In the diagram below, two semicircles and one quarter-circle are inscribed in a unit square.
Inside one of the partitioned sections, infinitely many circles are drawn that are tangent to each other and to the boundaries. Let r n denote the radius of the n th largest circle.
Express r n in terms of n .
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.
Could u please explain how one obtains the general relationship of k(n) and n given the dynamic relationship between k(n+1) and k(n) ?
Log in to reply
For simplicity, we can just check that, if k n = 4 1 ( 4 n 2 + 1 2 n + 1 7 ) then k n + 1 = k n + 1 + 4 ( k n − 2 ) = 4 1 ( 4 n 2 + 2 0 n + 3 3 ) which proves the result by induction.
A better answer is to note that k n + 1 − 2 = k n − 2 + 1 and since k 1 − 2 = 2 5 we deduce that k n − 2 = 2 1 ( 2 n + 3 ) , from which the result is immediate.
Log in to reply
But how do u get the 4n^2+12n+...in the first place ? Is it pure observation / pattern recognition ?
Log in to reply
@Sundar R – My first solution is one of checking that the answer works...
My second approach finds a way of writing the inductive relation which enables us to solve it...
@Sundar R – Given k 1 = 4 3 3 and k n + 1 = k n + 1 + 4 ( k n − 2 ) , we can calculate the first few terms as 4 3 3 , 4 5 7 , 4 8 9 , 4 1 2 9 , 4 1 7 7 , . . . Then using the Method of Differences , we find that the first difference is 6 , 8 , 1 0 , 1 2 , . . . and the second difference is 2 , 2 , 2 , . . . Since the second difference is constant, the relationship can be expressed as a quadratic in the form of k = a n 2 + b n + c . Using the first three terms, when n = 1 then k = 4 3 3 , when n = 2 then k = 4 5 7 , and when n = 3 then k = 4 8 9 , we have a system of equations a + b + c = 4 3 3 , 4 a + 2 b + c = 4 5 7 , and 9 a + 3 b + c = 4 8 9 , and solving these we can find that a = 1 , b = 3 , and c = 4 1 7 , meaning the relationship can be explicitly written as k n = n 2 + 3 n + 4 1 7 or k n = 4 1 ( 4 n 2 + 1 2 n + 1 7 ) .
Log in to reply
@David Vreken – Just testing the first five values in a sequence, and using the Method of Differences on those first few terms, is not enough to know that the pattern holds for all terms in the sequence. Plenty of elementary maths problems are stated like "what are the next two terms in the sequence 1,3,5,7,9,...?" (and assume that just because a pattern holds for the first few terms, it keeps on holding). This is not always true.
Example: What is the next term in the sequence 1,2,4,8,16,...?
The answer could be 32, but it need not be. If the rule defining the sequence was "what is the largest number of regions into which a disk can be divided if n joints on its circumference are joined by straight lines?", the answer is 3 1 .
Getting the recurrence relation k n + 1 − 2 = k n − 2 + 1 proves that the Method of Differences works for the sequence k n − 2 for all n . This is what enables me to calculate k n precisely.
Log in to reply
@Mark Hennings – Yes, you are correct, and I apologize for the incomplete argument. Once you know that the relationship k n = 4 1 ( 4 n 2 + 1 2 n + 1 7 ) holds for the first few terms, you can prove that the relationship holds for all terms by observing that k n = 4 1 ( ( 2 n + 3 ) 2 + 8 ) and that k n + 1 = 4 1 ( 4 ( n + 1 ) 2 + 1 2 ( n + 1 ) + 1 7 ) = 4 1 ( ( 2 n + 5 ) 2 + 8 ) , which means that 2 n + 5 = 4 ( k n + 1 − 2 ) = 4 ( k n − 2 ) + 2 and from this you can derive that k n + 1 = k n + 1 + 4 ( k n − 2 ) .
But I now see that your method of just using ( k n + 1 − 2 ) = ( k n − 2 ) + 1 is much shorter than mine.
Thanks. I saw the above only after i posted my reply to ur solution.
I apologize for frequently asking, but I'm curious. How exactly you deduce k n − 2 = 2 1 ( 2 n + 3 ) ? How do you justify letting k n − 2 be of the form a n + b ?
Log in to reply
@Uros Stojkovic – It follows by induction from the identity k n + 1 − 2 = k n − 2 + 1 and the known value of k 1 .
