Equating:
2 ∑ s = 0 n s ( n s ) 2 = 2 n ( 2 n − 1 n − 1 ) . 2 \sum_{s=0}^{n} s \binom{n}{s}^2 = 2n \binom{2n-1}{n-1}. 2 ∑ s = 0 n s ( s n ) 2 = 2 n ( n − 1 2 n − 1 ) .
Using the identity ( 2 n − 1 n − 1 ) = n 2 n ( 2 n n ) = 1 2 ( 2 n n ) \binom{2n-1}{n-1} = \frac{n}{2n} \binom{2n}{n} = \frac{1}{2}\binom{2n}{n} ( n − 1 2 n − 1 ) = 2 n n ( n 2 n ) = 2 1 ( n 2 n ) :
2 ∑ s = 0 n s ( n s ) 2 = 2 n ⋅ 1 2 ( 2 n n ) = n ( 2 n n ) . 2 \sum_{s=0}^{n} s \binom{n}{s}^2 = 2n \cdot \frac{1}{2} \binom{2n}{n} = n\binom{2n}{n}. 2 ∑ s = 0 n s ( s n ) 2 = 2 n ⋅ 2 1 ( n 2 n ) = n ( n 2 n ) .
Relabelling s s s as t t t :
∑ t = 1 n 2 t ( n t ) 2 = n ( 2 n n ) . ■ \sum_{t=1}^{n} 2t \binom{n}{t}^2 = n \binom{2n}{n}. \qquad \blacksquare ∑ t = 1 n 2 t ( t n ) 2 = n ( n 2 n ) . ■
Topic : 三角方程 Trigonometric Equations | Difficulty : Standard | Marks : 20
5 Show that if α \alpha α is a solution of the equation
5 cos x + 12 sin x = 7 , 5\cos x + 12\sin x = 7, 5 cos x + 12 sin x = 7 ,
then either
cos α = 35 − 12 120 169 \cos \alpha = \frac{35 - 12\sqrt{120}}{169} cos α = 169 35 − 12 120
or cos α \cos \alpha cos α has one other value which you should find.
Prove carefully that if π / 2 < α < π \pi/2 < \alpha < \pi π /2 < α < π , then α < 3 π / 4 \alpha < 3\pi/4 α < 3 π /4 .
Model Solution
Finding the possible values of cos α \cos\alpha cos α . Since 5 cos α + 12 sin α = 7 5\cos\alpha + 12\sin\alpha = 7 5 cos α + 12 sin α = 7 , we have
sin α = 7 − 5 cos α 12 . ( 1 ) \sin\alpha = \frac{7 - 5\cos\alpha}{12}. \qquad (1) sin α = 12 7 − 5 c o s α . ( 1 )
Substituting into sin 2 α + cos 2 α = 1 \sin^2\alpha + \cos^2\alpha = 1 sin 2 α + cos 2 α = 1 :
( 7 − 5 cos α 12 ) 2 + cos 2 α = 1. \left(\frac{7 - 5\cos\alpha}{12}\right)^2 + \cos^2\alpha = 1. ( 12 7 − 5 c o s α ) 2 + cos 2 α = 1.
Expanding:
49 − 70 cos α + 25 cos 2 α 144 + cos 2 α = 1. \frac{49 - 70\cos\alpha + 25\cos^2\alpha}{144} + \cos^2\alpha = 1. 144 49 − 70 c o s α + 25 c o s 2 α + cos 2 α = 1.
Multiply through by 144 144 144 :
49 − 70 cos α + 25 cos 2 α + 144 cos 2 α = 144. 49 - 70\cos\alpha + 25\cos^2\alpha + 144\cos^2\alpha = 144. 49 − 70 cos α + 25 cos 2 α + 144 cos 2 α = 144.
169 cos 2 α − 70 cos α − 95 = 0. ( 2 ) 169\cos^2\alpha - 70\cos\alpha - 95 = 0. \qquad (2) 169 cos 2 α − 70 cos α − 95 = 0. ( 2 )
Applying the quadratic formula with a = 169 a = 169 a = 169 , b = − 70 b = -70 b = − 70 , c = − 95 c = -95 c = − 95 :
cos α = 70 ± 4900 + 4 × 169 × 95 338 = 70 ± 4900 + 64220 338 = 70 ± 69120 338 . \cos\alpha = \frac{70 \pm \sqrt{4900 + 4 \times 169 \times 95}}{338} = \frac{70 \pm \sqrt{4900 + 64220}}{338} = \frac{70 \pm \sqrt{69120}}{338}. cos α = 338 70 ± 4900 + 4 × 169 × 95 = 338 70 ± 4900 + 64220 = 338 70 ± 69120 .
Simplifying 69120 \sqrt{69120} 69120 : since 69120 = 576 × 120 69120 = 576 \times 120 69120 = 576 × 120 , we get 69120 = 24 120 \sqrt{69120} = 24\sqrt{120} 69120 = 24 120 . Therefore
cos α = 70 ± 24 120 338 = 35 ± 12 120 169 . \cos\alpha = \frac{70 \pm 24\sqrt{120}}{338} = \frac{35 \pm 12\sqrt{120}}{169}. cos α = 338 70 ± 24 120 = 169 35 ± 12 120 .
So the two possible values are
cos α = 35 − 12 120 169 or cos α = 35 + 12 120 169 . ✓ \cos\alpha = \frac{35 - 12\sqrt{120}}{169} \qquad \text{or} \qquad \cos\alpha = \frac{35 + 12\sqrt{120}}{169}. \qquad \checkmark cos α = 169 35 − 12 120 or cos α = 169 35 + 12 120 . ✓
Checking both values give valid solutions. Note 120 ≈ 10.954 \sqrt{120} \approx 10.954 120 ≈ 10.954 , so:
cos α = 35 − 12 120 169 ≈ 35 − 131.45 169 ≈ − 0.571 \cos\alpha = \frac{35 - 12\sqrt{120}}{169} \approx \frac{35 - 131.45}{169} \approx -0.571 cos α = 169 35 − 12 120 ≈ 169 35 − 131.45 ≈ − 0.571 : this gives sin α = 7 − 5 ( − 0.571 ) 12 = 9.855 12 > 0 \sin\alpha = \frac{7 - 5(-0.571)}{12} = \frac{9.855}{12} > 0 sin α = 12 7 − 5 ( − 0.571 ) = 12 9.855 > 0 , consistent with α \alpha α in the second quadrant.
cos α = 35 + 12 120 169 ≈ 166.45 169 ≈ 0.985 \cos\alpha = \frac{35 + 12\sqrt{120}}{169} \approx \frac{166.45}{169} \approx 0.985 cos α = 169 35 + 12 120 ≈ 169 166.45 ≈ 0.985 : this gives sin α = 7 − 5 ( 0.985 ) 12 = 2.075 12 > 0 \sin\alpha = \frac{7 - 5(0.985)}{12} = \frac{2.075}{12} > 0 sin α = 12 7 − 5 ( 0.985 ) = 12 2.075 > 0 , consistent with α \alpha α in the first quadrant.
