A Special Locus

Geometry Level 5

The figure shows a rectangle of side lengths A B = a {AB=a} and B C = b {BC=b} . A variable point P P outside the rectangle is made such that A X AX , X Y XY and Y B YB form a geometric progression. If θ \theta is the minimum measure of C P D \angle{CPD} satisfying the conditions, find the expression for tan θ \tan \theta in terms of a a and b b . Submit your answer with a = 5 a=5 and b = 3 b=3 .


The answer is 1.5.

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

4 solutions

Let E E be the foot of the perpendicular to A B AB from P P . Denote A X AX , X E XE , E Y EY , Y B YB and P E PE by k k , m m , n n , l l and h h respectively, as seem on the figure.
Let X P E = θ 1 \angle XPE={{\theta }_{1}} and Y P E = θ 2 \angle YPE={{\theta }_{2}} .

P E X \triangle PEX , D A X \triangle DAX and P E Y \triangle PEY , C B Y \triangle CBY are pairs of similar triangles, hence h b = m k = n l ( 1 ) \frac{h}{b}=\frac{m}{k}=\frac{n}{l} \ \ \ \ \ (1)

( 1 ) k = b m h and l = b n h ( 2 ) \left( 1 \right)\Rightarrow k=b\frac{m}{h} \ \ \ \text{and} \ \ \ l=b\frac{n}{h} \ \ \ \ \ (2) Since m k = m l = m + n k + l \dfrac{m}{k}=\dfrac{m}{l}=\dfrac{m+n}{k+l} , ( 1 ) h b = m + n k + l h b = m + n a ( m + n ) m h + n h = a ( m + n ) b ( 3 ) \left( 1 \right)\Rightarrow \frac{h}{b}=\frac{m+n}{k+l}\Rightarrow \frac{h}{b}=\frac{m+n}{a-\left( m+n \right)}\Rightarrow \frac{m}{h}+\frac{n}{h}=\frac{a-\left( m+n \right)}{b} \ \ \ \ \ (3) From the geometric progression,

B Y Y X = Y X X A Y X 2 = X A B Y ( m + n ) 2 = k l ( 2 ) ( m + n ) 2 = b 2 m h n h m h n h = ( m + n ) 2 b 2 ( 4 ) \begin{aligned} \frac{BY}{YX}=\frac{YX}{XA}\Rightarrow Y{{X}^{2}}=XA\cdot BY & \Rightarrow {{\left( m+n \right)}^{2}}=kl \\ & \overset{\left( 2 \right)}{\mathop{\Rightarrow }}\,{{\left( m+n \right)}^{2}}={{b}^{2}}\cdot \frac{m}{h}\cdot \frac{n}{h} \\ & \Rightarrow \frac{m}{h}\cdot \frac{n}{h}=\frac{{{\left( m+n \right)}^{2}}}{{{b}^{2}}} \ \ \ \ \ (4) \\ \end{aligned} Now, we have

tan θ = tan θ 1 + tan θ 2 1 tan θ 1 tan θ 2 = m h + n h 1 m h m h = ( 3 ) , ( 4 ) a ( m + n ) b 1 ( m + n ) 2 b 2 = b a ( m + n ) b 2 ( m + n ) 2 \tan \theta =\dfrac{\tan {{\theta }_{1}}+\tan {{\theta }_{2}}}{1-\tan {{\theta }_{1}}\cdot \tan {{\theta }_{2}}}=\dfrac{\dfrac{m}{h}+\dfrac{n}{h}}{1-\frac{m}{h}\cdot \dfrac{m}{h}}\overset{\left( 3 \right),\left( 4 \right)}{\mathop{=}}\,\dfrac{\dfrac{a-\left( m+n \right)}{b}}{1-\dfrac{{{\left( m+n \right)}^{2}}}{{{b}^{2}}}}=b\dfrac{a-\left( m+n \right)}{{{b}^{2}}-{{\left( m+n \right)}^{2}}} Setting x = m + n x=m+n , we have a function for tan θ \tan \theta in terms of x x :

