Non-Homogenous Diophantine Equation

a 2 + b + c = a b c \large a^2 + b + c = abc

How many triples ( a , b , c ) (a, b, c) of positive integers satisfy the above expression?


The answer is 9.

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.

3 solutions

Sharky Kesa
Dec 13, 2016

Rearrange this expression to get

b + c = a ( b c a ) b+c = a(bc-a)

Notice that the RHS should dominate as the variables get larger, especially as b b and c c get larger. Also note that if a b c a \geq bc , then there are no solutions because the R H S RHS is not positive but the LHS is.

Consider the graph of the parabola f ( x ) = x ( b c x ) f(x)=x(bc-x) . It is an upside-down parabola that crosses the x x -axis at x = 0 x=0 and x = b c x=bc . Thus, if f f is restricted to the interval 1 x b c 1 1 \leq x \leq bc-1 , the common minimum value being f ( 1 ) = b c 1 f(1)=bc-1 . Since 1 a b c 1 1\leq a \leq bc-1 , we have b + c = f ( a ) f ( 1 ) b+c=f(a)\geq f(1) . Thus, we must have

b c 1 b + c . bc-1 \leq b+c.

This certainly is a candidate for bounding! WLOG c b c\geq b . If b 3 b \geq 3 , then b c c + c + c > b + c + 1 bc \geq c+c+c > b+c+1 , which is a contradiction. Therefore, we must have either b = 1 b=1 or b = 2 b=2 .

Case 1: b = 1 b=1

The original equation can be rearranged to

c = a 2 + 1 a 1 = a + 1 + 2 a 1 c = \dfrac {a^2+1}{a-1} = a+1+\dfrac {2}{a-1}

Since 2 a 1 \frac {2}{a-1} must be an integer, this leads to a = 2 a=2 or a = 3 a=3 . The solutions corresponding to this case are then ( a , b , c ) = ( 2 , 1 , 5 ) (a, b, c) = (2, 1, 5) and ( 3 , 1 , 5 ) (3, 1, 5) .

Case 2: b = 2 b=2

The original equation can be rearranged to

c = a 2 + 2 2 a 1 c = \dfrac {a^2+2}{2a-1}

Here we will try to use a polynomial modulus to help simplify it.

a 2 + 2 0 ( m o d 2 a 1 ) 4 a 2 + 8 0 ( m o d 2 a 1 ) ( 2 a ) 2 + 8 0 ( m o d 2 a 1 ) 1 2 + 8 0 ( m o d 2 a 1 ) \begin{aligned} a^2 + 2 &\equiv 0 \pmod{2a-1}\\ 4a^2+8 &\equiv 0 \pmod{2a-1}\\ (2a)^2+8 &\equiv 0 \pmod{2a-1}\\ 1^2 + 8 &\equiv 0 \pmod{2a-1} \end{aligned}

Thus, 2 a + 1 9 2a+1 \mid 9 . Checking the six factors of 9, including the negative ones, gives rise to six solutions, namely, a = 4 , 1 , 0 , 1 , 2 , 5 a=-4, -1, 0, 1, 2, 5 . However, we wish a a to be positive, so a = 1 , 2 a=1, 2 or 5 5 . The solutions corresponding to this case are ( a , b , c ) = ( 1 , 2 , 3 ) , ( 2 , 2 , 2 ) (a, b, c) = (1, 2, 3), (2, 2, 2) and ( 5 , 2 , 3 ) (5, 2, 3) .

Thus, there are 5 5 solutions if b c b \leq c . We get another 5 if c b c \leq b , but ( 2 , 2 , 2 ) (2, 2 ,2) are common. Therefore, there are 9 solutions altogether.


Disclaimer: This problem and solution came from Problem Solving Tactics.

Why does this look familiar...

Wen Z - 4 years, 6 months ago
Kushal Bose
Dec 14, 2016

a 2 + b + c = a b c a 2 a . ( b c ) + ( b + c ) = 0 a^2+b+c=a b c \\ \implies a^2-a.(b c)+(b+c)=0 .

Let, a a has two solutions in positive integers m , n m,n , so,

( a m ) ( a n ) = a 2 a . ( b c ) + ( b + c ) (a-m)(a-n)=a^2-a.(b c)+(b+c) . Comparing coefficients we get

m + n = b c and m n = b + c m+n=b c \,\, \text{and} \,\, m n=b+c .This ordered pairs has a finite number of solutions.

