Year: 2012
Paper: 2
Question Number: 7
Course: UFM Pure
Section: Vectors
There were just over 1000 entries for paper II this year, almost exactly the same number as last year. Overall, the paper was found marginally easier than its predecessor, which means that it was pitched at exactly the level intended and produced the hoped-for outcomes. Almost 50 candidates scored 100 marks or more, with more than 400 gaining at least half marks on the paper. At the lower end of the scale, around a quarter of the entry failed to score more than 40 marks. It was pleasing to note that the advice of recent years, encouraging students not to make attempts at lots of early parts to questions but rather to spend their time getting to grips with the six that can count towards their paper total, was more obviously being heeded in 2012 than I can recall being the case previously. As in previous years, the pure maths questions provided the bulk of candidates' work, with relatively few efforts to be found at the applied ones. Questions 1 and 2 were the most popular questions, although each drew only around 800 "hits" – fewer than usual. Questions 3 – 5 & 8 were almost as popular (around 700), with Q6 attracting the interest of under 450 candidates and Q7 under 200. Q9 was the most popular applied question – and, as it turned out, the most successfully attempted question on the paper – with very little interest shown in the rest of Sections B or C.
Difficulty Rating: 1600.0
Difficulty Comparisons: 0
Banger Rating: 1516.0
Banger Comparisons: 1
Three distinct points, $X_1$, $X_2$ and $X_3$, with position vectors ${\bf x}_1$, ${\bf x}_2$ and ${\bf x}_3$ respectively, lie on a circle of radius 1 with its centre at the origin $O$. The point $G$ has position vector $\frac13({\bf x}_1+{\bf x}_2+{\bf x}_3)$. The line through $X_1$ and $G$ meets the circle again at the point $Y_1$ and the points $Y_2$ and $Y_3$ are defined correspondingly.
Given that $\overrightarrow{GY_1} =-\lambda_1\overrightarrow{GX_1}$, where $\lambda_1$ is a positive scalar, show that
\[
\overrightarrow{OY_1}= \tfrac13 \big(
(1-2\lambda_1){\bf x}_1 +(1+\lambda_1)({\bf x}_2+{\bf x}_3)\big)
\]
and hence that
\[
\lambda_1 = \frac
{3-\alpha-\beta-\gamma} {3+\alpha -2\beta-2\gamma}
\,,\]
where $\alpha = {\bf x}_2 \,.\, {\bf x}_3$, $\beta = {\bf x}_3\,.\, {\bf x}_1$ and $\gamma = {\bf x}_1\,.\, {\bf x}_2$.
Deduce that $\dfrac {GX_1}{GY_1} + \dfrac {GX_2}{GY_2} +
\dfrac {GX_3}{GY_3} =3
\,$.
\begin{center}
\begin{tikzpicture}
\def\a{2};
\coordinate (O) at (0,0);
\coordinate (A) at ({\a*cos(20)},{\a*sin(20)});
\coordinate (B) at ({\a*cos(150)},{\a*sin(150)});
\coordinate (C) at ({\a*cos(250)},{\a*sin(250)});
% Calculate Centroid G
\coordinate (G) at ($1/3*(A)+1/3*(B)+1/3*(C)$);
% 1. Draw the circle and name it
\draw[dashed, name path=circumcircle] (O) circle (\a);
% 2. Create an invisible path that extends through A and G in BOTH directions.
% Using -2 and 4 ensures it crosses the circle boundaries clearly.
\path[name path=median_a] ($(A)!-1!(G)$) -- ($(A)!3!(G)$);
\path[name path=median_b] ($(B)!-1!(G)$) -- ($(B)!3!(G)$);
\path[name path=median_c] ($(C)!-1!(G)$) -- ($(C)!3!(G)$);
% 3. Find intersections.
% We use 'sort by=median' to ensure point 1 is at one end and point 2 at the other.
