1
$\begingroup$

Solve the system of equations :

(EDIT : The problem does not say anything about the nature of $x$ and $y$ (integer, natural number ,..etc.) )

$4xy + 4(x^2 + y^2) + {\frac {3} { (x+y) ^ 2 } } = \frac {85} {3} $

$2x + {\frac {1} {x+y}} = \frac {13} {3}$

I do not know how to approach these types of problems. I tried finding value of $\frac {1} {x+y}$ in terms of $x$ and $y$, but it complicates the problem even more.

Can anyone provide a pointer to what should be done ?

  • 2
    Just checking: the first term in the second equation is $x$ and not $xy$ – 2017-01-12
  • 0
    Also, does the problem specify any restrictions on the type of values for x,y? For example, are x,y required to be integers? – 2017-01-12
  • 0
    There are no "restrictions" as such mentioned in the problem. It just asks you to "solve" the system of equations... – 2017-01-12
  • 1
    @QUANTUM: What do you mean by "it is $4xy$ only"? The question Anurag A asked was whether the first term in the second equation should be $2x$ or $2xy$? – 2017-01-12
  • 0
    @QUANTUM: Can you tell us the source of the problem? Is it from a book? – 2017-01-12
  • 0
    Please don't confuse me ... Everything I know about the problem is given in the question details ... Regarding the equations, they are as they have been specified... I have not made up anything. – 2017-01-12
  • 0
    IT is ... Question Number "244.5" – 2017-01-12
  • 0
    Regarding Anurag's comment, the $first$ term in the $second$ equation is $2x$ and not $x$. I have given the link for the source of problem. Anyone who is having problem can check it out... – 2017-01-12
  • 0
    At least provide the sort of $approach$ needed to solve such problems.... – 2017-01-12
  • 0
    Ok, I take back my claim that the problem is wrong (and I deleted my flawed comment). Anurag A's solution looks good. – 2017-01-12

2 Answers 2

6

Perhaps this can help: Let $x+y=v$ and $x-y=w$, then the equations can be written as \begin{align*} 3\left(v^2+\frac{1}{v^2}\right)+w^2 & = \frac{85}{3}\\ \left(v+\frac{1}{v}\right)+w & = \frac{13}{3}. \end{align*} Now let $v+\frac{1}{v}=t$, then the system can be rewritten as \begin{align*} 3t^2+w^2 & = \frac{85}{3}+6\\ t+w & = \frac{13}{3}. \end{align*} Now solve for $t$ and $w$...

  • 0
    Very nice. It should be $\frac {85}{3}-6$ though, should it not? – 2017-01-12
  • 0
    Yes, very nice. The answers for x,y are ugly, but at least they're symbolic. – 2017-01-12
  • 1
    @quasi, real solutions are $x = 2/3, y = -1/3$, $x = 2$, $y = 1$, not that ugly. – 2017-01-12
  • 0
    @Ennar -- I must have made an algebraic error -- will check. – 2017-01-12
  • 0
    @Ennar -- yes, those solutions work, my mistake. But there's also a non-real, complex solution which, though symbolic, is a little bit ugly. – 2017-01-12
  • 0
    @Ennar I think it should be +6, or am I making a silly error? – 2017-01-12
  • 0
    No, you are right, I'm the one that is silly, @Anurag A. – 2017-01-12
1

Less elegant than Anurag A's answer and using brute force.

Considering the equations $$4xy + 4(x^2 + y^2) + {\frac {3} { (x+y) ^ 2 } } = \frac {85} {3}\tag 1$$ $$2x + {\frac {1} {x+y}} = \frac {13} {3}\tag 2$$ extract $y$ from $(2)$; this gives $$y=\frac{-6 x^2+13 x-3}{6 x-13}\tag 3$$ Replace $(3)$ in $(1)$ an simplify to get $$\frac{8 (x-2) (3 x-2) \left(24 x^2-118 x+149\right)}{(13-6 x)^2}=0\tag 4$$ and the quadratic term does not show real solution (so,two real roots and two complex conjugate roots for $x$).

The solutions of $(4)$ are simple; when you have them, go back to $(3)$ for the corresponding $y$'s.