Both values are genuine solutions of the original equation.
Proving α < 3 π / 4 \alpha < 3\pi/4 α < 3 π /4 when π / 2 < α < π \pi/2 < \alpha < \pi π /2 < α < π . If π / 2 < α < π \pi/2 < \alpha < \pi π /2 < α < π , then sin α > 0 \sin\alpha > 0 sin α > 0 and cos α < 0 \cos\alpha < 0 cos α < 0 , so the relevant root is
cos α = 35 − 12 120 169 . \cos\alpha = \frac{35 - 12\sqrt{120}}{169}. cos α = 169 35 − 12 120 .
Since cos \cos cos is strictly decreasing on ( π / 2 , π ) (\pi/2, \pi) ( π /2 , π ) , the condition α < 3 π / 4 \alpha < 3\pi/4 α < 3 π /4 is equivalent to cos α > cos ( 3 π / 4 ) = − 2 2 \cos\alpha > \cos(3\pi/4) = -\frac{\sqrt{2}}{2} cos α > cos ( 3 π /4 ) = − 2 2 .
We need to show:
35 − 12 120 169 > − 2 2 . \frac{35 - 12\sqrt{120}}{169} > -\frac{\sqrt{2}}{2}. 169 35 − 12 120 > − 2 2 .
Multiply both sides by 338 = 2 × 169 > 0 338 = 2 \times 169 > 0 338 = 2 × 169 > 0 :
70 − 24 120 > − 169 2 . 70 - 24\sqrt{120} > -169\sqrt{2}. 70 − 24 120 > − 169 2 .
Rearranging:
70 + 169 2 > 24 120 . 70 + 169\sqrt{2} > 24\sqrt{120}. 70 + 169 2 > 24 120 .
Since both sides are positive, square both sides:
LHS 2 = 4900 + 2 × 70 × 169 2 + 28561 × 2 = 4900 + 23660 2 + 57122 = 62022 + 23660 2 . \text{LHS}^2 = 4900 + 2 \times 70 \times 169\sqrt{2} + 28561 \times 2 = 4900 + 23660\sqrt{2} + 57122 = 62022 + 23660\sqrt{2}. LHS 2 = 4900 + 2 × 70 × 169 2 + 28561 × 2 = 4900 + 23660 2 + 57122 = 62022 + 23660 2 .
RHS 2 = 576 × 120 = 69120. \text{RHS}^2 = 576 \times 120 = 69120. RHS 2 = 576 × 120 = 69120.
So we need 62022 + 23660 2 > 69120 62022 + 23660\sqrt{2} > 69120 62022 + 23660 2 > 69120 , i.e. 23660 2 > 7098 23660\sqrt{2} > 7098 23660 2 > 7098 .
Since 2 > 1.4 \sqrt{2} > 1.4 2 > 1.4 , we have 23660 2 > 23660 × 1.4 = 33124 > 7098 23660\sqrt{2} > 23660 \times 1.4 = 33124 > 7098 23660 2 > 23660 × 1.4 = 33124 > 7098 . This confirms the inequality.
Therefore cos α > − 2 2 \cos\alpha > -\frac{\sqrt{2}}{2} cos α > − 2 2 , which gives α < 3 π 4 \alpha < \frac{3\pi}{4} α < 4 3 π . ■ \qquad \blacksquare ■
Topic : 积分与换元法 / Integration and Substitution | Difficulty : Challenging | Marks : 20
6 Find d y d x \frac{dy}{dx} d x d y if
y = a x + b c x + d . (*) y = \frac{ax + b}{cx + d}. \qquad \text{(*)} y = c x + d a x + b . (*)
By using changes of variable of the form ( ∗ ) (*) ( ∗ ) , or otherwise, show that
∫ 0 1 1 ( x + 3 ) 2 ln ( x + 1 x + 3 ) d x = 1 6 ln 3 − 1 4 ln 2 − 1 12 , \int_{0}^{1} \frac{1}{(x + 3)^2} \ln \left( \frac{x + 1}{x + 3} \right) dx = \frac{1}{6} \ln 3 - \frac{1}{4} \ln 2 - \frac{1}{12}, ∫ 0 1 ( x + 3 ) 2 1 ln ( x + 3 x + 1 ) d x = 6 1 ln 3 − 4 1 ln 2 − 12 1 ,
and evaluate the integrals
∫ 0 1 1 ( x + 3 ) 2 ln ( x 2 + 3 x + 2 ( x + 3 ) 2 ) d x and ∫ 0 1 1 ( x + 3 ) 2 ln ( x + 1 x + 2 ) d x . \int_{0}^{1} \frac{1}{(x + 3)^2} \ln \left( \frac{x^2 + 3x + 2}{(x + 3)^2} \right) dx \quad \text{and} \quad \int_{0}^{1} \frac{1}{(x + 3)^2} \ln \left( \frac{x + 1}{x + 2} \right) dx. ∫ 0 1 ( x + 3 ) 2 1 ln ( ( x + 3 ) 2 x 2 + 3 x + 2 ) d x and ∫ 0 1 ( x + 3 ) 2 1 ln ( x + 2 x + 1 ) d x .
Model Solution
Part 1: Finding d y d x \frac{dy}{dx} d x d y
For y = a x + b c x + d y = \dfrac{ax + b}{cx + d} y = c x + d a x + b , apply the quotient rule:
d y d x = a ( c x + d ) − c ( a x + b ) ( c x + d ) 2 = a c x + a d − a c x − b c ( c x + d ) 2 = a d − b c ( c x + d ) 2 . ( † ) \frac{dy}{dx} = \frac{a(cx + d) - c(ax + b)}{(cx + d)^2} = \frac{acx + ad - acx - bc}{(cx + d)^2} = \frac{ad - bc}{(cx + d)^2}. \qquad (\dagger) d x d y = ( c x + d ) 2 a ( c x + d ) − c ( a x + b ) = ( c x + d ) 2 a c x + a d − a c x − b c = ( c x + d ) 2 a d − b c . ( † )
This shows that the derivative of a linear fractional function takes the form k ( c x + d ) 2 \dfrac{k}{(cx+d)^2} ( c x + d ) 2 k where k = a d − b c k = ad - bc k = a d − b c is a constant — exactly the factor appearing in each integral below.
Part 2: Evaluating I 1 = ∫ 0 1 1 ( x + 3 ) 2 ln ( x + 1 x + 3 ) d x I_1 = \displaystyle\int_{0}^{1} \frac{1}{(x+3)^2} \ln\!\left(\frac{x+1}{x+3}\right) dx I 1 = ∫ 0 1 ( x + 3 ) 2 1 ln ( x + 3 x + 1 ) d x
The form of ( † ) (\dagger) ( † ) suggests the substitution u = x + 1 x + 3 u = \dfrac{x+1}{x+3} u = x + 3 x + 1 (a linear fractional function with a = 1 , b = 1 , c = 1 , d = 3 a = 1, b = 1, c = 1, d = 3 a = 1 , b = 1 , c = 1 , d = 3 ).
