Weitzenböck's inequality should not be confused with Weitzenböck identity.
In mathematics, Weitzenböck's inequality, named after Roland Weitzenböck, states that for a triangle of side lengths
a
b
c
\Delta
a2+b2+c2\geq4\sqrt{3}\Delta.
Equality occurs if and only if the triangle is equilateral. Pedoe's inequality is a generalization of Weitzenböck's inequality. The Hadwiger–Finsler inequality is a strengthened version of Weitzenböck's inequality.
Rewriting the inequality above allows for a more concrete geometric interpretation, which in turn provides an immediate proof.[1]
\sqrt{3 | |
Now the summands on the left side are the areas of equilateral triangles erected over the sides of the original triangle and hence the inequation states that the sum of areas of the equilateral triangles is always greater than or equal to threefold the area of the original triangle.
\Deltaa+\Deltab+\Deltac\geq3\Delta.
This can now be shown by replicating area of the triangle three times within the equilateral triangles. To achieve that the Fermat point is used to partition the triangle into three obtuse subtriangles with a
120\circ
120\circ
120\circ
The proof of this inequality was set as a question in the International Mathematical Olympiad of 1961. Even so, the result is not too difficult to derive using Heron's formula for the area of a triangle:
\begin{align} \Delta&{}=
1 | |
4 |
\sqrt{(a+b+c)(a+b-c)(b+c-a)(c+a-b)}\\[4pt] &{}=
1 | |
4 |
\sqrt{2(a2b2+a2c2+b2c2)-(a4+b4+c4)}. \end{align}
It can be shown that the area of the inner Napoleon's triangle, which must be nonnegative, is[2]
\sqrt{3 | |
so the expression in parentheses must be greater than or equal to 0.
This method assumes no knowledge of inequalities except that all squares are nonnegative.
\begin{align} {}&(a2-b2)2+(b2-c2)2+(c2-a2)2\geq0\\[5pt] {}\iff&2(a4+b4+c4)-2(a2b2+a2c2+b2c2)\geq0\\[5pt] {}\iff&
4(a4+b4+c4) | |
3 |
\geq
4(a2b2+a2c2+b2c2) | |
3 |
\\[5pt] {}\iff&
(a4+b4+c4)+2(a2b2+a2c2+b2c2) | |
3 |
\geq2(a2b2+a2c2+b2c2)-(a4+b4+c4)\\[5pt] {}\iff&
(a2+b2+c2)2 | |
3 |
\geq(4\Delta)2, \end{align}
and the result follows immediately by taking the positive square root of both sides. From the first inequality we can also see that equality occurs only when
a=b=c
This proof assumes knowledge of the AM–GM inequality.
\begin{align} &&(a-b)2+(b-c)2+(c-a)2&\geq&&0\\ \iff&&2a2+2b2+2c2&\geq&&2ab+2bc+2ac\\ \iff&&3(a2+b2+c2)&\geq&&(a+b+c)2\\ \iff&&a2+b2+c2&\geq&&\sqrt{3(a+b+c)\left(
a+b+c | |
3 |
\right)3}\\ ⇒ &&a2+b2+c2&\geq&&\sqrt{3(a+b+c)(-a+b+c)(a-b+c)(a+b-c)}\\ \iff&&a2+b2+c2&\geq&&4\sqrt3\Delta. \end{align}
As we have used the arithmetic-geometric mean inequality, equality only occurs when
a=b=c
Write
x=\cotA,c=\cotA+\cotB>0
S=\cotA+\cotB+\cotC=c+
1-x(c-x) | |
c |
cS=c2-xc+x
| ||||
| ||||
\right) |
S\ge\sqrt{3}
\cotA=
b2+c2-a2 | |
4\Delta |
S= | a2+b2+c2 |
4\Delta |
On Some Geometrical Inequalities. The Mathematical Gazette, Vol. 26, No. 272 (Dec., 1942), pp. 202-208 (JSTOR)
Über eine Ungleichung in der Dreiecksgeometrie. Mathematische Zeitschrift, Volume 5, 1919, pp. 137-146 (online copy at Göttinger Digitalisierungszentrum)