MHB Inequality Involves The Sides Of Triangle

AI Thread Summary
The discussion revolves around proving the inequality \( a^2yz + b^2xz + c^2xy \le 0 \) for the sides of a triangle \( a, b, c \) and real numbers \( x, y, z \) that sum to zero. Participants express gratitude for contributions and attempt to clarify their solutions. One user acknowledges a misunderstanding of the problem and admits their proposed solution is invalid. The conversation highlights the collaborative nature of problem-solving in mathematical discussions. Overall, the focus remains on addressing the inequality and refining the approach to the proof.
anemone
Gold Member
MHB
POTW Director
Messages
3,851
Reaction score
115
Let $a,\,b$ and $c$ be the sides of a triangle and $x,\,y$ and $z$ are real numbers such that $x+ y+ z = 0$. Prove that $a^2yz +b^2xz+c^2xy\le 0$.
 
Mathematics news on Phys.org
Here is my solution.
Without loss of generality, assume $c$ is the longest side of the triangle. If the triangle is obtuse or right, then $c^2 \ge a^2 + b^2$. So since $z = -x - y$, then $a^2yz + b^2xz + c^2xy = -(a^2y^2 - b^2x^2) + (c - a^2 - b^2)xy.$ Now $a^2y^2 + b^2x^2 \ge 2ab|xy| = [(a + b)^2 - a^2 - b^2]|xy| \ge (c^2 - a^2 - b^2)|xy|,$ where the last step follows from the triangle inequality $a + b > c$. Thus $-(a^2y^2 + b^2x^2) \le -(c^2 - a^2 - b^2)|xy|$, forcing

$$-(a^2y^2 + b^2x^2) + (c^2 - a^2 - b^2)xy \le -(c^2 - a^2 - b^2)|xy| + (c^2 - a^2 - b^2)xy = (c^2 - a^2 - b^2)[-|xy| + xy] \le 0.$$

Now suppose the triangle is acute. Then $c^2 < a^2 + b^2$. Since $c$ is the longest side, we also have $b^2 < a^2 + c^2$, and $a^2 < b^2 + c^2$. We can write

$$xy = \frac{z^2 - x^2 - y^2}{2},\quad yz = \frac{x^2 - y^2 - z^2}{2},\quad \text{and} \quad xz = \frac{y^2 - x^2 - z^2}{2},$$

which are derived from the equations $z^2 = (x + y)^2$, $x^2 = (y + z)^2$, and $y^2 = (x + z)^2$, respectively. Thus

$$a^2yz + b^2xz + c^2xy = -\frac{(b^2 + c^2 - a^2)x^2 + (c^2 + a^2 - b^2)y^2 + (a^2 + b^2 - c^2)z^2}{2} \le 0.\quad \text{QED}$$
 
Well done Euge! And thanks for participating!

Here's my solution:

Since $a,\,b$ and $c$ are the lengths of a triangle we have:

$a^2≤(b+c)^2, \,b^2≤(a+c)^2, \,c^2≤(a+b)^2$ and so

$\begin{align*}a^2x+b^2y+c^2z&\le (b+c)^2x+(a+c)^2y+(a+b)^2z\\&=a^2(y+z)+b^2(x+z)+c^2(x+y)+2abz+2acy+2bcx\\&=a^2(-x)+b^2(-y)+c^2(-z)+2abz+2acy+2bcx\end{align*}$

and so our mission now has turned into proving

$2(a^2x+b^2y+c^2z)\le 2abz+2acy+2bcx$, i.e.

$a^2x+b^2y+c^2z \le abz+acy+bcx$, but that is true from the Chebyshev's inequality which says if

$ab\ge ac \ge bc$ and $x\ge y \ge z$,

$$abz+acy+bcx \le \frac{(ab+ac+bc)(x+y+z)}{3}=0$$ since $x+y+z=0.$

Therefore we have:

$a^2x+b^2y+c^2z \le abz+acy+bcx \le 0$ (Q.E.D.)
 
anemone said:
Well done Euge! And thanks for participating!

