Symmetry options for ADFs

Report any suspected bugs and unexpected behavior here
Post Reply
topefo
Posts: 29
Joined: Fri Nov 02, 2018 12:33 pm
Security question 1: No
Security question 2: 96

Symmetry options for ADFs

Post by topefo » Thu Jan 03, 2019 8:39 pm

Hi!

I am looking into the calculation of discontinuity factors, and using the "set adf" option to
eliminate any stochastic variations between equivalent sides/corners by symmetry.

However, for me the options 3 and 5 seem a bit strange. According to the Serpent wiki we have:

Option Eq. Sides Eq. Corners
3 N-E, S-W SW-NE
5 N-W, S-E NW-SE

However, I would expect the following:

For diagonal mirror symmetry with symmetry line going from the NW corner to the SE corner: N-W, S-E sides and SW-NE corners are equivalent.
For diagonal mirror symmetry with symmetry line going from the NE corner to the SW corner: N-E, S-W sides and NW-SE corners are equivalent.

For me it seems that the eq. corners have been mixed between options 3 and 5, or am I missing something here?

Cordially,
Petri

Ville Valtavirta
Posts: 536
Joined: Fri Sep 07, 2012 1:43 pm
Security question 1: No
Security question 2: 92

Re: Symmetry options for ADFs

Post by Ville Valtavirta » Tue Jan 08, 2019 6:13 pm

Hi Petri,

you are correct.

Checking the actual workings from applygcsymmetries.c shows that:

set adf UNI SURF 3
set adf UNI SURF nesw
set adf UNI SURF swne

indicates a reflective symmetry line going between NE and SW corners. This means that N-E sides are equal as are S-W sides (the second column of the table was correct). Finally, the SE-NW corners are equivalent as you said.

The second case

set adf UNI SURF 3
set adf UNI SURF nwse
set adf UNI SURF senw

indicates a reflective symmetry line between the NW and SE corners yielding equivalence between N-W and S-E and the corners NE-SW.

The wikipage should now be correct regarding both the square and hexagonal surfaces.

-Ville

Post Reply