By ( † ) (\dagger) ( † ) :
d u d x = 1 ⋅ 3 − 1 ⋅ 1 ( x + 3 ) 2 = 2 ( x + 3 ) 2 , \frac{du}{dx} = \frac{1 \cdot 3 - 1 \cdot 1}{(x+3)^2} = \frac{2}{(x+3)^2}, d x d u = ( x + 3 ) 2 1 ⋅ 3 − 1 ⋅ 1 = ( x + 3 ) 2 2 ,
so 1 ( x + 3 ) 2 d x = 1 2 d u \dfrac{1}{(x+3)^2}\, dx = \dfrac{1}{2}\, du ( x + 3 ) 2 1 d x = 2 1 d u .
Changing limits: when x = 0 x = 0 x = 0 , u = 1 3 u = \frac{1}{3} u = 3 1 ; when x = 1 x = 1 x = 1 , u = 1 2 u = \frac{1}{2} u = 2 1 .
I 1 = 1 2 ∫ 1 / 3 1 / 2 ln u d u . I_1 = \frac{1}{2}\int_{1/3}^{1/2} \ln u\, du. I 1 = 2 1 ∫ 1/3 1/2 ln u d u .
Integrating by parts (∫ ln u d u = u ln u − u \int \ln u\, du = u\ln u - u ∫ ln u d u = u ln u − u ):
I 1 = 1 2 [ u ln u − u ] 1 / 3 1 / 2 . I_1 = \frac{1}{2}\Big[u\ln u - u\Big]_{1/3}^{1/2}. I 1 = 2 1 [ u ln u − u ] 1/3 1/2 .
At u = 1 2 u = \frac{1}{2} u = 2 1 :
1 2 ln 1 2 − 1 2 = − 1 2 ln 2 − 1 2 . \frac{1}{2}\ln\frac{1}{2} - \frac{1}{2} = -\frac{1}{2}\ln 2 - \frac{1}{2}. 2 1 ln 2 1 − 2 1 = − 2 1 ln 2 − 2 1 .
At u = 1 3 u = \frac{1}{3} u = 3 1 :
1 3 ln 1 3 − 1 3 = − 1 3 ln 3 − 1 3 . \frac{1}{3}\ln\frac{1}{3} - \frac{1}{3} = -\frac{1}{3}\ln 3 - \frac{1}{3}. 3 1 ln 3 1 − 3 1 = − 3 1 ln 3 − 3 1 .
Difference:
( − 1 2 ln 2 − 1 2 ) − ( − 1 3 ln 3 − 1 3 ) = 1 3 ln 3 − 1 2 ln 2 − 1 6 . \left(-\frac{1}{2}\ln 2 - \frac{1}{2}\right) - \left(-\frac{1}{3}\ln 3 - \frac{1}{3}\right) = \frac{1}{3}\ln 3 - \frac{1}{2}\ln 2 - \frac{1}{6}. ( − 2 1 ln 2 − 2 1 ) − ( − 3 1 ln 3 − 3 1 ) = 3 1 ln 3 − 2 1 ln 2 − 6 1 .
Therefore:
I 1 = 1 2 ( 1 3 ln 3 − 1 2 ln 2 − 1 6 ) = 1 6 ln 3 − 1 4 ln 2 − 1 12 . (shown) I_1 = \frac{1}{2}\left(\frac{1}{3}\ln 3 - \frac{1}{2}\ln 2 - \frac{1}{6}\right) = \frac{1}{6}\ln 3 - \frac{1}{4}\ln 2 - \frac{1}{12}. \qquad \text{(shown)} I 1 = 2 1 ( 3 1 ln 3 − 2 1 ln 2 − 6 1 ) = 6 1 ln 3 − 4 1 ln 2 − 12 1 . (shown)
Part 3: Evaluating I 2 = ∫ 0 1 1 ( x + 3 ) 2 ln ( x 2 + 3 x + 2 ( x + 3 ) 2 ) d x I_2 = \displaystyle\int_{0}^{1} \frac{1}{(x+3)^2} \ln\!\left(\frac{x^2+3x+2}{(x+3)^2}\right) dx I 2 = ∫ 0 1 ( x + 3 ) 2 1 ln ( ( x + 3 ) 2 x 2 + 3 x + 2 ) d x
Factor x 2 + 3 x + 2 = ( x + 1 ) ( x + 2 ) x^2 + 3x + 2 = (x+1)(x+2) x 2 + 3 x + 2 = ( x + 1 ) ( x + 2 ) and use ln A B C 2 = ln A C + ln B C \ln\dfrac{AB}{C^2} = \ln\dfrac{A}{C} + \ln\dfrac{B}{C} ln C 2 A B = ln C A + ln C B :
ln ( x + 1 ) ( x + 2 ) ( x + 3 ) 2 = ln x + 1 x + 3 + ln x + 2 x + 3 . \ln\frac{(x+1)(x+2)}{(x+3)^2} = \ln\frac{x+1}{x+3} + \ln\frac{x+2}{x+3}. ln ( x + 3 ) 2 ( x + 1 ) ( x + 2 ) = ln x + 3 x + 1 + ln x + 3 x + 2 .
So I 2 = I 1 + J I_2 = I_1 + J I 2 = I 1 + J where
J = ∫ 0 1 1 ( x + 3 ) 2 ln x + 2 x + 3 d x . J = \int_{0}^{1} \frac{1}{(x+3)^2}\ln\frac{x+2}{x+3}\, dx. J = ∫ 0 1 ( x + 3 ) 2 1 ln x + 3 x + 2 d x .
For J J J , substitute v = x + 2 x + 3 v = \dfrac{x+2}{x+3} v = x + 3 x + 2 (with a = 1 , b = 2 , c = 1 , d = 3 a=1, b=2, c=1, d=3 a = 1 , b = 2 , c = 1 , d = 3 ). By ( † ) (\dagger) ( † ) :
d v d x = 3 − 2 ( x + 3 ) 2 = 1 ( x + 3 ) 2 , \frac{dv}{dx} = \frac{3 - 2}{(x+3)^2} = \frac{1}{(x+3)^2}, d x d v = ( x + 3 ) 2 3 − 2 = ( x + 3 ) 2 1 ,
so 1 ( x + 3 ) 2 d x = d v \dfrac{1}{(x+3)^2}\, dx = dv ( x + 3 ) 2 1 d x = d v .
Limits: x = 0 ⇒ v = 2 3 x = 0 \Rightarrow v = \frac{2}{3} x = 0 ⇒ v = 3 2 ; x = 1 ⇒ v = 3 4 x = 1 \Rightarrow v = \frac{3}{4} x = 1 ⇒ v = 4 3 .
J = ∫ 2 / 3 3 / 4 ln v d v = [ v ln v − v ] 2 / 3 3 / 4 . J = \int_{2/3}^{3/4} \ln v\, dv = \Big[v\ln v - v\Big]_{2/3}^{3/4}. J = ∫ 2/3 3/4 ln v d v = [ v ln v − v ] 2/3 3/4 .