Here's my solution:

Since $a,\,b$ and $c$ are the lengths of a triangle we have:

$a^2≤(b+c)^2, \,b^2≤(a+c)^2, \,c^2≤(a+b)^2$ and so

$\begin{align*}a^2x+b^2y+c^2z&\le (b+c)^2x+(a+c)^2y+(a+b)^2z\\&=a^2(y+z)+b^2(x+z)+c^2(x+y)+2abz+2acy+2bcx\\&=a^2(-x)+b^2(-y)+c^2(-z)+2abz+2acy+2bcx\end{align*}$

and so our mission now has turned into proving

$2(a^2x+b^2y+c^2z)\le 2abz+2acy+2bcx$, i.e.

$a^2x+b^2y+c^2z \le abz+acy+bcx$, but that is true from the Chebyshev's inequality which says if

$ab\ge ac \ge bc$ and $x\ge y \ge z$,

$$abz+acy+bcx \le \frac{(ab+ac+bc)(x+y+z)}{3}=0$$ since $x+y+z=0.$

Therefore we have:

$a^2x+b^2y+c^2z \le abz+acy+bcx \le 0$ (Q.E.D.)

Thankyou very much anemone and Euge
The purpose of the challenge was to prove, that $a^2yz+b^2zx+c^2xy \le 0$ but it seems to me, that anemones end result is a different one?? Maybe I am wrong ...? Thankyou very much for clearing the matter. :o
 
Last edited:
Hi lfdahl,

It appears anemone tried to prove a different inequality than the one in the given problem, but even so there is an error in that solution: since $x$, $y$, and $z$ are arbitrary real numbers, it is not necessarily true that $a^2x + b^2y + c^2z \le (b + c)^2x + (c + a)^2y + (a + b)^2z$. In fact, if $x$, $y$, and $z$ are negative, then $a^2x + b^2y + c^2z > (b + c)^2x + (c + a)^2y + (a + b)^2z$. Take for instance $a = 3$, $b = 4$, $c = 5$, and $x = y = z = -1$. Then $a^2x + b^2y + c^2z = -50$, but $(b + c)^2x + (c + a)^2y + (a + b)^2z = -155 < -50$.
 
Thank you Euge, but

$x = y = z = -1$ is not allowed, albeit: $x + y + z = 0$ :confused:
 
Oops, I accidentally edited your post with my message, rather than post mine with a "reply with quote"! Ifdahl, if you could please, rewrite your statement over mine in an edit.

Edit: I tried to put back your question -- does it look correct? My response was this:

Sorry, I overlooked the condition $x + y + z = 0$. But the result will not hold: If $a = 3$, $b = 4$, $c = 5$, $x = -1$, $y = 1$, and $z = 0$, then $x + y + z = 0$ and the triangle with sides $a,b,c$ form a right triangle, but $a^2x + b^2y + c^2z = -9 + 16 + 0 = 7 > 0$.
 
Thankyou Euge for the edit!My suggested solution:

WLOG, I can choose $x,y$ and $z$ such, that: $yz \ge zx \ge xy$. On the other hand, I can also choose the sides in a triangle, such that: $a^2 \le b^2 \le c^2$.

Applying Chebyshev´s sum inequality:\[3\left ( a^2yz+b^2zx+c^2xy \right ) \le (a^2+b^2+c^2)(xy+zx+yz) \\\\ 2\left ( a^2yz+b^2zx+c^2xy \right ) \le a^2x(y+z)+b^2y(x+z)+c^2z(x+y)=-a^2x^2-b^2y^2-c^2z^2\leq 0\]Thus \[a^2yz+b^2zx+c^2xy \leq 0\]
 
Last edited:
lfdahl said:
Thankyou Euge for the edit!My suggested solution:

WLOG, I can choose $x,y$ and $z$ such, that: $yz \ge zx \ge xy$. On the other hand, I can also choose the sides in a triangle, such that: $a^2 \le b^2 \le c^2$.