f ( x ) = b a x b 2 x 2 f\left( x \right)=b\frac{a-x}{{{b}^{2}}-{{x}^{2}}} For the domain of f f we claim that dom ( f ) = ( 0 , a 3 ] \text{dom}\left( f \right)=\left( 0,\dfrac{a}{3} \right] .
Proof: Let r r be the common ratio of the geometric progression, i.e. l x = x k = r > 0 \dfrac{l}{x}=\dfrac{x}{k}=r>0 . Then, l = r x l=rx , k = x r k=\dfrac{x}{r} , thus we have l + x + k = a r x + x + x r = a x = r r 2 + r + 1 a l+x+k=a\Rightarrow rx+x+\frac{x}{r}=a\Rightarrow x=\frac{r}{{{r}^{2}}+r+1}a An easy study of the function g ( r ) = r r 2 + r + 1 g\left( r \right)=\dfrac{r}{{{r}^{2}}+r+1} shows that its range is ran ( g ) = ( 0 , 1 3 ] \text{ran}\left( g \right)=\left( 0,\dfrac{1}{3} \right] , hence the domain of f f is ( 0 , a 3 ] \left( 0,\dfrac{a}{3} \right] .

The derivative of f f is
f ( x ) = b ( b 2 x 2 ) 2 ( x 2 + 2 a x b 2 ) {f}'\left( x \right)=\dfrac{b}{{{\left( {{b}^{2}}-{{x}^{2}} \right)}^{2}}}\left( -{{x}^{2}}+2ax-{{b}^{2}} \right)

The discriminant of the quadratic is Δ = 4 ( a 2 b 2 ) \Delta =4\left( {{a}^{2}}-{{b}^{2}} \right) .

Case 1 : a b a\le b
Then Δ 0 \Delta \le 0 and f ( x ) < 0 {f}'\left( x \right) <0 for all x ( 0 , a 3 ) x\in \left( 0,\dfrac{a}{3} \right) with an exception of a single root if Δ = 0 \Delta = 0 . Hence f f is a decreasing function and, since f f is continuous, its range is [ f ( a 3 ) , lim x 0 + f ( x ) ) \left[ f\left( \dfrac{a}{3} \right),\underset{x\to {{0}^{+}}}{\mathop{\lim }}\,f\left( x \right) \right) Hence,
f min = f ( a 3 ) = b a a 3 b 2 ( a 3 ) 2 = 6 a b 9 b 2 a 2 {{f}_{\min }}=f\left( \dfrac{a}{3} \right)=b\dfrac{a-\dfrac{a}{3}}{{{b}^{2}}-{{\left( \dfrac{a}{3} \right)}^{2}}}=\dfrac{6ab}{9{{b}^{2}}-{{a}^{2}}}

Case 2 : a > b a>b
Then,

x 2 + 2 a x b 2 = 0 x = 2 a ± 2 a 2 b 2 2 = a a 2 b 2 -{{x}^{2}}+2ax-{{b}^{2}}=0\Leftrightarrow x=\frac{-2a\pm 2\sqrt{{{a}^{2}}-{{b}^{2}}}}{-2}=a\mp \sqrt{{{a}^{2}}-{{b}^{2}}} The value x = a + a 2 b 2 x=a+\sqrt{{{a}^{2}}-{{b}^{2}}} is greater than a a , hence it is rejected, therefore the only value we have to deal with is x = a + a 2 b 2 x=a+\sqrt{{{a}^{2}}-{{b}^{2}}} .

Case 2i : a a 2 b 2 a 3 a-\sqrt{{{a}^{2}}-{{b}^{2}}}\ge \frac{a}{3} Then, 2 a 3 a 2 b 2 4 a 2 9 a 2 b 2 5 a 2 9 b 2 \dfrac{2a}{3}\ge \sqrt{{{a}^{2}}-{{b}^{2}}}\Rightarrow 4{{a}^{2}}\ge 9{{a}^{2}}-{{b}^{2}}\Rightarrow 5{{a}^{2}}\le 9{{b}^{2}} and in this case x 2 + 2 a x b 2 < 0 -{{x}^{2}}+2ax-{{b}^{2}}<0 for all x ( 0 , a 3 ) x\in \left( 0,\dfrac{a}{3} \right) , thus, like in case 1, f f is a decreasing function and f min = 6 a b 9 b 2 a 2 {{f}_{\min }}=\dfrac{6ab}{9{{b}^{2}}-{{a}^{2}}} .

