i i

“87-S1-sols” — 2020/2/4 — 21:50 — page 1 — #1

i i

87-S1-sols Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16

A Pure Mathematics

1 (i)
f 0 (x) = aeax cos(bx) − beax sin(bx)
= eax (a cos(bx) − b sin(bx))
=0
ax
e = 0 no solution
a cos(bx) − b sin(bx) = 0
1  a 
x= nπ + arctan
b b
1
xn = (nπ + c)
b
a


where c = arctan b
.

(ii) We have
nπ+c
f(xn ) = e b cos(nπ + c)
Consider
f(xn+1 ) ea((n+1)π+c)/b cos((n + 1)π + c) aπ cos((n + 1)π + c)
= a(nπ+c)/b
=eb
f(xn ) e cos(nπ + c) cos(nπ + c)
We know
cos(y + π) = cos(y) cos(π) − sin(π) sin(y) = − cos(y)
It follows
f(xn+1 ) aπ cos((n + 1)π + c) aπ
=eb = −e b
f(xn ) cos(nπ + c)
OLD SOLUTION:
   a 
1 a
(nπ+arctan( ab ))
  a 
f nπ + arctan =e b + cos nπ + arctan
b b b
anπ 1 a
a
= e b × e b arctan( b ) × (cos nπ) arctan
b
Now if n is even, a
e b × e b arctan( b ) × arctan
anπ 1 a
.
b
If n is odd the expression will be:
a
anπ 1
arctan( ab )
−e b ×e b × arctan .
b
Now consider the next turning point directly after the one shown above ie. (n + 1)π
1  a 
x= (n + 1)π + arctan .
b b
a
f(x) = e b × e b arctan( b ) × cos((n + 1)π) × arctan
a(n+1)π 1 a

b
which simplifies to
a
e b × e b × e b arctan( b ) × cos((n + 1)π) × arctan
aπ anπ 1 a

b
CasperYC.weebly.com/STEP Page 1 of 16 Updated: February 4, 2020

i i

i i
 i i

“87-S1-sols” — 2020/2/4 — 21:50 — page 2 — #2

i i

87-S1-sols Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16

So, depending on from the last part if n is even or odd, it is now the opposite here, which means
that this next turning point will have the opposite sign from the previous one. It follows that
   a 
aπ 1
−e fb nπ + arctan
b b

Since this is true for n and n + 1 then all the turning points must be part of a geometric progression
aπ
with common ratio: −e b as required.

(iii) See below.

40

20

-20

-40
-2 0 2 4

CasperYC.weebly.com/STEP Page 2 of 16 Updated: February 4, 2020

i i

i i
 i i

“87-S1-sols” — 2020/2/4 — 21:50 — page 3 — #3

i i

87-S1-sols Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16

2 (i) In the right angle triangle as shown in the diagram.

r
r

Pi/n R

π


We get that r = R sin n
.

(ii) Let Y = “area of the n sided-polygon” − “area of the inner (radius = R − r) circle” − “area of
the intersection of the outer circles and the n-sided polygon”.
√ √
Looking at the same triangle as before, it has opposite side = R2 − r2 and so has area 12 r R2 − r2 .
There are 2n of√ these triangles making up the polygon, so we end up with the area of the n-sided
polygon as nr R2 − r2 .
The area of the inner circle is just π(R − r)2 .
Finally, the intersection each outer  with the n-sided
 circle polygon is a sector with angle π − 2π/n,
2 π−2π/n 2 1 1


which therefore has area πr 2π
= πr 2 − n . As there are n circles, the total area is
nπr2 21 − n1 .

Hence
√
 
2 2 1 1
Y = nr R2 − r2 − π(R − r) − nπr −
2 n

(iii) Similarly, X = “area of the outer (radius = R + r) circle” − “area of the n-sided polygon” − “area
of the small circles that is outside the n-sided polygon”. We find
√
 
2 2 1 1
X = π(R + r) − nr R2 − r2 − nπr +
2 n

Obviously Z = nπr2 .