At v = 3 4 v = \frac{3}{4} v = 4 3 :
3 4 ln 3 4 − 3 4 = 3 4 ln 3 − 3 4 ln 2 − 3 4 . \frac{3}{4}\ln\frac{3}{4} - \frac{3}{4} = \frac{3}{4}\ln 3 - \frac{3}{4}\ln 2 - \frac{3}{4}. 4 3 ln 4 3 − 4 3 = 4 3 ln 3 − 4 3 ln 2 − 4 3 .
At v = 2 3 v = \frac{2}{3} v = 3 2 :
2 3 ln 2 3 − 2 3 = 2 3 ln 2 − 2 3 ln 3 − 2 3 . \frac{2}{3}\ln\frac{2}{3} - \frac{2}{3} = \frac{2}{3}\ln 2 - \frac{2}{3}\ln 3 - \frac{2}{3}. 3 2 ln 3 2 − 3 2 = 3 2 ln 2 − 3 2 ln 3 − 3 2 .
Difference:
( 3 4 ln 3 − 3 4 ln 2 − 3 4 ) − ( 2 3 ln 2 − 2 3 ln 3 − 2 3 ) \left(\frac{3}{4}\ln 3 - \frac{3}{4}\ln 2 - \frac{3}{4}\right) - \left(\frac{2}{3}\ln 2 - \frac{2}{3}\ln 3 - \frac{2}{3}\right) ( 4 3 ln 3 − 4 3 ln 2 − 4 3 ) − ( 3 2 ln 2 − 3 2 ln 3 − 3 2 )
= ( 3 4 + 2 3 ) ln 3 − ( 3 4 + 2 3 ) ln 2 − ( 3 4 − 2 3 ) = \left(\frac{3}{4} + \frac{2}{3}\right)\ln 3 - \left(\frac{3}{4} + \frac{2}{3}\right)\ln 2 - \left(\frac{3}{4} - \frac{2}{3}\right) = ( 4 3 + 3 2 ) ln 3 − ( 4 3 + 3 2 ) ln 2 − ( 4 3 − 3 2 )
= 17 12 ln 3 − 17 12 ln 2 − 1 12 . = \frac{17}{12}\ln 3 - \frac{17}{12}\ln 2 - \frac{1}{12}. = 12 17 ln 3 − 12 17 ln 2 − 12 1 .
Therefore:
I 2 = I 1 + J = ( 1 6 ln 3 − 1 4 ln 2 − 1 12 ) + ( 17 12 ln 3 − 17 12 ln 2 − 1 12 ) I_2 = I_1 + J = \left(\frac{1}{6}\ln 3 - \frac{1}{4}\ln 2 - \frac{1}{12}\right) + \left(\frac{17}{12}\ln 3 - \frac{17}{12}\ln 2 - \frac{1}{12}\right) I 2 = I 1 + J = ( 6 1 ln 3 − 4 1 ln 2 − 12 1 ) + ( 12 17 ln 3 − 12 17 ln 2 − 12 1 )
= ( 2 12 + 17 12 ) ln 3 − ( 3 12 + 17 12 ) ln 2 − 2 12 = 19 12 ln 3 − 5 3 ln 2 − 1 6 . = \left(\frac{2}{12} + \frac{17}{12}\right)\ln 3 - \left(\frac{3}{12} + \frac{17}{12}\right)\ln 2 - \frac{2}{12} = \frac{19}{12}\ln 3 - \frac{5}{3}\ln 2 - \frac{1}{6}. = ( 12 2 + 12 17 ) ln 3 − ( 12 3 + 12 17 ) ln 2 − 12 2 = 12 19 ln 3 − 3 5 ln 2 − 6 1 .
Part 4: Evaluating I 3 = ∫ 0 1 1 ( x + 3 ) 2 ln ( x + 1 x + 2 ) d x I_3 = \displaystyle\int_{0}^{1} \frac{1}{(x+3)^2} \ln\!\left(\frac{x+1}{x+2}\right) dx I 3 = ∫ 0 1 ( x + 3 ) 2 1 ln ( x + 2 x + 1 ) d x
Use ln x + 1 x + 2 = ln x + 1 x + 3 − ln x + 2 x + 3 \ln\dfrac{x+1}{x+2} = \ln\dfrac{x+1}{x+3} - \ln\dfrac{x+2}{x+3} ln x + 2 x + 1 = ln x + 3 x + 1 − ln x + 3 x + 2 , so I 3 = I 1 − J I_3 = I_1 - J I 3 = I 1 − J :
I 3 = ( 1 6 ln 3 − 1 4 ln 2 − 1 12 ) − ( 17 12 ln 3 − 17 12 ln 2 − 1 12 ) I_3 = \left(\frac{1}{6}\ln 3 - \frac{1}{4}\ln 2 - \frac{1}{12}\right) - \left(\frac{17}{12}\ln 3 - \frac{17}{12}\ln 2 - \frac{1}{12}\right) I 3 = ( 6 1 ln 3 − 4 1 ln 2 − 12 1 ) − ( 12 17 ln 3 − 12 17 ln 2 − 12 1 )
= ( 2 12 − 17 12 ) ln 3 + ( − 3 12 + 17 12 ) ln 2 + ( − 1 12 + 1 12 ) = \left(\frac{2}{12} - \frac{17}{12}\right)\ln 3 + \left(-\frac{3}{12} + \frac{17}{12}\right)\ln 2 + \left(-\frac{1}{12} + \frac{1}{12}\right) = ( 12 2 − 12 17 ) ln 3 + ( − 12 3 + 12 17 ) ln 2 + ( − 12 1 + 12 1 )
= − 5 4 ln 3 + 7 6 ln 2. = -\frac{5}{4}\ln 3 + \frac{7}{6}\ln 2. = − 4 5 ln 3 + 6 7 ln 2.
