Difference between revisions of "Van Aubel's Theorem"

(Proof 1: Complex Numbers)
(Proof 1: Complex Numbers)
 
(2 intermediate revisions by the same user not shown)
Line 6: Line 6:
  
 
== Proof 1: Complex Numbers==
 
== Proof 1: Complex Numbers==
[asy]
+
<asy>
 
size(220);
 
size(220);
 
import TrigMacros;
 
import TrigMacros;
 
rr_cartesian_axes(-3,8,-2,8,complexplane=true,usegrid = false);
 
rr_cartesian_axes(-3,8,-2,8,complexplane=true,usegrid = false);
pair U, V, WW, Z, O, P, Q, R, SS;
+
pair A, B, C, D, O, P, Q, R, SS;
 
O = (0,0) ;
 
O = (0,0) ;
 
A = (2,1.5);
 
A = (2,1.5);
Line 21: Line 21:
 
draw(B--(B + rotate(-90)*(C-B))--(C + rotate(90)*(B-C))--C);
 
draw(B--(B + rotate(-90)*(C-B))--(C + rotate(90)*(B-C))--C);
 
draw(C--(C + rotate(-90)*(D-C))--(D + rotate(90)*(C-D))--D);
 
draw(C--(C + rotate(-90)*(D-C))--(D + rotate(90)*(C-D))--D);
draw(D--(D + rotate(-90)*(A-D))--(A + rotate(90)*(D-A))--D);
+
draw(D--(D + rotate(-90)*(A-D))--(A + rotate(90)*(D-A))--A);
  
 
P = (B + (A + rotate(-90)*(B-A)))/2;
 
P = (B + (A + rotate(-90)*(B-A)))/2;
Line 30: Line 30:
 
//draw(WW--Y,red);
 
//draw(WW--Y,red);
 
//draw(X--Z,blue);
 
//draw(X--Z,blue);
dot("<math>a</math>",U,SW);
+
dot("$a$",A,SW);
dot("<math>b</math>",V,2*E);
+
dot("$b$",B,2*E);
dot("<math>c</math>",WW,E);
+
dot("$c$",C,E);
dot("<math>d</math>",Z,NNW);
+
dot("$d$",D,NNW);
  
dot("<math>p</math>",P,E);
+
dot("$p$",P,E);
dot("<math>q</math>",Q,S);
+
dot("$q$",Q,S);
dot("<math>r</math>",R,N);
+
dot("$r$",R,N);
dot("<math>s</math>",SS,S);
+
dot("$s$",SS,S);
[/asy]
+
</asy>
 
Putting the diagram on the complex plane, let any point <math>X</math> be represented by the complex number <math>x</math>.  Note that <math>\angle PAB = \frac{\pi}{4}</math> and that <math>PA = \frac{\sqrt{2}}{2}AB</math>, and similarly for the other sides of the quadrilateral.  Then we have
 
Putting the diagram on the complex plane, let any point <math>X</math> be represented by the complex number <math>x</math>.  Note that <math>\angle PAB = \frac{\pi}{4}</math> and that <math>PA = \frac{\sqrt{2}}{2}AB</math>, and similarly for the other sides of the quadrilateral.  Then we have
  
Line 61: Line 61:
 
\end{eqnarray*}</cmath>
 
\end{eqnarray*}</cmath>
  
Finally, we have <math>(p-r) = i(q-s) = e^{i \pi/2}(q-r)</math>, which implies <math>PR = QS</math> and <math>PR \perp QS</math>, as desired.
+
Finally, we have <math>(p-r) = i(q-s) = e^{i \pi/2}(q-s)</math>, which implies <math>PR = QS</math> and <math>PR \perp QS</math>, as desired.
  
 
==See Also==
 
==See Also==
 
[[Category:Theorems]]
 
[[Category:Theorems]]

Latest revision as of 13:01, 4 March 2023

Theorem

On each side of quadrilateral $ABCD$, construct an external square and its center: $ABA'B'$, $BCB'C'$, $CDC'D'$, $DAD'A'$; yielding centers $P_{AB}, P_{BC}, P_{CD}, P_{DA}$. Van Aubel's Theorem states that the two line segments connecting opposite centers are perpendicular and equal length: $P_{AB}P_{CD} = P_{BC}P_{DA}$, and $\overline{P_{AB}P_{CD}} \perp \overline{P_{BC}P_{DA}}$.

Proofs

Proof 1: Complex Numbers

[asy] size(220); import TrigMacros; rr_cartesian_axes(-3,8,-2,8,complexplane=true,usegrid = false); pair A, B, C, D, O, P, Q, R, SS; O = (0,0) ; A = (2,1.5); B= (4,1.8); C = (5.3,3); D= (3,5.3);  draw(A--B--C--D--cycle); draw(A--(A + rotate(-90)*(B-A))--(B + rotate(90)*(A-B))--B); draw(B--(B + rotate(-90)*(C-B))--(C + rotate(90)*(B-C))--C); draw(C--(C + rotate(-90)*(D-C))--(D + rotate(90)*(C-D))--D); draw(D--(D + rotate(-90)*(A-D))--(A + rotate(90)*(D-A))--A);  P = (B + (A + rotate(-90)*(B-A)))/2; Q = (C + (B + rotate(-90)*(C-B)))/2; R = (D + (C + rotate(-90)*(D-C)))/2; SS = (A + (D + rotate(-90)*(A-D)))/2;  //draw(WW--Y,red); //draw(X--Z,blue); dot("$a$",A,SW); dot("$b$",B,2*E); dot("$c$",C,E); dot("$d$",D,NNW);  dot("$p$",P,E); dot("$q$",Q,S); dot("$r$",R,N); dot("$s$",SS,S); [/asy] Putting the diagram on the complex plane, let any point $X$ be represented by the complex number $x$. Note that $\angle PAB = \frac{\pi}{4}$ and that $PA = \frac{\sqrt{2}}{2}AB$, and similarly for the other sides of the quadrilateral. Then we have


\begin{eqnarray*}  p &=& \frac{\sqrt{2}}{2}(b-a)e^{i \frac{\pi}{4}}+a \\ q &=& \frac{\sqrt{2}}{2}(c-b)e^{i \frac{\pi}{4}}+b \\ r &=& \frac{\sqrt{2}}{2}(d-c)e^{i \frac{\pi}{4}}+c \\ s &=& \frac{\sqrt{2}}{2}(a-d)e^{i \frac{\pi}{4}}+d \end{eqnarray*}

From this, we find that \begin{eqnarray*} p-r &=& \frac{\sqrt{2}}{2}(b-a)e^{i \frac{\pi}{4}}+a - \frac{\sqrt{2}}{2}(d-c)e^{i \frac{\pi}{4}}-c \\ &=& \frac{1+i}{2}(b-d) + \frac{1-i}{2}(a-c). \end{eqnarray*} Similarly, \begin{eqnarray*} q-s &=& \frac{\sqrt{2}}{2}(c-b)e^{i \frac{\pi}{4}}+a - \frac{\sqrt{2}}{2}(a-d)e^{i \frac{\pi}{4}}-c \\ &=& \frac{1+i}{2}(c-a) + \frac{1-i}{2}(b-d). \end{eqnarray*}

Finally, we have $(p-r) = i(q-s) = e^{i \pi/2}(q-s)$, which implies $PR = QS$ and $PR \perp QS$, as desired.

See Also