(iv) Finally
X + Y = π(R + r)2 − π(R − r)2 − nπr2 = 4πRr − nπr2 .
So
X +Y 4πRr − nπr2 4R
= 2
= −1
Z nπr nr
As r = R sin πn , for large n,
πR 4R 4R 4
r≈ =⇒ → =
n nr πR π
So
X +Y 4
→ − 1 =⇒ n → ∞.
Z π

CasperYC.weebly.com/STEP Page 3 of 16 Updated: February 4, 2020

i i

i i
 i i

“87-S1-sols” — 2020/2/4 — 21:50 — page 4 — #4

i i

87-S1-sols Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16

3 (i) Let
dy
y dv x dx −y
y = xv =⇒ v = =⇒ = 2
x dx x
Therefore
dv 1 + x2 v 2 2 dy x(1 + y 2 ) y dy 1
x3 + x2 v = =⇒ x = =⇒ =
dx (1 + x2 )v dx y(1 + x2 ) 1 + y 2 dx x(1 + x2 )

Integrate to get
y 1 x
Z Z
dy = − dx
1 + y2 x 1 + x2
1 1
ln (1 + y 2 ) = ln |x| − ln (1 + x2 ) + constant
2 2
constant = ln(2) (x = 1, v = 1 =⇒ y = 1)
2
 
2 4x
ln (1 + y ) = ln
1 + x2
4x2
1 + x2 v 2 =
1 + x2
3x2 − 1
x2 v 2 =
1 + x2
3 1
v2 = 2
− 2
1+x x (1 + x2 )

(ii) Hence as x → ∞, v → 0.

4
S = log2 e − log4 e + log16 e . . .
log2 e log2 e log2 e
= − + ...
1  2 4 
1 1 1 1 1
= log2 e 1 − + − + − ...
2 4 8 16 32
Let
1 1 1 4
P =1+ + ... = 1 =
4 16 1− 4
3
and
1 1 1 − 12 2
N =− − − ... = 1 = −
2 8 32 1− 4 3
Hence  
4 2 2 2 1
S = (P + N )(log2 e) = − log2 e = log2 e =⇒ S = = √
3 3 3 3 ln 2 ln 2 2

CasperYC.weebly.com/STEP Page 4 of 16 Updated: February 4, 2020

i i

i i
 i i

“87-S1-sols” — 2020/2/4 — 21:50 — page 5 — #5

i i

87-S1-sols Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16

5 (i) Let x = α cos2 θ + β sin2 θ, so we have

dx = 2(β − α) cos θ sin θ

Also we note now

(x − α) = (β − α) sin2 θ and (β − x) = (β − α) cos2 θ

The integral becomes

π π
β
1 2(β − α) sin θ cos θ
Z Z Z
2 2
ψ= p dx = p dθ = 2 dθ = π
α (x − α)(β − x) 0 (β − α)2 sin2 θ cos2 θ 0

(ii) If α < β, the integral becomes −π.

(iii) Let
1 1
x= =⇒ dx = − 2 dy
y y
We then have
β
1
Z
ω= p dx
α x (x − α)(β − x)
1
1
Z
α
= q dy
1
β y ( y1 − α)(β − y1 )
1
1
Z
α
= p dy
1
β
(1 − αy)(βy − 1)
1
1
Z
α
= q dy
1 1 1
β αβ( α − y)(y − β )
1
=√ ψ
αβ
π
=√
αβ

CasperYC.weebly.com/STEP Page 5 of 16 Updated: February 4, 2020

i i

i i
 i i

“87-S1-sols” — 2020/2/4 — 21:50 — page 6 — #6

i i

87-S1-sols Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16

6 (i) Let Z s Z t
F(s) = f(x) dx and G(t) = g(y) dy
0 0

where g(y) = x and y = f(x).

If A is the point (s, 0) and C is (0, t), then F(s) = area OAP and G(t) = area OCQ. Clearly the
sum of these areas is equal to or greater than OADC = st. Hence, F(s) + G(t) ≥ st. Equality
will occur if t = f(s).