Summary of results:
∫ 0 1 1 ( x + 3 ) 2 ln ( x + 1 x + 3 ) d x = 1 6 ln 3 − 1 4 ln 2 − 1 12 \int_{0}^{1} \frac{1}{(x+3)^2} \ln\!\left(\frac{x+1}{x+3}\right) dx = \frac{1}{6}\ln 3 - \frac{1}{4}\ln 2 - \frac{1}{12} ∫ 0 1 ( x + 3 ) 2 1 ln ( x + 3 x + 1 ) d x = 6 1 ln 3 − 4 1 ln 2 − 12 1
∫ 0 1 1 ( x + 3 ) 2 ln ( x 2 + 3 x + 2 ( x + 3 ) 2 ) d x = 19 12 ln 3 − 5 3 ln 2 − 1 6 \int_{0}^{1} \frac{1}{(x+3)^2} \ln\!\left(\frac{x^2+3x+2}{(x+3)^2}\right) dx = \frac{19}{12}\ln 3 - \frac{5}{3}\ln 2 - \frac{1}{6} ∫ 0 1 ( x + 3 ) 2 1 ln ( ( x + 3 ) 2 x 2 + 3 x + 2 ) d x = 12 19 ln 3 − 3 5 ln 2 − 6 1
∫ 0 1 1 ( x + 3 ) 2 ln ( x + 1 x + 2 ) d x = − 5 4 ln 3 + 7 6 ln 2 \int_{0}^{1} \frac{1}{(x+3)^2} \ln\!\left(\frac{x+1}{x+2}\right) dx = -\frac{5}{4}\ln 3 + \frac{7}{6}\ln 2 ∫ 0 1 ( x + 3 ) 2 1 ln ( x + 2 x + 1 ) d x = − 4 5 ln 3 + 6 7 ln 2
Topic : 曲线分析与作图 / Curve Analysis and Sketching | Difficulty : Challenging | Marks : 20
7 The curve C C C has equation
y = x ( x 2 − 2 x + a ) , y = \frac{x}{\sqrt{(x^2 - 2x + a)}} , y = ( x 2 − 2 x + a ) x ,
where the square root is positive. Show that, if a > 1 a > 1 a > 1 , then C C C has exactly one stationary point.
Sketch C C C when (i) a = 2 a = 2 a = 2 and (ii) a = 1 a = 1 a = 1 .
Model Solution
We have y = x x 2 − 2 x + a y = \dfrac{x}{\sqrt{x^2 - 2x + a}} y = x 2 − 2 x + a x , where the square root is positive.
Finding stationary points. Let u = x 2 − 2 x + a u = x^2 - 2x + a u = x 2 − 2 x + a , so y = x u − 1 / 2 y = x \, u^{-1/2} y = x u − 1/2 . By the product rule and chain rule:
d y d x = u − 1 / 2 + x ⋅ ( − 1 2 ) u − 3 / 2 ⋅ ( 2 x − 2 ) = 1 u − x ( x − 1 ) u 3 / 2 \frac{dy}{dx} = u^{-1/2} + x \cdot \left(-\frac{1}{2}\right) u^{-3/2} \cdot (2x - 2) = \frac{1}{\sqrt{u}} - \frac{x(x-1)}{u^{3/2}} d x d y = u − 1/2 + x ⋅ ( − 2 1 ) u − 3/2 ⋅ ( 2 x − 2 ) = u 1 − u 3/2 x ( x − 1 )
Combining over the common denominator u 3 / 2 u^{3/2} u 3/2 :
d y d x = u − x ( x − 1 ) u 3 / 2 = ( x 2 − 2 x + a ) − ( x 2 − x ) u 3 / 2 = a − x ( x 2 − 2 x + a ) 3 / 2 \frac{dy}{dx} = \frac{u - x(x-1)}{u^{3/2}} = \frac{(x^2 - 2x + a) - (x^2 - x)}{u^{3/2}} = \frac{a - x}{(x^2 - 2x + a)^{3/2}} d x d y = u 3/2 u − x ( x − 1 ) = u 3/2 ( x 2 − 2 x + a ) − ( x 2 − x ) = ( x 2 − 2 x + a ) 3/2 a − x
Stationary points require d y / d x = 0 dy/dx = 0 d y / d x = 0 , so a − x = 0 a - x = 0 a − x = 0 , giving x = a x = a x = a . We must also check that this value lies in the domain, i.e., u = x 2 − 2 x + a > 0 u = x^2 - 2x + a > 0 u = x 2 − 2 x + a > 0 .
When a > 1 a > 1 a > 1 , the discriminant of x 2 − 2 x + a x^2 - 2x + a x 2 − 2 x + a is 4 − 4 a < 0 4 - 4a < 0 4 − 4 a < 0 , so u > 0 u > 0 u > 0 for all x x x . Thus x = a x = a x = a is in the domain. Since d y / d x = 0 dy/dx = 0 d y / d x = 0 only when x = a x = a x = a (the denominator is always positive), C C C has exactly one stationary point .
Value at the stationary point. At x = a x = a x = a :
y ( a ) = a a 2 − 2 a + a = a a 2 − a = a a ( a − 1 ) = a a − 1 y(a) = \frac{a}{\sqrt{a^2 - 2a + a}} = \frac{a}{\sqrt{a^2 - a}} = \frac{a}{\sqrt{a(a-1)}} = \frac{\sqrt{a}}{\sqrt{a-1}} y ( a ) = a 2 − 2 a + a a = a 2 − a a = a ( a − 1 ) a = a − 1 a
Asymptotic behaviour. For large ∣ x ∣ |x| ∣ x ∣ :
y = x x 2 − 2 x + a = x ∣ x ∣ 1 − 2 / x + a / x 2 → x ∣ x ∣ y = \frac{x}{\sqrt{x^2 - 2x + a}} = \frac{x}{|x|\sqrt{1 - 2/x + a/x^2}} \to \frac{x}{|x|} y = x 2 − 2 x + a x = ∣ x ∣ 1 − 2/ x + a / x 2 x → ∣ x ∣ x
So y → 1 y \to 1 y → 1 as x → + ∞ x \to +\infty x → + ∞ and y → − 1 y \to -1 y → − 1 as x → − ∞ x \to -\infty x → − ∞ .
Part (i): a = 2 a = 2 a = 2 . Then y = x x 2 − 2 x + 2 y = \dfrac{x}{\sqrt{x^2 - 2x + 2}} y = x 2 − 2 x + 2 x , defined for all x x x since x 2 − 2 x + 2 = ( x − 1 ) 2 + 1 > 0 x^2 - 2x + 2 = (x-1)^2 + 1 > 0 x 2 − 2 x + 2 = ( x − 1 ) 2 + 1 > 0 .
The stationary point is at x = 2 x = 2 x = 2 , y = 2 ≈ 1.41 y = \sqrt{2} \approx 1.41 y = 2 ≈ 1.41 . Checking the sign of d y / d x = ( 2 − x ) / u 3 / 2 dy/dx = (2-x)/u^{3/2} d y / d x = ( 2 − x ) / u 3/2 : positive for x < 2 x < 2 x < 2 (increasing) and negative for x > 2 x > 2 x > 2 (decreasing), so this is a maximum.
For x < 0 x < 0 x < 0 : x 2 − 2 x + 2 > x 2 x^2 - 2x + 2 > x^2 x 2 − 2 x + 2 > x 2 (since − 2 x > 0 -2x > 0 − 2 x > 0 ), so ∣ y ∣ < 1 |y| < 1 ∣ y ∣ < 1 . The curve approaches y = − 1 y = -1 y = − 1 from below as x → − ∞ x \to -\infty x → − ∞ and rises through the origin.
For x > 0 x > 0 x > 0 and large: y = x x 2 − 2 x + 2 > 1 y = \frac{x}{\sqrt{x^2-2x+2}} > 1 y = x 2 − 2 x + 2 x > 1 (since x 2 > x 2 − 2 x + 2 x^2 > x^2 - 2x + 2 x 2 > x 2 − 2 x + 2 when x > 1 x > 1 x > 1 ), and y → 1 y \to 1 y → 1 from above as x → + ∞ x \to +\infty x → + ∞ . The curve rises from the origin, reaches a maximum of 2 \sqrt{2} 2 at x = 2 x = 2 x = 2 , then decreases toward y = 1 y = 1 y = 1 from above.
