Courtesy of my friends at A. Mylne & Co (Yacht & Commercial Vessel Design since 1896, part of Ace Marine Ltd, Ivy Cottage, Church Lane, Limekilns, Fife, KY11 3HP, UK; www.mylne.com
) and their 3D modelling package, I have managed to confirm my suspicions concerning the effects of small angular misalignment and applied twist to a blade fitted in the SE-76 jig.
Attached are two jpeg images of the same situation created in the 3D modelling package. These show a 25mm (1") wide 5mm thick chisel being ground at 30 degrees. Also applied is a 2 degree horizontal misalignment (applied with an 80mm centre of rotation from the blade cutting edge - explained below) and simultaneously a 2 degree twist in the vertical plane. These nominal 2 degree error sizes have been used to illustrate the combined effect.
The 80mm was an arbitrary size as the Tormek book says the blade should protrude from the jig by 50 to 75 mm (2 - 3"), and allowing a bit more to get to the axis of the universal jig gives near enough 80mm.
These images very closely illustrate a result I recently experienced where there is a change in the ground edge and two angled faces like an inverted 'v'.
Another result from this particular issue is that one side of the chisel is 2mm shorter than the other, and one corner is 1mm low.
By playing with different small errors and whether they are additive or subtractive (direction of rotation), then various long corners and trapezoidal ground faces can be created, confirming my previously posted views of compound-compound angle effects. I have seen most of the effects in my recent experiences of sharpening!
It is satisfying to prove my suspicions are correct, and it takes only small errors, even 1 degree, to have some noticeable effects.
So all I can advise, is to keep monitoring the grind using felt tip marking, adjusting the blade in the jig and pressure changes etc to counteract any undesirable grinding.
Many thanks to David Gray and Albert Montserrat at Mylne & Co for their assistance.
**I thought 'Attachments and other options' below would allow me to attach the jpegs, but apparently not. Is it possible to submit a picture?