(ii) Now take

f(x) = sin(x), g(y) = sin−1 (y),
then
G(t) = st − F(s) and s = sin−1 (t)
Z t Z s
−1 −1
sin (y) dy = t sin (t) − sin(x) dx = t sin−1 (t) − [− cos(x)]s0 = t sin−1 (t) − (− cos(s) + 1)
0 0
and Z t
sin−1 (y) dy = t sin−1 (t) − (1 − cos(sin−1 (t)))
0

(iii) Differentiating both sides to get

t 1
sin−1 (t) = √ + sin−1 (t) + sin(sin−1 (t)) × √ = sin−1 (t)
1 − t2 1 − t2

CasperYC.weebly.com/STEP Page 6 of 16 Updated: February 4, 2020

i i

i i
 i i

“87-S1-sols” — 2020/2/4 — 21:50 — page 7 — #7

i i

87-S1-sols Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16

7 (i) We rewrite the sum as following

10   10
X (2 + 4k)π X (2+4k)π
S= sin = ei 23

k=0
23 k=0
10
2π 4kπ
X
= ei 23 ei 23

k=0
10
X
k 2π
= eiu e2iu u=
k=0
23
1 − (e2iu )11
 
iu
=e
1 − e2iu
e11iu (e−11iu − e11iu )
= eiu
eiu (e−iu − eiu )
e11iu (e−11iu − e11iu )
=
e−iu − eiu

cos(x) + i sin(x) = eix

−ix
=⇒ e−ix − eix = −2i sin(x)
cos(x) − i sin(x) = e
Hence
−2i sin(11u) sin(11u)
S = e11iu = (cos(11u) + i sin(11u))
−2i sin(u) sin(u)
As we are only interested in the complex part of the term, we have

sin2 22π 2 π π




sin2 (11u) 23

sin sin 1 π
S= = 2π
= π

23 π
= 23

π
= tan
sin(u) sin 23 2 sin 23 cos 23 2 cos 23
2 23

(ii) Similarly, rewrite the sum as

10   X 10 10  10
(2 + 4k)π (2+4k)π
k
(−1)k ei( 23 ) = e(i 23 )
X 2π X 4π
X
k
k
S= (−1) sin = −ei 23 = e(iu) −ei2u
k=0
23 k=0 k=0 k=0

and
1 − (−e2iu )11 1 + e22iu e11iu (e−11iu + e11iu ) (e−11iu + e11iu )
     
iu iu iu
S=e =e =e = e11iu
1 − (−e2iu ) 1 + e2iu eiu (e−iu + eiu ) (e−iu + eiu )

As before ,we can show that

e−ix + eix = 2 cos(x)
It follows that
1
2 cos(11u) sin(11u) cos(11u) 2
sin(22u)
S = (cos(11u) + i sin(11u)) = =
2 cos(u) cos(u) cos(u)
2π
Putting u = 23
,
sin 44π 2π



− sin
 
23
23
1 2π
S= = = − tan
2 cos 2π
23
2 cos 2π
23
2 23

CasperYC.weebly.com/STEP Page 7 of 16 Updated: February 4, 2020

i i

i i
 i i

“87-S1-sols” — 2020/2/4 — 21:50 — page 8 — #8

i i

87-S1-sols Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16

1
8 x= t
will not work when x = 0, which is sure to happen since the integration is between [−1, 1].

(i) Let
dx
x = tan θ =⇒ = sec2 θ
dθ
1 π/4 π/4 π/4
1 sec2 θ 1 1 1 π
Z Z Z Z
2
dx = dθ = cos θ dθ = cos(2θ) + dθ = +
−1 (1 + x2 )2 −π/4 (1 + tan2 θ)2 −π/4 −π/4 2 2 2 4

(ii) Let
dt
t = tan θ =⇒ = sec2 θ
dθ
1 π/4 π/4 π/4
−t2 − tan2 θ
 
1 1 1 π
Z Z Z Z
dt = ·sec2 θ dθ = 2
− sin θ dθ = cos(2θ) − dθ = −
−1 (1 + t2 )2 −π/4 sec4 θ −π/4 −π/4 2 2 2 4

CasperYC.weebly.com/STEP Page 8 of 16 Updated: February 4, 2020

i i

i i
 i i

“87-S1-sols” — 2020/2/4 — 21:50 — page 9 — #9

i i

87-S1-sols Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16

9 (i) If P is an arbitrary point on the line segment AB with position vector p then p = a + k(b − a)
for some real number k, i.e. p = (1 − k)a + kb. Putting ρ(1 − k) and σ = k gives the required
result.

