The real reason for printing the knot 8_18 yesterday was to compare it to a print of 8_19, the first non-alternating knot in the standard knot table.
STL file: http://www.geekhaus.com/makerhome/day67_knot819.stl
Tinkercad link: https://tinkercad.com/things/icZH9zTcivd-day-67-knot-819/edit
Thingiverse link: http://www.thingiverse.com/thing:175023
Settings: The same MakerWare custom knot-slicing profile used yesterday.
Technical notes: As always, this knot was made in Mathematica using Cantarella data. Compare this model of 8_19 with the 3D model of 8_18. Can we see something in this optimal 3D configuration of 8_19 that would suggest it has no alternating 2D projections?