Sketch for a = 2 a = 2 a = 2 : the curve passes through the origin, rises to a maximum of 2 \sqrt{2} 2 at x = 2 x = 2 x = 2 , then decreases toward the horizontal asymptote y = 1 y = 1 y = 1 (approached from above) as x → ∞ x \to \infty x → ∞ . For negative x x x , the curve rises from the asymptote y = − 1 y = -1 y = − 1 (approached from below) through the origin.
Part (ii): a = 1 a = 1 a = 1 . Then y = x x 2 − 2 x + 1 = x ( x − 1 ) 2 = x ∣ x − 1 ∣ y = \dfrac{x}{\sqrt{x^2 - 2x + 1}} = \dfrac{x}{\sqrt{(x-1)^2}} = \dfrac{x}{|x - 1|} y = x 2 − 2 x + 1 x = ( x − 1 ) 2 x = ∣ x − 1∣ x , defined for x ≠ 1 x \neq 1 x = 1 .
For x < 1 x < 1 x < 1 : y = x 1 − x = − 1 + 1 1 − x y = \dfrac{x}{1 - x} = -1 + \dfrac{1}{1 - x} y = 1 − x x = − 1 + 1 − x 1 , which is increasing (from y → − 1 y \to -1 y → − 1 as x → − ∞ x \to -\infty x → − ∞ to y → − ∞ y \to -\infty y → − ∞ as x → 1 − x \to 1^- x → 1 − ).
For x > 1 x > 1 x > 1 : y = x x − 1 = 1 + 1 x − 1 y = \dfrac{x}{x - 1} = 1 + \dfrac{1}{x - 1} y = x − 1 x = 1 + x − 1 1 , which is decreasing (from y → + ∞ y \to +\infty y → + ∞ as x → 1 + x \to 1^+ x → 1 + to y → 1 y \to 1 y → 1 as x → + ∞ x \to +\infty x → + ∞ ).
There is a vertical asymptote at x = 1 x = 1 x = 1 , with y → − ∞ y \to -\infty y → − ∞ from the left and y → + ∞ y \to +\infty y → + ∞ from the right. The horizontal asymptotes are y = − 1 y = -1 y = − 1 (as x → − ∞ x \to -\infty x → − ∞ ) and y = 1 y = 1 y = 1 (as x → + ∞ x \to +\infty x → + ∞ ). There are no stationary points (the derivative 1 ( x − 1 ) 2 > 0 \frac{1}{(x-1)^2} > 0 ( x − 1 ) 2 1 > 0 on each piece). The curve passes through the origin.
Topic : 三角级数与求和 / Trigonometric Series and Summation | Difficulty : Hard | Marks : 20
8 Prove that
∑ k = 0 n sin k θ = cos 1 2 θ − cos ( n + 1 2 ) θ 2 sin 1 2 θ . (*) \sum_{k=0}^{n} \sin k\theta = \frac{\cos \frac{1}{2}\theta - \cos(n + \frac{1}{2})\theta}{2 \sin \frac{1}{2}\theta} . \qquad \text{(*)} ∑ k = 0 n sin k θ = 2 s i n 2 1 θ c o s 2 1 θ − c o s ( n + 2 1 ) θ . (*)
(i) Deduce that, when n n n is large,
∑ k = 0 n sin ( k π n ) ≈ 2 n π . \sum_{k=0}^{n} \sin \left( \frac{k\pi}{n} \right) \approx \frac{2n}{\pi} . ∑ k = 0 n sin ( n k π ) ≈ π 2 n .
(ii) By differentiating ( ∗ ) (*) ( ∗ ) with respect to θ \theta θ , or otherwise, show that, when n n n is large,
∑ k = 0 n k sin 2 ( k π 2 n ) ≈ ( 1 4 + 1 π 2 ) n 2 . \sum_{k=0}^{n} k \sin^2 \left( \frac{k\pi}{2n} \right) \approx \left( \frac{1}{4} + \frac{1}{\pi^2} \right) n^2 . ∑ k = 0 n k sin 2 ( 2 n k π ) ≈ ( 4 1 + π 2 1 ) n 2 .
[The approximations, valid for small θ \theta θ , sin θ ≈ θ \sin \theta \approx \theta sin θ ≈ θ and cos θ ≈ 1 − 1 2 θ 2 \cos \theta \approx 1 - \frac{1}{2}\theta^2 cos θ ≈ 1 − 2 1 θ 2 may be assumed.]
Model Solution
Proof of the summation formula. Multiply both sides of the identity by 2 sin θ 2 2\sin\frac{\theta}{2} 2 sin 2 θ (assuming sin θ 2 ≠ 0 \sin\frac{\theta}{2} \neq 0 sin 2 θ = 0 ). The left side becomes
2 sin θ 2 ∑ k = 0 n sin k θ = ∑ k = 0 n 2 sin k θ sin θ 2 . 2\sin\tfrac{\theta}{2} \sum_{k=0}^{n} \sin k\theta = \sum_{k=0}^{n} 2\sin k\theta \sin\tfrac{\theta}{2}. 2 sin 2 θ ∑ k = 0 n sin k θ = ∑ k = 0 n 2 sin k θ sin 2 θ .
Using the product-to-sum identity 2 sin A sin B = cos ( A − B ) − cos ( A + B ) 2\sin A\sin B = \cos(A-B) - \cos(A+B) 2 sin A sin B = cos ( A − B ) − cos ( A + B ) :
2 sin k θ sin θ 2 = cos ( k − 1 2 ) θ − cos ( k + 1 2 ) θ . 2\sin k\theta \sin\tfrac{\theta}{2} = \cos\!\bigl(k - \tfrac{1}{2}\bigr)\theta - \cos\!\bigl(k + \tfrac{1}{2}\bigr)\theta. 2 sin k θ sin 2 θ = cos ( k − 2 1 ) θ − cos ( k + 2 1 ) θ .