Applying Chebyshev´s sum inequality:\[3\left ( a^2yz+b^2zx+c^2xy \right ) \le (a^2+b^2+c^2)(xy+zx+yz) \\\\ 2\left ( a^2yz+b^2zx+c^2xy \right ) \le a^2x(y+z)+b^2y(x+z)+c^2z(x+y)=-a^2x^2-b^2y^2-c^2z^2\leq 0\]Thus \[a^2yz+b^2zx+c^2xy \leq 0\]
in fact $3\left ( a^2yz+b^2zx+c^2xy \right ) \le (a^2+b^2+c^2)(xy+zx+yz)\leq 0$
for $ xy+yz+zx\leq 0$
hence
$\left ( a^2yz+b^2zx+c^2xy \right ) \leq 0$
 
  • #10
anemone said:
Let $a,\,b$ and $c$ be the sides of a triangle and $x,\,y$ and $z$ are real numbers such that $x+ y+ z = 0$. Prove that $a^2yz +b^2xz+c^2xy\le 0$.
my solution:
$\because (x+y+z)^2=x^2+y^2+z^2+2(xy+yz+zx)=0\\
\therefore yz+zx+xy\leq 0----(1)\\
(1)\times a^2:a^2yz+a^2zx+a^2xy\leq 0---(2)\\
(1)\times b^2:b^2yz+b^2zx+b^2xy\leq 0---(3)\\
(1)\times c^2:c^2yz+c^2zx+c^2xy\leq 0---(4)\\
(2)+(3)+(4): (a^2+b^2+c^2)(yz+zx+xy)\leq 0---(5)\\$
now we may set $yz\geq zx\geq xy$
and $a^2\leq b^2\leq c^2 (or \,\,a \leq b\leq c)$
applying Chebyshev's sum inequality we get:
$3(a^2yz+b^2zx+c^2xy)\leq (a^2+b^2+c^2)(yz+zx+xy)\leq 0$
that is :
$a^2yz+b^2zx+c^2xy\leq 0$
 
  • #11
Albert said:
my solution:
$\because (x+y+z)^2=x^2+y^2+z^2+2(xy+yz+zx)=0\\
\therefore yz+zx+xy\leq 0----(1)\\
(1)\times a^2:a^2yz+a^2zx+a^2xy\leq 0---(2)\\
(1)\times b^2:b^2yz+b^2zx+b^2xy\leq 0---(3)\\
(1)\times c^2:c^2yz+c^2zx+c^2xy\leq 0---(4)\\
(2)+(3)+(4): (a^2+b^2+c^2)(yz+zx+xy)\leq 0---(5)\\$
now we may set $yz\geq zx\geq xy$
and $a^2\leq b^2\leq c^2 (or \,\,a \leq b\leq c)$
applying Chebyshev's sum inequality we get:
$3(a^2yz+b^2zx+c^2xy)\leq (a^2+b^2+c^2)(yz+zx+xy)\leq 0$
that is :
$a^2yz+b^2zx+c^2xy\leq 0$
applying law of symmetry you can set
either $yz\geq zx\geq xy$ or $a^2\leq b^2\leq c^2 (or \,\,a \leq b\leq c)$ but not both as a gets tied to yz , b to xz , c to xy and any random combinations
 
  • #12
kaliprasad said:
applying law of symmetry you can set
either $yz\geq zx\geq xy$ or $a^2\leq b^2\leq c^2 (or \,\,a \leq b\leq c)$ but not both as a gets tied to yz , b to xz , c to xy and any random combinations
sure ! with different assumption,we may obtain distinct result
 
Last edited:
  • #13
Sorry folks! Apparently my solution doesn't work as I have somehow misinterpreted the problem, which was silly mistake from my part...:o

Sorry too for my late reply as I have been pretty busy for both family and personal issue!
 
  • #14
kaliprasad said:
applying law of symmetry you can set
either $yz\geq zx\geq xy$ or $a^2\leq b^2\leq c^2 (or \,\,a \leq b\leq c)$ but not both as a gets tied to yz , b to xz , c to xy and any random combinations

You´re right, kaliprasad, therefore my suggested solution in #8 is not valid!
 
Back
Top