[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference rusure::math

Title:Mathematics at DEC
Moderator:RUSURE::EDP
Created:Mon Feb 03 1986
Last Modified:Fri Jun 06 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:2083
Total number of notes:14613

1264.0. "Integer pentagons" by CIVAGE::LYNN (Lynn Yarbrough @WNP DTN 427-5663) Mon Jul 02 1990 18:34

Here's a couplet of problems, the first quite easy, the second quite hard:

1) Find a regular pentagon for which each vertex has at least one integer 
coordinate.

2) Find a regular pentagon for which each vertex has an integer y-
coordinate.

At present I don't know whether the second has a solution. I haven't found 
one yet.

Lynn Yarbrough 
T.RTitleUserPersonal
Name
DateLines
1264.1TRACE::GILBERTOwnership ObligatesTue Jul 03 1990 20:5965
First, I'll shift the problem slightly, to an equivalent one.

	Find a regular pentagon in the complex plane all of whose vertices
	have rational real parts.  Or show that no such pentagon can exist.

It can't exist.  Proof follows:

	Let the vertices be p_0, p_1, p_2, p_3, and p_4, (the underscore is
	used to indicate subscripts), in order counterclockwise.  Without loss
	of generality, let p_0 = 0.  Let p_1 = V = V_r + i V_i.  And also let
	W = cos(2pi/5) + i sin(2pi/5), so that multiplying by W represents a
	2pi/5 rotation counterclockwise (2pi/5 = 72�).
	let p_1 = V = V_r + i V_i.  Then:

		p_0 = 0
		p_1 = V
		p_2 = V + V*W
		p_3 = V + V*W + V*W�
		p_4 = V + V*W + V*W� + V*W�

	Now, we expand all of the above.  In so doing, the following table
	is useful:

		k	sin(k pi/5)		cos(k pi/5)
		-	-----------		-----------
		0	0			 1
		1	sqrt((5-sqrt(5))/8)	 (sqrt(5)+1)/4
		2	sqrt((5+sqrt(5))/8)	 (sqrt(5)-1)/4
		3	sqrt((5+sqrt(5))/8)	-(sqrt(5)-1)/4
		4	sqrt((5-sqrt(5))/8)	-(sqrt(5)+1)/4
		5	0			-1

	We get:
		p_0 = 0
		p_1 = V_r + i V_i

		p_2 = {V_r (3+sqrt(5))/4 - V_i sqrt((5+sqrt(5))/8)}
		  + i {V_i (3+sqrt(5))/4 + V_r sqrt((5+sqrt(5))/8)}

		p_3 = {V_r/2 - V_i sqrt(5+2 sqrt(5))/2}
		  + i {V_i/2 + V_r sqrt(5+2 sqrt(5))/2}

		p_4 = {V_r (1-sqrt(5))/4 - V_i sqrt((5+sqrt(5))/8)}
		  + i {V_i (1-sqrt(5))/4 - V_r sqrt((5+sqrt(5))/8))

	And so we take just the real parts of these, to see whether they
	might all be rational.  Let R_i = Re(p_i), the real part of each p_i;
	we also want each of these R_i to be rational.
	
		R_1 = V_r

		R_3 = R_1/2 - V_i sqrt(5+2 sqrt(5))/2, so
		V_i = (R_1 - 2 R_3) sqrt((5-2 sqrt(5))/5)

		R_2 = R_1 (3+sqrt(5))/4 - (R_1 - 2 R_3) sqrt((3-sqrt(5))/8)
		    = R_1 (3+sqrt(5))/4 - (R_1 - 2 R_3) (sqrt(5)-1)/4, so

		R_2 = R_1 + R_3 (sqrt(5)-1)/2

		sqrt(5) = 2 (R_2 - R_1)/R_3 + 1

	Thus, if R_1, R_2, and R_3 are rational, we've discovered that sqrt(5)
	is really a rational number!  The alternative is that no four vertices
	(we never got to R_4) of a regular pentagon in the plane may have
	rational x coordinates.