Case 2ii : a a 2 b 2 < a 3 5 a 2 > 9 b 2 a-\sqrt{{{a}^{2}}-{{b}^{2}}}<\dfrac{a}{3}\Leftrightarrow 5{{a}^{2}}>9{{b}^{2}}
Then f f has a minimum at x 0 = a a 2 b 2 {{x}_{0}}=a-\sqrt{{{a}^{2}}-{{b}^{2}}} , because f ( x ) < 0 {f}'\left( x \right)<0 for x ( 0 , x 0 ) x\in \left(0, {{x}_{0}} \right) and f ( x ) > 0 {f}'\left( x \right)>0 for x ( x 0 , a 3 ] x\in \left( {{x}_{0}},\dfrac{a}{3} \right] .

The minimum value is

f min = f ( a a 2 b 2 ) = b a 2 b 2 b 2 ( a a 2 b 2 ) 2 = b a 2 b 2 b 2 a 2 + b 2 + 2 a a 2 b 2 = b a 2 b 2 b 2 a 2 + b 2 + 2 a a 2 b 2 = 1 2 b a 2 b 2 a 2 b 2 ( a a 2 b 2 ) = 1 2 b ( a a 2 b 2 ) = b ( a + a 2 b 2 ) 2 ( a 2 a 2 + b 2 ) = a + a 2 b 2 2 b \begin{aligned} {{f}_{\min }}=f\left( a-\sqrt{{{a}^{2}}-{{b}^{2}}} \right) & =\dfrac{b\sqrt{{{a}^{2}}-{{b}^{2}}}}{{{b}^{2}}-{{\left( a-\sqrt{{{a}^{2}}-{{b}^{2}}} \right)}^{2}}} \\ & =\dfrac{b\sqrt{{{a}^{2}}-{{b}^{2}}}}{{{b}^{2}}-{{a}^{2}}+{{b}^{2}}+2a\sqrt{{{a}^{2}}-{{b}^{2}}}} \\ & =\dfrac{b\sqrt{{{a}^{2}}-{{b}^{2}}}}{{{b}^{2}}-{{a}^{2}}+{{b}^{2}}+2a\sqrt{{{a}^{2}}-{{b}^{2}}}} \\ & =\dfrac{1}{2}\cdot \dfrac{b\sqrt{{{a}^{2}}-{{b}^{2}}}}{\sqrt{{{a}^{2}}-{{b}^{2}}}\left( a-\sqrt{{{a}^{2}}-{{b}^{2}}} \right)} \\ & =\dfrac{1}{2}\cdot \dfrac{b}{\left( a-\sqrt{{{a}^{2}}-{{b}^{2}}} \right)} \\ & =\dfrac{b\left( a+\sqrt{{{a}^{2}}-{{b}^{2}}} \right)}{2\left( {{a}^{2}}-{{a}^{2}}+{{b}^{2}} \right)} \\ & =\dfrac{a+\sqrt{{{a}^{2}}-{{b}^{2}}}}{2b} \\ \end{aligned}

With a = 5 a=5 and b = 3 b=3 we are in case 2ii, hence our answer is: f min = 5 + 5 2 3 2 2 × 3 = 3 2 = 1.5 {{f}_{\min }}=\dfrac{5+\sqrt{{{5}^{2}}-{{3}^{2}}}}{2\times 3}=\dfrac{3}{2}=\boxed{1.5}

The minimum value of θ \theta is different for the cases where 5 a 2 < 9 b 2 5a^2<9b^2 .

The expression that you have calculated works only for the cases where 5 a 2 > 9 b 2 5a^2>9b^2 .

And, θ \theta can also attain a minimum value for cases where a < b a<b .

Digvijay Singh - 5 months, 1 week ago

Log in to reply

Τhanks for the notes, you are right. Ι have corrected my solution.

Thanos Petropoulos - 5 months, 1 week ago

Let θ min \theta_\text{min} be the minimum value of θ \theta .

For 5 a 2 < 9 b 2 5a^2<9b^2 , tan ( θ min ) = 6 a b 9 b 2 a 2 \tan(\theta_\text{min})=\dfrac{6ab}{9b^2-a^2} .

For 5 a 2 > 9 b 2 5a^2>9b^2 , tan ( θ min ) = a + a 2 b 2 2 b \tan(\theta_\text{min})=\dfrac{a+\sqrt{a^2-b^2}}{2b} .

Digvijay Singh - 5 months, 1 week ago

Log in to reply

@Digvijay Singh - I think, in the wording of the problem, you should include the restriction that b a 2 b\ge \frac{a}{2} , otherwise θ \theta can get equal to 90 90{}^\circ and also can take values greater than 90 90{}^\circ . Since lim θ π 2 + ( tan θ ) = \underset{\theta \to {{\frac{\pi }{2}}^{+}}}{\mathop{\lim }}\,\left( \tan \theta \right)=-\infty , tan θ \tan \theta does not have a minimum in this case.