The sum telescopes:
k = 0 : cos θ 2 − cos θ 2 = 0 ( since cos ( − θ 2 ) = cos θ 2 ) k = 0: \quad \cos\tfrac{\theta}{2} - \cos\tfrac{\theta}{2} = 0 \quad (\text{since } \cos(-\tfrac{\theta}{2}) = \cos\tfrac{\theta}{2}) k = 0 : cos 2 θ − cos 2 θ = 0 ( since cos ( − 2 θ ) = cos 2 θ )
k = 1 : cos θ 2 − cos 3 θ 2 k = 1: \quad \cos\tfrac{\theta}{2} - \cos\tfrac{3\theta}{2} k = 1 : cos 2 θ − cos 2 3 θ
k = 2 : cos 3 θ 2 − cos 5 θ 2 k = 2: \quad \cos\tfrac{3\theta}{2} - \cos\tfrac{5\theta}{2} k = 2 : cos 2 3 θ − cos 2 5 θ
⋮ \vdots ⋮
k = n : cos ( n − 1 2 ) θ − cos ( n + 1 2 ) θ k = n: \quad \cos\!\bigl(n - \tfrac{1}{2}\bigr)\theta - \cos\!\bigl(n + \tfrac{1}{2}\bigr)\theta k = n : cos ( n − 2 1 ) θ − cos ( n + 2 1 ) θ
All intermediate terms cancel, leaving
2 sin θ 2 ∑ k = 0 n sin k θ = cos θ 2 − cos ( n + 1 2 ) θ . 2\sin\tfrac{\theta}{2} \sum_{k=0}^{n} \sin k\theta = \cos\tfrac{\theta}{2} - \cos\!\bigl(n + \tfrac{1}{2}\bigr)\theta. 2 sin 2 θ ∑ k = 0 n sin k θ = cos 2 θ − cos ( n + 2 1 ) θ .
Dividing both sides by 2 sin θ 2 2\sin\frac{\theta}{2} 2 sin 2 θ :
∑ k = 0 n sin k θ = cos θ 2 − cos ( n + 1 2 ) θ 2 sin θ 2 . ■ \sum_{k=0}^{n} \sin k\theta = \frac{\cos\frac{\theta}{2} - \cos\!\bigl(n + \frac{1}{2}\bigr)\theta}{2\sin\frac{\theta}{2}}. \qquad \blacksquare ∑ k = 0 n sin k θ = 2 s i n 2 θ c o s 2 θ − c o s ( n + 2 1 ) θ . ■
Part (i)
Set θ = π n \theta = \dfrac{\pi}{n} θ = n π in the formula:
∑ k = 0 n sin k π n = cos π 2 n − cos ( n + 1 2 ) π n 2 sin π 2 n . \sum_{k=0}^{n} \sin\frac{k\pi}{n} = \frac{\cos\frac{\pi}{2n} - \cos\!\bigl(n + \frac{1}{2}\bigr)\frac{\pi}{n}}{2\sin\frac{\pi}{2n}}. ∑ k = 0 n sin n k π = 2 s i n 2 n π c o s 2 n π − c o s ( n + 2 1 ) n π .
Since ( n + 1 2 ) π n = π + π 2 n \bigl(n + \frac{1}{2}\bigr)\frac{\pi}{n} = \pi + \frac{\pi}{2n} ( n + 2 1 ) n π = π + 2 n π , we have
cos ( π + π 2 n ) = − cos π 2 n , \cos\!\bigl(\pi + \tfrac{\pi}{2n}\bigr) = -\cos\tfrac{\pi}{2n}, cos ( π + 2 n π ) = − cos 2 n π ,
so the numerator becomes cos π 2 n − ( − cos π 2 n ) = 2 cos π 2 n \cos\frac{\pi}{2n} - \bigl(-\cos\frac{\pi}{2n}\bigr) = 2\cos\frac{\pi}{2n} cos 2 n π − ( − cos 2 n π ) = 2 cos 2 n π , giving
∑ k = 0 n sin k π n = 2 cos π 2 n 2 sin π 2 n = cot π 2 n . ( † ) \sum_{k=0}^{n} \sin\frac{k\pi}{n} = \frac{2\cos\frac{\pi}{2n}}{2\sin\frac{\pi}{2n}} = \cot\frac{\pi}{2n}. \qquad (\dagger) ∑ k = 0 n sin n k π = 2 s i n 2 n π 2 c o s 2 n π = cot 2 n π . ( † )
For large n n n , π 2 n \frac{\pi}{2n} 2 n π is small. Using sin ϕ ≈ ϕ \sin\phi \approx \phi sin ϕ ≈ ϕ and cos ϕ ≈ 1 − ϕ 2 2 \cos\phi \approx 1 - \frac{\phi^2}{2} cos ϕ ≈ 1 − 2 ϕ 2 with ϕ = π 2 n \phi = \frac{\pi}{2n} ϕ = 2 n π :
cot π 2 n = cos π 2 n sin π 2 n ≈ 1 − π 2 8 n 2 π 2 n = 2 n π ( 1 − π 2 8 n 2 ) ≈ 2 n π . ■ \cot\frac{\pi}{2n} = \frac{\cos\frac{\pi}{2n}}{\sin\frac{\pi}{2n}} \approx \frac{1 - \frac{\pi^2}{8n^2}}{\frac{\pi}{2n}} = \frac{2n}{\pi}\!\left(1 - \frac{\pi^2}{8n^2}\right) \approx \frac{2n}{\pi}. \qquad \blacksquare cot 2 n π = s i n 2 n π c o s 2 n π ≈ 2 n π 1 − 8 n 2 π 2 = π 2 n ( 1 − 8 n 2 π 2 ) ≈ π 2 n . ■
Part (ii)
We need ∑ k = 0 n k sin 2 k π 2 n \displaystyle\sum_{k=0}^{n} k\sin^2\frac{k\pi}{2n} k = 0 ∑ n k sin 2 2 n k π . Using sin 2 ϕ = 1 − cos 2 ϕ 2 \sin^2\phi = \frac{1 - \cos 2\phi}{2} sin 2 ϕ = 2 1 − c o s 2 ϕ with ϕ = k π 2 n \phi = \frac{k\pi}{2n} ϕ = 2 n k π :
∑ k = 0 n k sin 2 k π 2 n = 1 2 ∑ k = 0 n k − 1 2 ∑ k = 0 n k cos k π n . ( ‡ ) \sum_{k=0}^{n} k\sin^2\frac{k\pi}{2n} = \frac{1}{2}\sum_{k=0}^{n} k - \frac{1}{2}\sum_{k=0}^{n} k\cos\frac{k\pi}{n}. \qquad (\ddagger) ∑ k = 0 n k sin 2 2 n k π = 2 1 ∑ k = 0 n k − 2 1 ∑ k = 0 n k cos n k π . ( ‡ )
First sum.
∑ k = 0 n k = n ( n + 1 ) 2 ≈ n 2 2 . \sum_{k=0}^{n} k = \frac{n(n+1)}{2} \approx \frac{n^2}{2}. ∑ k = 0 n k = 2 n ( n + 1 ) ≈ 2 n 2 .
Second sum. We compute ∑ k = 0 n k cos k θ \sum_{k=0}^{n} k\cos k\theta ∑ k = 0 n k cos k θ by differentiating ( ∗ ) (*) ( ∗ ) with respect to θ \theta θ . Write
S ( θ ) = ∑ k = 0 n sin k θ = cos θ 2 − cos ( n + 1 2 ) θ 2 sin θ 2 . S(\theta) = \sum_{k=0}^{n} \sin k\theta = \frac{\cos\frac{\theta}{2} - \cos(n+\frac{1}{2})\theta}{2\sin\frac{\theta}{2}}. S ( θ ) = ∑ k = 0 n sin k θ = 2 s i n 2 θ c o s 2 θ − c o s ( n + 2 1 ) θ .