\fill [name intersections={of=circumcircle and median_a, sort by=median, by={Y1_other, Y1}}];
\fill [name intersections={of=circumcircle and median_b, sort by=median, by={Y2_other, Y2}}];
\fill [name intersections={of=circumcircle and median_c, sort by=median, by={Y3, Y3_other}}];
% Draw the line through the triangle (A to Y1)
\draw[blue] (A) -- (Y1);
\draw[blue] (B) -- (Y2);
\draw[blue] (C) -- (Y3);
% Draw points and labels
\filldraw (A) circle (1.5pt) node[right] {$X_1$};
\filldraw (B) circle (1.5pt) node[left] {$X_2$};
\filldraw (C) circle (1.5pt) node[below left] {$X_3$};
\filldraw (G) circle (1.5pt) node[below] {$G$};
\filldraw (O) circle (1.5pt) node[right] {$O$};
\filldraw (Y1) circle (1.5pt) node[below left] {$Y_1$};
\filldraw (Y2) circle (1.5pt) node[below right] {$Y_2$};
\filldraw (Y3) circle (1.5pt) node[above] {$Y_3$};
\end{tikzpicture}
\end{center}
\begin{align*}
&& \mathbf{y}_1 &= \overrightarrow{OG}+\overrightarrow{GY_1} \\
&&&= \frac13(\mathbf{x}_1+\mathbf{x}_2+\mathbf{x}_3) -\lambda_1 \left (\mathbf{x}_1 - \frac13(\mathbf{x}_1+\mathbf{x}_2+\mathbf{x}_3)\right) \\
&&&= \frac13 \left ( (1-2\lambda_1)\mathbf{x}_1+(1+\lambda_1)(\mathbf{x}_2+\mathbf{x}_3)\right) \\
&& 1 &= \mathbf{y}_1 \cdot \mathbf{y}_1 \\
&&&= \frac13 \left ( (1-2\lambda_1)\mathbf{x}_1+(1+\lambda_1)(\mathbf{x}_2+\mathbf{x}_3)\right) \cdot \frac13 \left ( (1-2\lambda_1)\mathbf{x}_1+(1+\lambda_1)(\mathbf{x}_2+\mathbf{x}_3)\right) \\
&&&= \frac19\left ( (1-2\lambda_1)^2+2(1+\lambda_1)^2 + 2(1-2\lambda_1)(1+\lambda_1)(\mathbf{x}_1 \cdot \mathbf{x}_2+\mathbf{x}_1 \cdot \mathbf{x}_3) + 2(1+\lambda_1)^2 \mathbf{x}_2 \cdot \mathbf{x}_3 \right) \\
\Rightarrow && 9 &= (1-2\lambda_1)^2+2(1+\lambda_1)^2 + 2(1-2\lambda_1)(1+\lambda_1)(\beta+\gamma) + 2(1+\lambda_1)^2 \alpha \\
&&&= 3+6\lambda_1^2+2(\beta+\gamma)-2(\beta+\gamma)\lambda_1 - 4\lambda_1^2(\beta+\gamma) + 2\alpha+4\lambda_1\alpha + 2\lambda_1^2 \alpha \\
&& 0 &= (-6+2(\alpha+\beta+\gamma))+2(2\alpha-(\beta+\gamma))\lambda_1 + (6+2(\alpha-2(\beta+\gamma)))\lambda_1^2 \\
\Rightarrow && 0 &= ((\alpha+\beta+\gamma)-3)+(2\alpha-(\beta+\gamma))\lambda_1 + (3+\alpha-2(\beta+\gamma))\lambda_1^2 \\
&&&= (\lambda_1+1)((3+\alpha-2(\beta+\gamma))\lambda_1+ ((\alpha+\beta+\gamma)-3)) \\
\Rightarrow && \lambda_1 &= \frac{3-(\alpha+\beta+\gamma)}{3+\alpha-2(\beta+\gamma)}
\end{align*}
as required.
Since $\dfrac {GX_1}{GY_1} = \frac1{\lambda_1}$ we must have,
\begin{align*}
&& \frac {GX_1}{GY_1} + \frac {GX_2}{GY_2} +
\frac {GX_3}{GY_3} &= \frac1{\lambda_1}+\frac1{\lambda_2}+\frac1{\lambda_3} \\
&&&= \frac{(3+\alpha-2\beta-2\gamma)+(3+\beta-2\gamma-2\alpha)+3+\gamma-2\alpha-2\beta)}{3-\alpha-\beta-\gamma} \\
&&&= \frac{9-3(\alpha+\beta+\gamma)}{3-(\alpha+\beta+\gamma)} \\
&&&= 3
\end{align*}
This question was the least popular of the pure maths questions by a considerable margin, and attempts at it were usually fairly poor. In fact, very few candidates got beyond the opening (given) result. The barrier to further progress was almost invariably the failure to realise that all points on a circle centre O and radius 1 have position vectors that satisfy x·x = 1.