Thanos Petropoulos - 5 months, 1 week ago

Log in to reply

The question does not ask you to find the minimum of tan θ \tan{\theta} .

The question asks to find the tangent of the minimum angle θ min \theta_\text{min} , which is always finite, since θ min \theta_\text{min} is always less than 9 0 90^\circ .

Digvijay Singh - 5 months, 1 week ago

@Thanos Petropoulos I think you meant P E Y C B Y \triangle PEY \approx \triangle \color{#D61F06} CBY . (third line from the top)

N. Aadhaar Murty - 5 months, 1 week ago

Log in to reply

Thanks, I edited it.

Thanos Petropoulos - 5 months, 1 week ago
David Vreken
Jan 5, 2021

Let the midpoint of A B AB be at the origin, with A ( 1 2 a , 0 ) A(\frac{1}{2}a, 0) , B ( 1 2 a , 0 ) B(-\frac{1}{2}a, 0) , C ( 1 2 a , b ) C(-\frac{1}{2}a, -b) , and D ( 1 2 a , b ) D(\frac{1}{2}a, -b) . Also draw P R C D PR \perp CD with R R on C D CD and Q Q on A B AB , and let θ 1 = C P R \theta_1 = \angle CPR and θ 2 = D P R \theta_2 = \angle DPR .

We can show that the locus of points for P P lie on the upper half of an ellipse with a major axis of a a , a minor axis of b b , and centered at the origin:

Using a parametric representation of the ellipse, P P will have coordinates P ( 1 2 a cos t , 1 2 b sin t ) P(\frac{1}{2}a \cos t, \frac{1}{2}b \sin t) .

That means tan θ 1 = C R P R = 1 2 a cos t + 1 2 a b + 1 2 b sin t = a ( 1 + cos t ) b ( 2 + sin t ) \tan \theta_1 = \cfrac{CR}{PR} = \cfrac{\frac{1}{2}a \cos t + \frac{1}{2}a}{b + \frac{1}{2}b \sin t} = \cfrac{a(1 + \cos t)}{b(2 + \sin t)} and tan θ 2 = R D P R = 1 2 a 1 2 a cos t b + 1 2 b sin t = a ( 1 cos t ) b ( 2 + sin t ) \tan \theta_2 = \cfrac{RD}{PR} = \cfrac{\frac{1}{2}a - \frac{1}{2}a \cos t}{b + \frac{1}{2}b \sin t} = \cfrac{a(1 - \cos t)}{b(2 + \sin t)} .

Since P Y X P C D \triangle PYX \sim \triangle PCD by AA similarity, Y X C D = P Q P R \cfrac{YX}{CD} = \cfrac{PQ}{PR} , or Y X a = 1 2 b sin t b + 1 2 b sin t \cfrac{YX}{a} = \cfrac{\frac{1}{2}b \sin t}{b + \frac{1}{2}b \sin t} , which solves to Y X = a sin t 2 + sin t YX = \cfrac{a \sin t}{2 + \sin t} .

As alternate interior angles of parallel lines, B C Y = Y P R = θ 1 \angle BCY = \angle YPR = \theta_1 , so that by B C Y \triangle BCY , B Y = B C tan θ 1 = b a ( 1 + cos t ) b ( 2 + sin t ) = a ( 1 + cos t ) 2 + sin t BY = BC \tan \theta_1 = b \cdot \cfrac{a(1 + \cos t)}{b(2 + \sin t)} = \cfrac{a(1 + \cos t)}{2 + \sin t} .

Similarly, X D A = Q P X = θ 2 \angle XDA = \angle QPX = \theta_2 , so that by X D A \triangle XDA , X A = A D tan θ 2 = b a ( 1 cos t ) b ( 2 + sin t ) = a ( 1 cos t ) 2 + sin t XA = AD \tan \theta_2 = b \cdot \cfrac{a(1 - \cos t)}{b(2 + \sin t)} = \cfrac{a(1 - \cos t)}{2 + \sin t} .

Since ( 1 + cos t ) ( 1 cos t ) = sin 2 t (1 + \cos t)(1 - \cos t) = \sin^2 t , B Y X A = Y X 2 BY \cdot XA = YX^2 , which means B Y BY , Y X YX , and X A XA are in a geometric progression.