(ii) Similarly, an arbitrary point on the line segment P C will have position vector αp + βc with
α + β = 1. Now consider any point R inside the triangle ABC and let the line CR meet AB at P ,
then r = αp + βc = α(ρa + σb) + βc = λa + µb + νc with λ + µ + ν = α(ρ + σ) + β = α + β = 1.

(iii) See diagram below.

CasperYC.weebly.com/STEP Page 9 of 16 Updated: February 4, 2020

i i

i i
 i i

“87-S1-sols” — 2020/2/4 — 21:50 — page 10 — #10

i i

87-S1-sols Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16

B Mechanics

√
x2 −1
10 (i) Resolving veritcally, 2T cos θ = mg. We have sin θ = x1 , so cos θ = x
.

√
Length of streched string = π + 2 arcsin(1/x) + 2 x2 − 1. Using Hooke’s law,
√
λ(2θ + 2 x2 − 1 − π)
T = .
2π
Therefore
√
   
mg λ 1 2
= arcsin +2 x −1−π
2 cos θ π x

(ii) Solve to get √

3 π
x = 2 =⇒ cos θ = =⇒ θ =
2 6
mg λ π
 √  6πmg
√ = + 2 3 − π =⇒ λ =
3 π 6 36 − 5π

CasperYC.weebly.com/STEP Page 10 of 16 Updated: February 4, 2020

i i

i i
 i i

“87-S1-sols” — 2020/2/4 — 21:50 — page 11 — #11

i i

87-S1-sols Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16

11 (i) Using the diagram below.

Resolving radially for motion in circle

mv 2
mg cos θ − R =
r
By conservation of energy
1 2
mv = mgr(1 − cos θ) =⇒ R = mg(3 cos θ − 2)
2
Particle loses contact when R = 0, i.e. θ = arccos 23 , which is the angle of the velocity to the

horizontal.

(ii) From start particle now falls a distance 2r so loss of potential energy is 2mgr. Horizontal com-
2v
ponent  is v cos θ = 3 , so if vertical component is V , then gain in kinetic energy is
 of velocity
1 4v 2
2
m 9
+ V 2 and
2
v 2 = 2gr(1 − cos θ) = gr
3
hence
4v 2 308gr
+ V 2 = 4gr =⇒ V 2 =
9 81
308mgr
so vertical component of momentum at floor is 81 .

CasperYC.weebly.com/STEP Page 11 of 16 Updated: February 4, 2020

i i

i i
 i i

“87-S1-sols” — 2020/2/4 — 21:50 — page 12 — #12

i i

87-S1-sols Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16

12 As we can see from the sketch below.

2d h

2d h

(i) Let time to first bounce be T , then

s
1 2h
h = gT 2 =⇒ T =
2 g

Verical velocity just before first bounce is

p
gT = 2gh

so immediately after bounce it is p

egT = e 2gh
The horizontal velocity is constant, so since horizontal distance between first and second bounce is
twice the distance from start to first bounce, then time between first and second bounce is 2T . So
we will now have
1 2gT 2 − h 4h − h 3
−h = (egT )2T − g4T 2 =⇒ e = 2
= =
2 2gT 4h 4

(ii) In order for the particle to continue to land in the middle of every step, each bounce must be
identical t the second one, so time between bounce must be constant. Let veritcal velocity after
second bounce be v2 , then just before it must have been e−1 v2 . where e is the coefficient of
restitution for the subsequent bounces.
Using v 2 = u2 + 2as, we thus have from first to second bounce
√
−1 2 9 25 −1 5 2gh
(e v2 ) = v12 + 2gh = gh + 2gh = gh =⇒ e v2 =
8 8 4
and since we want
3p
v2 = v1 = 2gh
4
3
it follows that the partical can only continue hitting the middle of each step if e = 5
for the second
and subsequent steps.

