A Potpourri of Math Problems
by djmathman, Aug 30, 2013, 3:39 AM
from three different subjects. That's what I dealt with today.
Also combo isn't on here because combo is evil
Short
Solution
This LTE thing is pretty cool
Also combo isn't on here because combo is evil

MOSP 1999 wrote:
Prove that for any positive real numbers 
![\[\dfrac1{a+b}+\dfrac1{b+c}+\dfrac1{c+a}+\dfrac1{2\sqrt[3]{abc}}\geq\dfrac{(a+b+c+\sqrt[3]{abc})^2}{(a+b)(b+c)(c+a)}.\]](//latex.artofproblemsolving.com/3/9/f/39f12ad42239ffd89a5f390ba67b83304cdf77ec.png)

![\[\dfrac1{a+b}+\dfrac1{b+c}+\dfrac1{c+a}+\dfrac1{2\sqrt[3]{abc}}\geq\dfrac{(a+b+c+\sqrt[3]{abc})^2}{(a+b)(b+c)(c+a)}.\]](http://latex.artofproblemsolving.com/3/9/f/39f12ad42239ffd89a5f390ba67b83304cdf77ec.png)
Short
Note that after some manipulation the LHS becomes
![\[\dfrac{c^2}{c^2(a+b)}+\dfrac{a^2}{a^2(b+c)}+\dfrac{b^2}{b^2(c+a)}+\dfrac{(\sqrt[3]{abc})^2}{2abc},\]](//latex.artofproblemsolving.com/e/d/e/edea28d55d50d4f187bf2cbd4cab2759452062f7.png)
which by Titu's Lemma is greater than or equal to
![\[\dfrac{(a+b+c+\sqrt[3]{abc})^2}{a^2(b+c)+b^2(c+a)+c^2(a+b)+2abc}=\dfrac{(a+b+c+\sqrt[3]{abc})^2}{(a+b)(b+c)(c+a)},\]](//latex.artofproblemsolving.com/9/c/6/9c66941c795b164a72de57df8fecaef6702763eb.png)
as desired.
![\[\dfrac{c^2}{c^2(a+b)}+\dfrac{a^2}{a^2(b+c)}+\dfrac{b^2}{b^2(c+a)}+\dfrac{(\sqrt[3]{abc})^2}{2abc},\]](http://latex.artofproblemsolving.com/e/d/e/edea28d55d50d4f187bf2cbd4cab2759452062f7.png)
which by Titu's Lemma is greater than or equal to
![\[\dfrac{(a+b+c+\sqrt[3]{abc})^2}{a^2(b+c)+b^2(c+a)+c^2(a+b)+2abc}=\dfrac{(a+b+c+\sqrt[3]{abc})^2}{(a+b)(b+c)(c+a)},\]](http://latex.artofproblemsolving.com/9/c/6/9c66941c795b164a72de57df8fecaef6702763eb.png)
as desired.

Panafrican 2013.1 wrote:
Let
be a convex quadrilateral with
parallel to
. Let
and
be the midpoints of
and
, respectively. Prove that if
, then
.









Solution
Let
, and let
be the feet of the altitudes from
to
respectively. Note that by the angle equality we have that
and
are isogonal conjugates of each other, but since
is a median of
it follows that
is the
-symmedian of said triangle. Therefore
. But since
,
, and multiplying these two ratios together gives
. Therefore
is the
-symmedian of
and
are isogonal conjugates
as desired. 





















Bulgaria 1998, PEN A17 wrote:
Let
and
be natural numbers such that
![\[A=\dfrac{(m+3)^n+1}{3m}\]](//latex.artofproblemsolving.com/c/c/6/cc6ea4853ef98b5c073d7e292dfbfa0773d19e8d.png)
is an integer. Prove that
is odd.


![\[A=\dfrac{(m+3)^n+1}{3m}\]](http://latex.artofproblemsolving.com/c/c/6/cc6ea4853ef98b5c073d7e292dfbfa0773d19e8d.png)
is an integer. Prove that

This LTE thing is pretty cool
If
is odd, then there is nothing to prove, as
and
are both odd. If
is even, then taking mod
gives
is odd. Therefore, since
we have by LTE
![\[v_2(A)=v_2[(m+3)^n+1]-v_2(3m)=v_2(m+4)+v_2(n)-v_2(m)=v_2(m+4)-v_2(m).\]](//latex.artofproblemsolving.com/9/d/4/9d4b032598b3a414be71e21651bde8fca96e6415.png)
If
then
and
is odd. Otherwise, we need to go a different path.
Note that in the binomial expansion of
, all but the
and
terms are divisible by
. Therefore
. If
is to be even, since
we must have
. If
then by bounding we can easily see that
must be really small and nothing works. Otherwise,
, and
, but this is not possible since
. Therefore
is always odd, as desired. 







![\[v_2(A)=v_2[(m+3)^n+1]-v_2(3m)=v_2(m+4)+v_2(n)-v_2(m)=v_2(m+4)-v_2(m).\]](http://latex.artofproblemsolving.com/9/d/4/9d4b032598b3a414be71e21651bde8fca96e6415.png)
If



Note that in the binomial expansion of















This post has been edited 11 times. Last edited by djmathman, Sep 5, 2013, 2:30 PM