Whoopee Cushion - designed by Bob Keller ©1999
Angles for R.I. = 1.54 151 facets + 33 facets on girdle = 184
3-fold, mirror-image symmetry 120 index
L/W = 1.005 T/W = 0.627 T/L = 0.624 P/W = 0.507 C/W = 0.095 H/W = (P+C)/W+0.02 = 0.622 P/H = 0.816 C/H = 0.152
Vol./W^3 = 0.216 Brightness at 0 degrees tilt for R.I. = 1.54
COS = 89.2 ISO = 93.2

Whoopee Cushion is a meetpoint design for intermediate to advanced faceters. It is faceted using a 120 index and utilizes barion design techniques to combine a round pavilion with its triangular, cushion shaped girdle. Very steep barion facets are required to avoid light leaks around the edges with lower RI materials, and the barion angles used on many designs are often compromised for the sake of mounting practicalities. Whoopee Cushion employs steep barion angles to maximize the light return. It provides small, equilaterally positioned cut corners to facilitate mounting as well as eliminate sharp, damage prone points. An OMNI preform is used to generate the girdle outline.

A design ancestor used a simpler, step cut crown but it seemed no matter what angles I used or how I shifted the step widths around I just couldn't eliminate some dark areas in the corners. Elaborating it with a brilliant pattern illuminated the corners and really gussied up the crown to my eye. Whoopee Cushion is a bright cut which will perform well with material that's more than lightly color saturated. Generally, the darker the material, the brighter the cut needs to be. However, the darker the material, the smaller the stone needs to be. A go-no go test I like is to polish small windows if necessary after sawing out. I then place the rough over a page of a Bible. If I can't read it through the rough, I figure God is telling me I'm not done with the saw yet.

Whoopee Cushion is recommended for intermediate to advanced faceters because it has a relatively large number of facets (184) and is intended for large stones. Larger stones take longer to cut (and polish) and more facets just present more opportunities for making errors. While the pavilion mains are simple to cut, it has 33 of them which must be brought to an accurate centerpoint at the culet as the first step in generating the preform.

The e1 and f1 split main facets on the crown entail small index cheats to meet adjacent facets properly - 1/10 tooth (0.3°) plus and minus. The index cheater on your machine is probably graduated, but it is probably not calibrated. Fractional indexes are typically dialed in on-the-fly using an iterative, incremental, shave-look-adjust "floating in" process. That doesn't entail any rock science, but you need to be familiar with your machine and know which way you have to turn the index cheater to cause a facet to cut deeper on one side or the other.

Shown above is the smoky crystal from which I'm currently cutting this design. If memory serves it was from Madagascar. In spite of a 'cathedral' termination, it was too hammered to be desirable as a specimen. I procured it inexpensively at a show for my rough box. It weighed 64.8 grams and measured about 3x7 cm before I started whacking on it with my trim saw. Above right is the dopped stone after I removed the bulk of the flawed material and very rough 'sawforming'.

I endeavored to make a careful study of the rough and finally settled on an orientation to yield the largest/best stone. It looks like a 20mm or so sized eye-clean stone after I cut away remaining veils, cracks, bubbles and other flaws with further preforming on the lap with a 50-60 Dyna Disk - it's going to be close with a couple of veils near the culet.

The cutting instructions for Whoopee Cushion are also available as a plain text file.

p1 42.00 002-006-010-014-018-022-026-030-034-038-
Cut to culet centerpoint
p2 42.00 120-040-080 Cut to culet centerpoint
g2 90.00 120-040-080 g2 meets p2 - establish size, girdle line
g3 90.00 011-029-051-069-091-109 g3 meets corresponding p1 - preforms girdle outline
g4 90.00 013-027-053-067-093-107 g4 meets corresponding p1 - preforms girdle outline
g5 90.00 015-025-055-065-095-105 g5 meets corresponding p1 - preforms girdle outline
g6 90.00 017-023-057-063-097-103 g6 meets corresponding p1 - preforms girdle outline
g7 90.00 019-021-059-061-099-101 g7 meets corresponding p1 - preforms girdle outline
At left is the pavilion after preforming. I used a 50-60 Dyna Disk to hog off the bulk of the material, followed by 180 and then a well-worn Crystalite 360 channel lap, which is cutting more like 600-1200 than a 360.