Substituting tan θ 1 = a ( 1 + cos t ) b ( 2 + sin t ) \tan \theta_1 = \cfrac{a(1 + \cos t)}{b(2 + \sin t)} and tan θ 2 = a ( 1 cos t ) b ( 2 + sin t ) \tan \theta_2 = \cfrac{a(1 - \cos t)}{b(2 + \sin t)} into tan ( θ 1 + θ 2 ) = tan θ 1 + tan θ 2 1 tan θ 1 tan θ 2 \tan (\theta_1 + \theta_2) = \cfrac{\tan \theta_1 + \tan \theta_2}{1 - \tan \theta_1 \tan \theta_2} gives tan θ = 2 a b ( 2 + sin t ) b 2 ( 2 + sin t ) a 2 sin 2 t \tan \theta = \cfrac{2ab(2 + \sin t)}{b^2(2 + \sin t) - a^2\sin^2 t} .

The minimum and maximum θ \theta values will occur when d d t ( tan θ ) = a b cos t ( ( a 2 b 2 ) sin 2 t + 4 ( a 2 b 2 ) sin t 4 b 2 ) ( a sin t b ( sin t + 2 ) ) 2 ( a sin t + b ( sin t + 2 ) ) 2 = 0 \cfrac{d}{dt}(\tan \theta) = \cfrac{ab \cos t((a^2 - b^2) \sin^2 t + 4(a^2 - b^2) \sin t - 4b^2)}{(a \sin t - b(\sin t + 2))^2(a \sin t + b(\sin t + 2))^2} = 0 , which is either at cos t = 0 \cos t = 0 (which solves to sin t = 1 \sin t = 1 ), or at ( a 2 b 2 ) sin 2 t + 4 ( a 2 b 2 ) sin t 4 b 2 ) = 0 (a^2 - b^2) \sin^2 t + 4(a^2 - b^2) \sin t - 4b^2) = 0 (which solves to sin t = 2 + 2 a a 2 b 2 \sin t = -2 + \cfrac{2a}{\sqrt{a^2 - b^2}} ).

For the latter, since 1 sin t 1 -1 \leq \sin t \leq 1 , then 2 + 2 a a 2 b 2 1 -2 + \cfrac{2a}{\sqrt{a^2 - b^2}} \leq 1 which solves to 5 a 2 9 b 2 5a^2 \geq 9b^2 . Some careful analysis of the derivative function shows that the minimum occurs at sin t = 1 \sin t = 1 when 5 a 2 < 9 b 2 5a^2 < 9b^2 and at sin t = 2 + 2 a a 2 b 2 \sin t = -2 + \cfrac{2a}{\sqrt{a^2 - b^2}} when 5 a 2 9 b 2 5a^2 \geq 9b^2 .

Substituting these values of sin t \sin t back into tan θ = 2 a b ( 2 + sin t ) b 2 ( 2 + sin t ) a 2 sin 2 t \tan \theta = \cfrac{2ab(2 + \sin t)}{b^2(2 + \sin t) - a^2\sin^2 t} gives us the following:

  • when 5 a 2 < 9 b 2 5a^2 < 9b^2 , the minimum θ \theta occurs when sin t = 1 \sin t = 1 , and tan θ min = 2 a b ( 2 + 1 ) b 2 ( 2 + 1 ) a 2 1 2 = 6 a b 9 b 2 a 2 \tan \theta_{\text{min}} = \cfrac{2ab(2 + 1)}{b^2(2 + 1) - a^2 \cdot 1^2} = \cfrac{6ab}{9b^2 - a^2}

  • when 5 a 2 9 b 2 5a^2 \geq 9b^2 , the minimum θ \theta occurs when sin t = 2 + 2 a a 2 b 2 \sin t = -2 + \cfrac{2a}{\sqrt{a^2 - b^2}} , and tan θ min = 2 a b ( 2 a a 2 b 2 ) b 2 ( 2 a a 2 b 2 ) a 2 ( 2 + 2 a a 2 b 2 ) 2 = a + a 2 b 2 2 b \tan \theta_{\text{min}} = \cfrac{2ab(\frac{2a}{\sqrt{a^2 - b^2}})}{b^2(\frac{2a}{\sqrt{a^2 - b^2}}) - a^2 \cdot (-2 + \frac{2a}{\sqrt{a^2 - b^2}})^2} = \cfrac{a + \sqrt{a^2 - b^2}}{2b}

In this problem, a = 5 a = 5 and b = 3 b = 3 , so 5 a 2 9 b 2 5a^2 \geq 9b^2 , and tan θ min = 5 + 5 2 3 2 2 3 = 1.5 \tan \theta_{\text{min}} = \cfrac{5 + \sqrt{5^2 - 3^2}}{2 \cdot 3} = \boxed{1.5} .