CasperYC.weebly.com/STEP Page 12 of 16 Updated: February 4, 2020

i i

i i
 i i

“87-S1-sols” — 2020/2/4 — 21:50 — page 13 — #13

i i

87-S1-sols Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16

13 For x ≥ 0,
d2 x
m = −mµ2 =⇒ x = A cos(µt) + B sin(µt)
dt2
dx
At t = 0, x = 0 and dt
= v0 . So
v0
x= sin(µt)
µ
For x < 0,
d2 x dx
2
= −κ =⇒ x = C + De−κt
dt dt
dx
At t = 0, x = 0 and dt
= v0 . So
v0
x= (1 − e−κt )
κ
Sketch is given below.

CasperYC.weebly.com/STEP Page 13 of 16 Updated: February 4, 2020

i i

i i
 i i

“87-S1-sols” — 2020/2/4 — 21:50 — page 14 — #14

i i

87-S1-sols Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16

C Probability and Statistics

14 (i) (b) and (c) could be records of a complete tournament since each case B has won two games in
succession. (a) cannot be since here A, B and C have each won one game.

(ii) Sequences in which tournament is still undecided after 5 games are as follows: ACBAC, BCABC,
1
so since each sequence has a probability of (1/2)5 = 32 , probability that tournament is undecided
1
after 5 games is 16 .
1
A wins in fewer than 5 games with the sequences AA and BCAA, so probability = 14 + 16 = 165
.
5
B wins in fewer than 5 games with the sequences ACBB and BB, so probability = 16 .
C wins in fewer than 5 games with the sequences ACC and BCC, so probability = 18 + 18 = 14 .

(iii) A wins evententually with the sequence AA, ACBAA, ACBACBAA etc or BCAA, BCABCAA
etc. So the required probability is
"  # "  #
2  5  8 4  7
1 1 1 1 1 5
= + + + ... + + + ... =
2 2 2 2 2 14

B wins eventually with sequences BB, BCABB, etc or ACBB, ACBACBB etc. So the required
5
probability is 14 . C wins eventually with the sequences ACC, ACBACC etc or BCC, BCABCC
etc. So the required probability is
1
8 4 2
=2× 1 = =
1− 8
14 7

CasperYC.weebly.com/STEP Page 14 of 16 Updated: February 4, 2020

i i

i i
 i i

“87-S1-sols” — 2020/2/4 — 21:50 — page 15 — #15

i i

87-S1-sols Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16

15 As shown in the diagram below.

(i) Distance of P from (1, 0) is X = 2 sin θ where θ is uniform in [0, π2 ].

p.d.f. of θ is thus π2 for 0 < θ ≤ π2 . So

1 π 4
Z
E(X) = 2 sin θ dθ =
π 0 π
Z π
1 2 π
Z
2 2
E(X ) = 4 sin θ dθ = 1 − cos(2θ) dθ = 2
π 0 π 0
 2
4
Var(X) = 2 −
π

(ii)         
4 2 −1 2 2 −1 2
P X> = P sin θ > = 1 − P θ < sin = sin
π π π π π

CasperYC.weebly.com/STEP Page 15 of 16 Updated: February 4, 2020

i i

i i
 i i

“87-S1-sols” — 2020/2/4 — 21:50 — page 16 — #16

i i

87-S1-sols Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16

16 (i)

P(same way twice and Preservative)

P(a member voting same way twice is a preservative) =
P(same way twice)
3
5
= 3 2 1
5
+ ×
5 2
3
=
4
P(member votes same way a third time given voted same way twice)
1
=P(member is a Preservative) + P(Member is a Progessive) ×
2
3 1 1
= + ×
4 4 2
7
=
8

(ii) Number of votes in favour of Preservative policy = 60(1 − α) + β where β is the number of
Progressives voting for the policy, i.e. β = B(40, 0.5) ≈ N (20, 10). So number of votes is

X = 60(1 − α) + β ≈ N (60(1 − α) + 20, 10)

P(getting a majority) = P(X > 50)

√
= P(80 − 60α + z 10 > 50)
 
60α − 30
=P z> √ < 0.8
10
60α − 30
√ > 0.8416
10
α > 0.544

CasperYC.weebly.com/STEP Page 16 of 16 Updated: February 4, 2020

i i

i i