Congrats!!
I was wondering if one could get closed-form expression just by directly solving recurrence relation k n + 1 = k n + 1 ± 2 k n − 2 and avoiding Method of Differences.
If there is a way of solving this recurrence relation, could you write it here or at least give instructions? Thanks in advance.
Log in to reply
I think the method of solution I have given is about as direct as possible.
Setting the bottom left corner of the square as the origin, we let x , y , r denote the x-coordinates, y-coordinates, and radius of the first circle.
because the circle touches the three larger circles, by the Pythagorean theorm we have
⎩ ⎪ ⎪ ⎪ ⎪ ⎨ ⎪ ⎪ ⎪ ⎪ ⎧ ( x − 1 ) 2 + y 2 = ( 1 − r ) 2 ( 1 ) x 2 + ( y − 2 1 ) 2 = ( 2 1 − r ) 2 ( 2 ) ( x − 2 1 ) 2 + y 2 = ( 2 1 + r ) 2 ( 3 )
by ( 1 ) − ( 3 ) we get x = 3 r
by ( 1 ) − ( 2 ) we get y = 5 r
substituding x = 3 r and y = 5 r in ( 1 ) we obtain an equation in r
( 3 r − 1 ) 2 + 2 5 r 2 = ( 1 − r ) 2
which yields r = 3 3 4
thus we see that r n = 4 n 2 + 1 2 n + 1 7 4 is the correct answer since it is the only option for which r 1 = 3 3 4
can you please explain where you apply the Pythagorean theorem 3 times? Thanks! - Actually got it - I think it would suffice to say that you draw lines from the center of 3 circles to first circle center and then you can derive the hypotenuses and the corresponding right triangles.. I was confused about (x-1)^2 which actually is (1-x)^2 even if the result is correct x-1 is less than zero
Or you could just gauge out of the picture that r1 has to be about 1/8
Wow! It was an easier explanation for the ones who do not know the Descartes theorem.
This is same as mine! I don't know such a deep theorem, so done by this way!
Use inversive geometry , by choosing O on button left corner of the square and r = 1 .
Then two semicircles and the quarter-circle will become three straight lines after transformation. And the circles you are looking for will all become circles with r n ′ = 1 / 4 , where C n = ( 4 3 , 4 3 + 2 n )
To find r n , you need to do inverse transform on each circles. And you will get,
r n = 2 1 ( ( 4 3 ) 2 + ( 4 3 + 2 n ) 2 − 4 1 1 − ( 4 3 ) 2 + ( 4 3 + 2 n ) 2 + 4 1 1 )
r n = 4 n 2 + 1 2 n + 1 7 4
By C n you meen the center of the inverse of the nth circle? It that case why do you don't refer to it as (C n)'? And very nice and simple solution!
It's cool man
Smart and simple solution.
I did it the same way as Daniel Xiang to get r 1 . Given the answer choices, this was the only possibility the worked. However, from that starting point, it is possible to continue on to derive the formula given as the answer.
Of Daniel's three numbered equations, the following two apply to all the circles:
\begin{aligned} (x-1)^2 + y^2 &= (1-r)^2 \tag{1}\\ (1/2-x)^2 + y^2 &= (1/2+r)^2 \tag{3} \end{aligned}
from which we still obtain x = 3 r . Plugging this result into the first equation, we obtain
y 2 = 4 r ( 1 − 2 r ) .
This doesn't simplify nicely for y , but if we define k ≡ y / r , then we can simplify the above equation as
k 2 k 2 + 8 r = 4 ( 1 / r − 2 ) = 4 / r = k 2 + 8 4
This leads to
x y = k 2 + 8 1 2 = k 2 + 8 4 k
Now we have all three relevant variables in terms of k . These relations still apply to all the circles. Next we need to find a relation between k n + 1 and k n . To do so, we use the distance between the centers of two sequential circles.