Note the three, prepolished p2 pavilion main facets at the 40, 80 and 120 index positions which establish the permanent centerpoint at the culet.

Using alternating patterns that place prepolished facets between still hazy ones helps me keep track of the center of the universe as the remaining 30 mains are prepolished and brought to meet with the culet.

I sometimes hear and read remarks to the effect that quartz is hard to polish. I don't find that to be the case and I suspect those who do may be trying to polish over still too rough a surface. It can take quite a while to polish if you start after a 1200 Crystalite, especially so if it is near new and the facet is large. I'm fond of a 1600 Dyna Disk for prepolishing quartz, which cuts about like a 3000 Crystalite. A cerium polish pretty much snaps in on quartz that's been brought to a fine prepolish.

The b3-b7 barion facets must be cut a slightly different angles for them to meet with both the g3-g7 and corresponding p1 facets. That's more of a requirement for the faceter than the stone. You could save some angle changes and time by cutting b3-b7 all at 85°, meeting with the g3-g7 facets and letting them fall where they may relative to the p1 facets. You won't be able to tell the difference from the crown side.

Shown at right is the polished pavilion, ready to transfer. I cleared that culet crud earlier than I'd guestimated and was able to coax a clean stone just shy of 24mm out of the crystal, measured across the girdle from cut corner to the side opposite. Love it when a plan comes together... The girdle facets are prepolished at this stage but won't be polished until after the transfer. Polishing the girdle after transferring helps fine tune the alignment and establish any cheat needed to level the girdle on the crown side.

I always get a feeling of accomplishment when I get to the transfer even though you're only "half" finished at that point.


b3 85.00 011-029-051-069-091-109 b3 meets corresponding p1, g3
b4 85.23 013-027-053-067-093-107 b4 meets corresponding p1, g4
b5 85.39 015-025-055-065-095-105 b5 meets corresponding p1, g5
b6 85.49 017-023-057-063-097-103 b6 meets corresponding p1, g6
b7 85.54 019-021-059-061-099-101 b7 meets corresponding p1, g7
c2 33.50 120-040-080 c2 meets g2 - creates upper girdle line
c3 34.00 011-029-051-069-091-109 c3 meets g3 - creates upper girdle line
c4 34.00 013-027-053-067-093-107 c4 meets g4 - creates upper girdle line
c5 34.00 015-025-055-065-095-105 c5 meets g5 - creates upper girdle line
c6 34.00 017-023-057-063-097-103 c6 meets g6 - creates upper girdle line
c7 34.00 019-021-059-061-099-101 c7 meets g7 - creates upper girdle line
d3 15.00 011-029-051-069-091-109 d3 meets c2,c3 - creates upper tier
d4 15.00 013-027-053-067-093-107 d4 meets c4 - creates upper tier
d5 15.00 015-025-055-065-095-105 d5 meets c5 - creates upper tier
d6 15.00 017-019-021-023-057-059-061-063-097-099-101-103 d6 meets c6 and c7 - creates upper tier
e1 32.81 013.9-053.9-093.9 - 026.1-066.1-106.1 e1 meets c4,c5,others - note 0.1 tooth (0.3°) index cheats
e2 32.96 018-022-058-062-098-102 e2 meets c6,c7,others
Cutting the table prior to applying the upper course of split mains requires you to meet with only three points. If you cut the upper course of split mains first, you will have to meet the table with 15 points simultaneously, which is unlikely to happen. I figure if I'm going to have to go back and touch up the upper course of mains after I cut the table anyway, it's easier to just cut them last. Most faceters and machines are a little short of perfect. Since the f1 and f2 facets are triangular, they're easy to 'float in' as may be required to meet any other points lacking in geometrical perfection due to accumulated errors. Use them as "error decumulators." ;)
t 0.00 Table table meets d3,d4
f1 16.00 013.9-053.9-093.9 - 026.1-066.1-106.1 f1 meets d4,d5,others - note 0.1 tooth (0.3°) index cheats
f2 15.85 018-022-058-062-098-102 f2 meets d6,t,others

If you cut a Whoopee Cushion of your own, be sure and let me know what you've cut and how you made out.

Index of Gem Designs

Perfect Transfer Index

Table of Contents

© Bob's Rock Shop Bob Keller