Then S ′ ( θ ) = ∑ k = 0 n k cos k θ S'(\theta) = \sum_{k=0}^{n} k\cos k\theta S ′ ( θ ) = ∑ k = 0 n k cos k θ . For the right side, let N = cos θ 2 − cos ( n + 1 2 ) θ N = \cos\frac{\theta}{2} - \cos(n+\frac{1}{2})\theta N = cos 2 θ − cos ( n + 2 1 ) θ and D = 2 sin θ 2 D = 2\sin\frac{\theta}{2} D = 2 sin 2 θ :
N ′ = − 1 2 sin θ 2 + ( n + 1 2 ) sin ( n + 1 2 ) θ , D ′ = cos θ 2 , N' = -\tfrac{1}{2}\sin\tfrac{\theta}{2} + (n+\tfrac{1}{2})\sin(n+\tfrac{1}{2})\theta, \qquad D' = \cos\tfrac{\theta}{2}, N ′ = − 2 1 sin 2 θ + ( n + 2 1 ) sin ( n + 2 1 ) θ , D ′ = cos 2 θ ,
S ′ ( θ ) = N ′ D − N D ′ D 2 . S'(\theta) = \frac{N'D - ND'}{D^2}. S ′ ( θ ) = D 2 N ′ D − N D ′ .
Now evaluate at θ = π n \theta = \frac{\pi}{n} θ = n π , where ( n + 1 2 ) θ = π + π 2 n (n+\frac{1}{2})\theta = \pi + \frac{\pi}{2n} ( n + 2 1 ) θ = π + 2 n π , giving sin ( n + 1 2 ) θ = − sin π 2 n \sin(n+\frac{1}{2})\theta = -\sin\frac{\pi}{2n} sin ( n + 2 1 ) θ = − sin 2 n π and cos ( n + 1 2 ) θ = − cos π 2 n \cos(n+\frac{1}{2})\theta = -\cos\frac{\pi}{2n} cos ( n + 2 1 ) θ = − cos 2 n π .
At θ = π n \theta = \frac{\pi}{n} θ = n π :
N = cos π 2 n − ( − cos π 2 n ) = 2 cos π 2 n , N = \cos\tfrac{\pi}{2n} - (-\cos\tfrac{\pi}{2n}) = 2\cos\tfrac{\pi}{2n}, N = cos 2 n π − ( − cos 2 n π ) = 2 cos 2 n π ,
N ′ = − 1 2 sin π 2 n + ( n + 1 2 ) ( − sin π 2 n ) = − ( n + 1 ) sin π 2 n , N' = -\tfrac{1}{2}\sin\tfrac{\pi}{2n} + (n+\tfrac{1}{2})(-\sin\tfrac{\pi}{2n}) = -(n+1)\sin\tfrac{\pi}{2n}, N ′ = − 2 1 sin 2 n π + ( n + 2 1 ) ( − sin 2 n π ) = − ( n + 1 ) sin 2 n π ,
D = 2 sin π 2 n , D ′ = cos π 2 n . D = 2\sin\tfrac{\pi}{2n}, \qquad D' = \cos\tfrac{\pi}{2n}. D = 2 sin 2 n π , D ′ = cos 2 n π .
Computing the pieces:
N ′ D = − 2 ( n + 1 ) sin 2 π 2 n , N D ′ = 2 cos 2 π 2 n , D 2 = 4 sin 2 π 2 n . N'D = -2(n+1)\sin^2\tfrac{\pi}{2n}, \qquad ND' = 2\cos^2\tfrac{\pi}{2n}, \qquad D^2 = 4\sin^2\tfrac{\pi}{2n}. N ′ D = − 2 ( n + 1 ) sin 2 2 n π , N D ′ = 2 cos 2 2 n π , D 2 = 4 sin 2 2 n π .
Therefore
∑ k = 0 n k cos k π n = − 2 ( n + 1 ) sin 2 π 2 n − 2 cos 2 π 2 n 4 sin 2 π 2 n = − n + 1 2 − 1 2 cot 2 π 2 n . ( § ′ ) \sum_{k=0}^n k\cos\frac{k\pi}{n} = \frac{-2(n+1)\sin^2\frac{\pi}{2n} - 2\cos^2\frac{\pi}{2n}}{4\sin^2\frac{\pi}{2n}} = -\frac{n+1}{2} - \frac{1}{2}\cot^2\frac{\pi}{2n}. \qquad (\S') ∑ k = 0 n k cos n k π = 4 s i n 2 2 n π − 2 ( n + 1 ) s i n 2 2 n π − 2 c o s 2 2 n π = − 2 n + 1 − 2 1 cot 2 2 n π . ( § ′ )
Approximating for large n n n . From part (i), cot π 2 n ≈ 2 n π \cot\frac{\pi}{2n} \approx \frac{2n}{\pi} cot 2 n π ≈ π 2 n , so cot 2 π 2 n ≈ 4 n 2 π 2 \cot^2\frac{\pi}{2n} \approx \frac{4n^2}{\pi^2} cot 2 2 n π ≈ π 2 4 n 2 . Thus
∑ k = 0 n k cos k π n ≈ − n 2 − 2 n 2 π 2 . \sum_{k=0}^n k\cos\frac{k\pi}{n} \approx -\frac{n}{2} - \frac{2n^2}{\pi^2}. ∑ k = 0 n k cos n k π ≈ − 2 n − π 2 2 n 2 .
Final assembly. Substituting into ( ‡ ) (\ddagger) ( ‡ ) :
∑ k = 0 n k sin 2 k π 2 n = 1 2 ∑ k = 0 n k − 1 2 ∑ k = 0 n k cos k π n \sum_{k=0}^n k\sin^2\frac{k\pi}{2n} = \frac{1}{2}\sum_{k=0}^n k - \frac{1}{2}\sum_{k=0}^n k\cos\frac{k\pi}{n} ∑ k = 0 n k sin 2 2 n k π = 2 1 ∑ k = 0 n k − 2 1 ∑ k = 0 n k cos n k π
≈ n 2 4 − 1 2 ( − n 2 − 2 n 2 π 2 ) = n 2 4 + n 4 + n 2 π 2 ≈ n 2 4 + n 2 π 2 = ( 1 4 + 1 π 2 ) n 2 . ■ \approx \frac{n^2}{4} - \frac{1}{2}\!\left(-\frac{n}{2} - \frac{2n^2}{\pi^2}\right) = \frac{n^2}{4} + \frac{n}{4} + \frac{n^2}{\pi^2} \approx \frac{n^2}{4} + \frac{n^2}{\pi^2} = \left(\frac{1}{4} + \frac{1}{\pi^2}\right)n^2. \qquad \blacksquare ≈ 4 n 2 − 2 1 ( − 2 n − π 2 2 n 2 ) = 4 n 2 + 4 n + π 2 n 2 ≈ 4 n 2 + π 2 n 2 = ( 4 1 + π 2 1 ) n 2 . ■