Nice observation that we have an ellipse involved and worth to be brought up. To be more precise, it can be an ellipse with a major axis of a a and a minor axis of b b or vise versa, or it can even be a circle in case a = b a=b . It is easy to show that only the points of this conic beget the geometric progression, hence the locus is indeed the upper part of the ellipse/circle.

Thanos Petropoulos - 5 months, 1 week ago
Guilherme Niedu
Jan 2, 2021

Let Y X = c \overline{YX} = c and their common ratio be p p , so that A X = c p \overline {AX} = \frac{c}{p} and Y B = c p \overline{YB} = cp . From that it follows:

c p + c + c p = a \large \displaystyle cp + c + \frac{c}{p} = a

p + 1 + 1 p = a c \large \displaystyle p + 1 + \frac{1}{p} = \frac{a}{c}

( i ) p + 1 p = a c c \color{#20A900} (i) \ \ \boxed{ \large \displaystyle p + \frac{1}{p} = \frac{a-c}{c}}

Also, let α \alpha and β \beta be marked as follows.

Notice also that B Y C = α \angle{BYC} = \alpha and A X D = β \angle{AXD} = \beta . So:

tan ( α ) = b c p \large \displaystyle \tan(\alpha) = \frac{b}{cp}

tan ( β ) = b p c \large \displaystyle \tan(\beta) = \frac{bp}{c}

So:

tan ( α + β ) = b c p + b p c 1 b c p b p c \large \displaystyle \tan(\alpha + \beta) = \frac{ \frac{b}{cp} + \frac{bp}{c}}{1 - \frac{b}{cp}\frac{bp}{c}}

tan ( α + β ) = b ( p + 1 p ) c c 2 b 2 \large \displaystyle \tan(\alpha + \beta) = b \left( p + \frac{1}{p} \right) \frac{c}{c^2-b^2}

Plugging ( i ) (i) into this:

tan ( α + β ) = b ( a c c ) c c 2 b 2 \large \displaystyle \tan(\alpha + \beta) = b\left( \frac{a-c}{c} \right) \frac{c}{c^2-b^2}

tan ( α + β ) = b ( a c ) c 2 b 2 \large \displaystyle \tan(\alpha + \beta) = \frac{b(a-c)}{c^2-b^2}

Since θ \theta and α + β \alpha+\beta add up to 180 180 degrees, tan ( θ ) = tan ( α + β ) \tan(\theta) = -\tan(\alpha + \beta) :

( i i ) tan ( θ ) = b ( c a ) c 2 b 2 \color{#20A900} (ii) \ \ \boxed{\large \displaystyle \tan(\theta) = \frac{b(c-a)}{c^2-b^2} }

To minimize θ \theta is the same as minimize tan ( θ ) \tan(\theta) , since 0 o < θ < 9 0 o 0^o < \theta < 90^o and, in that interval, tan ( θ ) \tan(\theta) is continuous and increasing . So, differentiating ( i i ) (ii) with respect to c c , one gets:

b ( c 2 b 2 ) ( b c a b ) 2 c ( c 2 b 2 ) 2 = 0 \large \displaystyle \frac{ b(c^2-b^2) - (bc-ab)2c}{ (c^2-b^2)^2} = 0

If c b c \neq b , then we can multiply both sides by ( c 2 b 2 ) 2 (c^2-b^2)^2 . Then:

c 2 2 a c + b 2 = 0 \large \displaystyle c^2 - 2ac + b^2 = 0

[ c ( a + a 2 b 2 ) ] [ c ( a a 2 b 2 ) ] = 0 \large \displaystyle [c - (a + \sqrt{a^2-b^2})]\cdot [c - (a - \sqrt{a^2-b^2})] = 0

Since c < a c < a :

c = a a 2 b 2 \color{#20A900} \boxed{ \large \displaystyle c = a - \sqrt{a^2-b^2}} .

