Cube Nets on a Grid

While attempting the latest Internet Puzzle Solvers Test, I found I needed to create a cube.  I have graph paper with 1-inch squares, measuring 9 by 7 inches.  What is the maximum number of unit cubes I could create from a single page, by cutting out 6-square Cube Nets?  Extension: What is the maximum number of unit cubes which could be made from an MxN sheet (M<=N<=10)?

There are 11 different Cube Nets which might be used in building a cube.  I found a listing at the Cube page at MathWorld:
http://mathworld.wolfram.com/Cube.html

Source: Original.


Solutions were received from Kirk Bresniker and Joseph DeVincentis.

Joseph sent the following configurations:

I found a way to make 9 in a 9x7 sheet. The theoretical maximum of 10 is probably impossible due to the apparently forced 2 wasted squares in two corners.

DDEEFFFXX
HDDEEEFFF
HHDDEIGGX
XHXIIIIGX
XHHIBAAGG
CCCBBBAAG
XXCCCBBAA

For other sizes, in most cases you can fit N cubes in any rectangle of area 6N+4 or more.
1xN does not fit any cube nets, and likewise for 2x2, 2x3, 2x4, and 3x3.
2xN fits one in 2x5, and one more for each additional three in length, as with 2x8:

AAABBBXX
XXAAABBB

3xN fits one in 3x4 (several ways), and one more for each additional two in length, as with 3x10:

AABBCCDDXX
XAABBCCDDX
XXAABBCCDD

4x4 fits two cubes:

XAAB
XABB
AABX
ABBX

4x6 fits three, and 4x7 fits four by repeating the 4x4 pattern. 4x9 fits five, and 4x10 fits six.

5x5 fits three (without D, below) and 5x6 fits four:

ABBBDX
AXBCDX
AABCDD
XABCXD
XACCCD

The pattern extends by repeating the inner portion to allow 5x8 to fit five cubes, and 5x9 to fit six.

6x7 fits six:

XAAXDDD
XADDDXE
AABEEEE
ABBCCEF
BBCCFFF
BCCFFXX

Seven cubes in a 5x10 (a different extension of the other 5xN solutions):

ABBBDEEEGX
AXBCDXEFGX
AABCDDEFGG
XABCXDEFXG
XACCCDFFFG

Seven cubes in 6x8 (an extension of a different four in 6x5 solution, and extendable in 3s by repeating the B-F-G group):

AAABBBXX
XCAAABBB
XCDDEFFG
CCDEEFGG
CDDEFFGX
CDEEFGGX

Eight cubes in 6x9:

XAACCXEEF
XACCEEEFF
AABCEGFFH
ABBCDGFHH
BBDDDGGHX
BDDXGGHHX

Nine cubes in a 6x10, also extendable in 3s:

CCAAABBBXX
DCCCAAABBB
DHHCIIEFFG
DDHHIEEFGG
XDHIIEFFGX
XDHIEEFGGX

Seven in a 7x7:

AADDDXX
BAAADDD
BBAFFXX
CBBGFFF
CCBGGGF
XCEEEGG
XCCXEEE

Eight in a 7x8:

AADDEEXX
BAADDEEX
BBAADDEE
CBBFFGGH
CCBFXGHH
XCFFGGHX
XCCFGHHX

Ten in a 7x10 (the rule possibly allows a very tight eleven):

AADDEEIIXX
BAADDEEIII
BBAADDEEHI
CBBFGGHHHH
CCBFXGGGJH
XCFFXXGJJJ
XCCFFXJJXX

Nine in an 8x8 (the rule possibly allows a very tight ten):

AADDEEXX
BAADDEEE
BBAADDFE
CBBGFFFF
CBXGGHHF
CCIIGGHH
XCXIIGHX
XCXXIIHX

Ten in an 8x9 (eleven may be possible):

ADDDEEEXX
AAADDDEEE
BBAAFFIGG
CBFFFIIGH
CBBFIIGGH
CCBJIXGHH
XCJJJXXHX
XCXXJJXHX

Twelve fit in an 8x10 by doing two 4x10 solutions one above the other.

Twelve in a 9x9:

ABBJJJXCC
AABXJCCCD
XABBJCDDD
IAABJDDXK
IIIIXKKKK
IXHHLFEEK
HHHGLFFEX
HGGGLXFEE
GGXLLLFFE

Thirteen in a 9x10 (didn't find a fourteen):


AAXBBXXXMX
DAAABBXXMX
DDDACBBLMM
EEDDCLLLLM
FEEECCKKLM
FFFEHCJKKK
GXFFHCJJJK 
GGGHHIIIJJ 
XXGGHHXIII 

Fifteen in a 10x10 can be done by placing a nine-in-6x10 solution above a six-in-4x10 solution.
The rule allows a very tight sixteen.

Probably unsolvable:  five in a 5x7, six in a 5x8, five in a 6x6.

Unsolved: eleven in a 7x10, ten in an 8x8, eleven in an 8x9, fourteen in a 9x10, sixteen in a 10x10.

Kirk wrote a computer program to find solutions using any of the various nets.  The program took an excessively long time for larger grids, and found he could achieve the same results with only four of the nets, so hypothesizes that only those 4 are needed for the best solutions

I'll assert that you only need the following four:

  XX
 XX
XX

   X
XXXX
   X

XXX
  XXX

XX
 XXX
   X

Using these four I've found the following.  Up to 6x9 this matches the exhaustive searches.

2 x 10 max 3 best 2
2 x 2 max 0 no solution
2 x 3 max 1 no solution
2 x 4 max 1 no solution
2 x 5 max 1 best 1 optimal
2 x 6 max 2 best 1
2 x 7 max 2 best 1
2 x 8 max 2 best 2 optimal
2 x 9 max 3 best 2
3 x 10 max 5 best 4
3 x 3 max 1 no solution
3 x 4 max 2 best 1
3 x 5 max 2 best 1
3 x 6 max 3 best 2
3 x 7 max 3 best 2
3 x 8 max 4 best 3
3 x 9 max 4 best 3
4 x 10 max 6 best 6 optimal
4 x 4 max 2 best 2 optimal
4 x 5 max 3 best 2
4 x 6 max 4 best 3
4 x 7 max 4 best 4 optimal
4 x 8 max 5 best 4
4 x 9 max 6 best 5
5 x 10 max 8 best 7
5 x 5 max 4 best 3
5 x 6 max 5 best 4
5 x 7 max 5 best 4
5 x 8 max 6 best 5
5 x 9 max 7 best 6
6 x 10 max 10 best 8
6 x 6 max 6 best 4
6 x 7 max 7 best 6
6 x 8 max 8 best 7
6 x 9 max 9 best 7
7 x 10 max 11 best 10
7 x 7 max 8 best 6
7 x 8 max 9 best 8
7 x 9 max 10 best 9
8 x 8 max 10 best 9

Mail to Ken