Showing posts with label Italian Mathematical Olympiad. Show all posts
Showing posts with label Italian Mathematical Olympiad. Show all posts

Wednesday, November 30, 2011

Italian Mathematical Olympiad 1992 - Problem 6

Problem:
Let $a$ and $b$ be integers. Prove that if $\sqrt[3]{a}+\sqrt[3]{b}$ is a non-zero rational number, then both $a$ and $b$ are perfect cubes.

Solution:
Let $\sqrt[3]{a}+\sqrt[3]{b}=q \in \mathbb{Q}^*$. From the hypotheses, $a$ and $b$ can't be both zero: without loss of generality we can suppose that $a \neq 0$. Seeking a contradiction, assume that $a$ is not a perfect cube. By Gauss' Lemma, the polynomial $x^3-a$ is irreducible over $\mathbb{Q}$ since it is irreducible over $\mathbb{Z}$: in fact if it was reducible over $\mathbb{Q}$, there would exist $\alpha \in \mathbb{Z}$ such that $\alpha^3 - a = 0$, which contradicts our assumption. Therefore $x^3-a$ is the minimum polynomial of $\sqrt[3]{a}$ over $\mathbb{Q}$ and $[\mathbb{Q}(\sqrt[3]{a}):\mathbb{Q}]=3$. Then, from the initial equation we obtain
$$\sqrt[3]{b}=q-\sqrt[3]{a} \iff (a+b-q^3)+3q^2\sqrt[3]{a} -3q\sqrt[3]{a^2} = 0$$ and since
$\{1,\sqrt[3]{a}, \sqrt[3]{a^2}\}$ is a basis over $\mathbb{Q}$, it must be $a+b=q=0$, which contradicts the hypotheses.  

Italian Mathematical Olympiad 1994 - Problem 2

Problem:
Find all integer solutions of the equation $y^2=x^3+16$.

Solution:
We immediately see that $(0,-4)$ and $(0,4)$ are solutions of the given equation. We show that there are no other solutions. Rewriting the equation, we have $$(y-4)(y+4)=x^3.$$ Since the difference of the two factors on the left hand side is $8$, then $\gcd(y-4,y+4) \in \{1,2,4,8\}$ and we have $$y-4=2^n a, \qquad y+4=2^n b$$ with $n \in \{0,1,2,3\}, a,b \in \mathbb{Z}^*$, $\gcd(a,b)=1$. If $n=1,2$, from $2^n(b-a)=8$, we must have $a,b$ odd, but since $2^{2n}ab=x^3$, this is impossible by Unique-Prime-Factorization Theorem. If $n=3$, then $b-a=1$, and from $2^6a(a+1)=x^3$, we have that $a(a+1)$ is a perfect cube, and since $\gcd(a,a+1)=1$, $a$ and $a+1$ must be both perfect cubes, clearly impossible. At last $n=0$, i.e. $\gcd(y-4,y+4)=1$ and both $y-4$ and $y+4$ are perfect cubes whose difference is $8$. But this is impossible: in fact if there exist $k,m \in \mathbb{Z}$ such that $y-4=k^3$ and $y+4=m^3$, then $k^3$ and $m^3$ have the same parity and they are relatively primes, so they must be odd, and also $k$ and $m$ must be odd. Therefore, $$(m-k)(m^2+mk+k^2)=8$$ and by parity we force $$\begin{array}{ccc} m^2+mk+k^2 & = & 1 \\ m-k & = & 8, \end{array}$$ which implies $m^2+k^2=22 \equiv 6 \pmod{8}$, contradiction.

Italian Mathematical Olympiad 1989 - Problem 1

Problem:
Decide if the equation $x^2+xy+y^2=2$ has integer solutions $(x,y)$ where $x$ e $y$ are both rational numbers. 

Solution:
Suppose that the equation $x^2+xy+y^2=2$ has at least one rational solution $\left(\dfrac{x_1}{x_2}, \dfrac{y_1}{y_2} \right)$, where we can suppose $(x_1,x_2), (y_1, y_2) \in \mathbb{Z} \times \mathbb{Z^\ast}$, with $x_1, x_2$ relatively primes and $y_1, y_2$ relatively primes.
We get $$\left(\frac{x_1}{x_2}\right)^2+\frac{x_1}{x_2}\frac{y_1}{y_2}+\left(\frac{y_1}{y_2}\right)^2=2
\Longleftrightarrow x^2_1y^2_2 + x_1x_2y_1y_2 + x^2_2y^2_1 = 2x^2_2y^2_2.$$
Setting $a=x_1y_2, b=x_2y_1, c=x_2y_2$, we obtain the equation $$a^2+ab+b^2=2c^2.$$ Suppose that this equation has an integer solution $(a, b, c)$. If this one is a solution, then also $(-a, -b, -c)$ is a solution. Since $c \neq 0$, we can assume  $c > 0$ and choose $\max (a, b, c)>0$ as small as possible. Now, it is clear that $a$ and $b$ are both even, otherwise we have $1 \equiv 0 \pmod{2}$ and also $c$ is even, otherwise we have $0 \equiv 2 \pmod{4}$, contradiction. So, $a=2a_0, b=2b_0, c=2c_0$, $a_0,b_0,c_0 \in \mathbb{N}, c_0 > 0$, therefore we have $$4a^2_0+4a_0b_0+4b^2_0=8c^2_0 \iff a^2_0+a_0b_0+b^2_0=2c^2_0.$$ Then $(a_0, b_0, c_0)$ is a solution of the equation and $\max(a_0,b_0,c_0) < \max (a, b, c)$ which contradicts the minimality of $\max (a, b, c)$.

Italian Mathematical Olympiad 1990 - Problem 5

Problem: 
Prove that, for every integer $x$, the number $x^2+5x+16$ is not divisible by $169$.

Solution:
If $x^2+5x+16$ would be divisible by $169$, then would be divisible by $13$. Then,
$$x^2+5x+16 \equiv x^2-8x+16 = (x-4)^2 \equiv 0 \pmod{13}.$$ So, if $x \neq 4 + 13k$, $k \in \mathbb{Z}$, then $x^2+5x+16$ is not divisible by $13$ and, a fortiori, by $169$. Now, let $x=4+13k$ with $k \in \mathbb{Z}$. Hence,
$$(4+13k)^2+5(4+13k)+16=169(k^2+k)+52 \equiv 52 \pmod{169}$$
then $x^2+5x+16$ is not divisible by $169$ for $x=4+13k$ and the desired conclusion follows.