Which makes for the given case:

c = 1 \color{#3D99F6} \boxed{\large \displaystyle c=1}

And:

tan ( θ ) = 3 2 = 1.5 \color{#3D99F6} \boxed{\large \displaystyle \tan(\theta) = \frac{3}{2} = 1.5}

The question asked to find a general solution. Could you maybe address the general problem rather than the specific one?

Digvijay Singh - 5 months, 1 week ago

Log in to reply

Sorry when I solved it the question was different. I'll edit the answer then

Guilherme Niedu - 5 months, 1 week ago

Mr @Digvijay Singh , I've edited it to make it general.

Guilherme Niedu - 5 months, 1 week ago
Pi Han Goh
Jan 1, 2021

Disclaimer: I've removed a flawed section of my solution because I made a wrong generalization.


To avoid the abuse of notation, I will relabel the vertex A A as A A' .

Let ( A X , X Y , Y B ) = ( A R , A , A R ) (A'X,XY, YB) = \left( \frac AR, A, AR \right) , where R > 1 R > 1 and a = A B = A R + A + A R = 5 a = AB = \frac AR + A + AR = 5 .

Then, Y C D = B Y C = tan 1 ( 3 A R ) \angle YCD = \angle BYC = \tan^{-1} \left( \dfrac 3{AR} \right) .
Similarly, X D C = A X D = tan 1 ( 3 A / R ) \angle XDC = \angle A'XD = \tan^{-1} \left( \dfrac 3{A/R} \right) .

With b = 3 b = 3 . C P D = π Y C D X D C = π [ tan 1 ( 3 A R ) + tan 1 ( 3 A / R ) ] \angle CPD = \pi - \angle YCD - \angle XDC = \pi - \left [ \tan^{-1} \left( \dfrac 3{AR} \right) + \tan^{-1} \left( \dfrac 3{A/R} \right) \right ]

Apply the compound angle formula, tan ( x + y ) = tan x + tan y 1 tan x tan y , \tan(x + y) = \dfrac{ \tan x + \tan y}{1 - \tan x \cdot \tan y } , the equation above simplifies to C P D = π tan 1 ( 3 A ( R 2 + 1 ) R ( A 2 9 ) ) \angle CPD = \pi - \tan^{-1} \left( \dfrac{3A(R^2 + 1)}{R(A^2 - 9)} \right)

Substitute A = 5 R + 1 / R + 1 A = \dfrac 5{R + 1/R + 1} gives

C P D = π + tan 1 [ 15 ( R 2 + 1 ) ( R 2 + R + 1 ) ( 3 R 2 2 R + 3 ) ( 3 R 2 + 8 R + 3 ) ] \angle CPD = \pi + \tan^{-1} \left [\dfrac{15(R^2+1)(R^2 + R + 1)}{(3R^2 -2R +3)(3R^2+8R+3)} \right ]

So we want to minimize f ( R ) : = 15 ( R 2 + 1 ) ( R 2 + R + 1 ) ( 3 R 2 2 R + 3 ) ( 3 R 2 + 8 R + 3 ) R > 1 f(R):= \dfrac{15(R^2+1)(R^2 + R + 1)}{(3R^2 -2R +3)(3R^2+8R+3)} \quad\quad R> 1

A simple derivative test shows that f ( R ) f(R) is minimized when R 2 4 R + 1 = 0 R = 2 + 3 R^2-4R+1 = 0 \Rightarrow R = 2 + \sqrt3 . Thus, min ( f ( R ) ) = f ( 2 + 3 ) = 3 2 \min(f(R)) = f(2 + \sqrt3) = \frac32 .

Hence, tan ( θ ( 5 , 3 ) ) = min ( C P D ) = tan [ π + tan 1 ( 3 2 ) ] = 3 2 = 1.5 \tan( \theta(5,3)) = \min( \angle CPD) = \tan \left [ \pi + \tan^{-1} \left( \frac32 \right) \right ] = \frac32 = \boxed{1.5}

You're wrong when you say that θ \theta exists only when a > b a>b . It can exist for all rectangles, but the expression for θ \theta will be different for the cases where 5 a 2 < 9 b 2 5a^2<9b^2 .

The expression that you have derived works only for rectangles in which 5 a 2 > 9 b 2 5a^2>9b^2 .

Can you find the other expression? Firstly, try finding the locus associated with P P .

Digvijay Singh - 5 months, 1 week ago

Whooops .... lemme try to rectify it ...

Pi Han Goh - 5 months, 1 week ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...