Firstly ,put b = c b=c we get a unique solution b = c = 2 b=c=2 .So, a ordered pair is ( 2 , 2 , 2 ) (2,2,2) . Next if b c b \neq c then

there will be two solutions ( b , c ) = ( 2 , 3 ) ; ( 1 , 5 ) (b,c)=(2,3);(1,5)

If one of m , n m,n is negative then ( a m ) ( a + n ) = a 2 a . ( b c ) + ( b + c ) (a-m)(a+n)=a^2-a.(b c)+(b+c) now

m n = b c m-n=b c and b + c = m n b+c=-m n .Here m m is a positive solution and n n is a negative solution.As b c > 0 bc >0 then m n > 0 m > n m-n >0 \implies m>n .Again b + c > 0 m n > 0 m n < 0 n < 0 b+c>0 \implies -mn >0 \implies mn<0 \implies n<0 .First we assume that n is a negative solution and last we get n < 0 n<0 so n n becomes positive.So both should be positive.

Lastly if both m , n m,n shouldn't be negative because solutions should be in positive integers.

So, the solutions are ( 2 , 2 , 2 ) ; ( 1 , 2 , 3 ) , ( 1 , 3 , 2 ) ; ( 2 , 1 , 5 ) ; ( 2 , 5 , 1 ) ; ( 3 , 1 , 5 ) ; ( 3 , 5 , 1 ) ; ( 5 , 2 , 3 ) ; ( 5 , 2 , 3 ) (2,2,2);(1,2,3),(1,3,2);(2,1,5);(2,5,1);(3,1,5);(3,5,1);(5,2,3);(5,2,3)

Kazem Sepehrinia
Dec 20, 2016

This is essentially the same as problem 4 of IMO 1994, because c = a 2 + b a b 1 c=\frac{a^2+b}{ab-1} Now we have a b 1 a 2 + b ab-1 | a^2+b or a b 1 a ( a 2 + b ) ( a b 1 ) = a 3 + 1 ab-1 | a(a^2+b)-(ab-1)=a^3+1 . Thus a 3 + 1 a b 1 \frac{a^3+1}{ab-1} must be an integer. @Wen Z

And I like this problem. Working with ( n , m ) (n, m) instead of ( a , b ) (a, b) .

First note that if m n 1 n 3 + 1 mn-1|n^3+1 then m n 1 m 3 n 3 1 + n 3 + 1 = n 3 ( m 3 + 1 ) mn-1|m^3n^3-1+n^3+1=n^3(m^3+1) It follows that m n 1 m 3 + 1 mn-1|m^3+1 because gcd ( m n 1 , n 3 ) = 1 \gcd(mn-1, n^3)=1 , so if ( m , n ) (m,n) is an answer ( n , m ) (n,m) will be too. If n = m n=m we get n 3 + 1 m n 1 = n 3 + 1 n 2 1 = n + 1 n 1 \frac{n^3+1}{mn-1}=\frac{n^3+1}{n^2-1}=n+\frac{1}{n-1} So n 1 1 n-1|1 and n = 2 n=2 . Therefore ( 2 , 2 ) (2,2) is an answer and from now on we can suppose m > n m>n . Let n = 1 n=1 , it gives m = 2 , 3 m=2,3 , we get four answers from here: ( 1 , 2 ) , ( 2 , 1 ) , ( 1 , 3 ) , ( 3 , 1 ) (1,2),(2,1),(1,3),(3,1) Note that if n 3 + 1 m n 1 = r \frac{n^3+1}{mn-1}=r then r n 1 r\stackrel{n}{\equiv} -1 , and r r have the form of k n 1 kn-1 for some positive integer k k . Since m > n m>n k n 1 = n 3 + 1 m n 1 < n 3 + 1 n 2 1 = n + 1 n 1 k n 1 < n + 1 n 1 k < 1 + 1 n + 1 n ( n 1 ) 2 k < 2 kn-1=\frac{n^3+1}{mn-1}<\frac{n^3+1}{n^2-1}=n+\frac{1}{n-1} \\ kn-1<n+\frac{1}{n-1} \\ k<1+\frac{1}{n}+\frac{1}{n(n-1)} \le 2 \\ k<2 Thus k = 1 k=1 and we have ( n 1 ) ( m n 1 ) = n 3 + 1 (n-1)(mn-1)=n^3+1 , hence n 1 n 3 + 1 n-1|n^3+1 or n 1 n 3 + 1 ( n 3 1 ) = 2 n 1 2 n-1|n^3+1-(n^3-1)=2 \\ n-1|2 As a result n = 2 n=2 or n = 3 n=3 which gives us 2 m 1 = 9 2m-1=9 or 2 ( 3 m 1 ) = 28 2(3m-1)=28 and both of them leads to m = 5 m=5 , so we get four other answers from here: ( 2 , 5 ) , ( 5 , 2 ) , ( 3 , 5 ) , ( 5 , 3 ) (2,5),(5,2),(3,5),(5,3) .

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...