( x n + 1 − x n ) 2 + ( y n + 1 − y n ) 2 = ( r n + 1 + r n ) 2
Subtracting off two copies of equation (1) (one for each circle), this becomes
x n + 1 + x n − 2 x n + 1 x n − 2 y n + 1 y n 4 r n + 1 + 4 r n − 2 y n + 1 y n k n + 1 2 + 8 1 6 + k n 2 + 8 1 6 − ( k n + 1 2 + 8 ) ( k n 2 + 8 ) 3 2 k n + 1 k n k n 2 + 8 + k n + 1 2 + 8 − 2 k n + 1 k n ( k n + 1 − k n ) 2 k n + 1 = 2 r n + 1 r n − r n + 1 − r n = 2 0 r n + 1 r n = ( k n + 1 2 + 8 ) ( k n 2 + 8 ) 3 2 0 = 2 0 = 4 = k n ± 2
Clearly the k n values are increasing, so k n + 1 = k n + 2 . And since k 1 = 5 , we find the simple result k n = 2 n + 3 . Then
r n = k n 2 + 8 4 = ( 2 n + 3 ) 2 + 8 4 = 4 n 2 + 1 2 n + 1 7 4
Use the method of substitution. Find the radius of r 1 analytically. Then put n = 1 in the options. The one that satisfies the answer found will be the right option. Just to be safe, find r 2 to eliminate 3 options.
I didn't get it right for I think I might have done a sign error somewhere, but the method by inversion of circles seems to be the way to go for me... (inverting with respect of a circle of radius 1 centred on the bottom left corner)...
Log in to reply
I like the fact that you want to complicate your way.
We can look at the area of the segment determined by the three circles (in which we place the infintely many little circles). We choose the origin of our coordinates to be the bottom left corner of the unit square. We parametrize the circles as ( x − 1 ) 2 + y 2 = 1 ( x − 2 1 ) 2 + y 2 = 4 1 x 2 + ( y − 2 1 ) 2 = 4 1 We can easily see that the intersection points are ( 0 , 0 ) , ( 2 1 , 2 1 ) , ( 5 2 , 5 4 ) .Then the area of the segment can be calculated: A = ∫ 0 5 2 d x 1 − ( x − 1 ) 2 + ∫ 5 2 2 1 d x ( 2 1 + 4 1 − x 2 ) − ∫ 0 2 1 d x 4 1 − ( x − 2 1 ) 2 The answer is 0 . 0 9 7 7 . If we add up all the areas of the circles inscribed in that segment, we should defintely get something smaller than A , but close enough. If we try the radii sequence r n = 4 n 2 + 1 2 n + 1 7 4 , we get: n = 1 ∑ ∞ π r n 2 = 0 . 0 7 5 4 This is the smallest radii sequence of all 4. We can also try to add up the areas of circles with the other sequences. We will find larger numbers than A. So, the third option is the correct.
Problem Loading...
Note Loading...
Set Loading...
Relevant wiki: Descartes' Circle Theorem
For any n ≥ 1 , the circle with radius 1 is the exterior Soddy circle to the three mutually tangent circles of radius 2 1 , r n , r n + 1 , where we are using the circle of radius 2 1 with centre on the bottom edge of the square. Using Descartes' Theorem, we have 2 ( k n 2 + k n + 1 2 + 2 2 + ( − 1 ) 2 ) ( k n − k n + 1 ) 2 = ( k n + k n + 1 + 2 − 1 ) 2 = 2 ( k n + k n + 1 ) − 9 ( ⋆ ) where k n = r n − 1 is the curvature of the circle with radius r n . If the circle with radius r 1 has centre ( x , y ) (where we set up coordinates with the origin at the bottom-left corner of the square), then ( x − 2 1 ) 2 + y 2 x 2 + ( y − 2 1 ) 2 ( x − 1 ) 2 + y 2 = ( 2 1 + r 1 ) 2 = ( 2 1 − r 1 ) 2 = ( 1 − r 1 ) 2 and hence x = 3 r 1 and y − x = 2 r 1 , so that y = 5 r 1 . Thus ( 3 r 1 − 1 ) 2 + 2 5 r 1 2 = ( 1 − r 1 ) 2 , and hence r 1 = 3 3 4 . It is clear from ( ⋆ ) that k n + 1 = k n + 1 ± 4 ( k n − 2 ) Only one of these roots is greater than k n , and it is clear that k n + 1 > k n . Thus we have k n + 1 k n + 1 − 2 k n − 2 k n = k n + 1 + 2 k n − 2 = k n − 2 + 1 = 2 1 ( 2 n + 3 ) = 4 1 ( 4 n 2 + 1 2 n + 1 7 ) and hence r n = 4 n 2 + 1 2 n + 1 7 4