All the methods you've been discussing to find the declination of a wall
assume that the INCLINATION
of the wall is exactly 90 deg, which is rarely the case... I measured
the inclinations of some buildings
in my neigbourhood and saw that most of them had inclinations varying
from 89 deg to 91 deg,
seldom 90 deg and, from time to time, more. An architect friend of mine
explained to me the
reasons for not giving that exact value but, in any case, there is no
point in imposing an accuracy of
+/- 0.5 deg in declination if we have +/- 1 deg in inclination.
According to D. Savoie ("La gnomonique" p 321) in medium latitudes we
can expect an error up to 4 min (!) for every degree of deviation. An
approximate formula is:
Err(H) = Err(D)* [ cos(Lat)*tan(SunDecl)*cos(H) - sin(Lat) ]
When I want to deal with exact values for declination and inclination
I use the routines at Yvon Masse's web or the program SundialAlign! but
probably you'll need to make a wedge to accurately place the dial.
Anselmo P. Serrada
41.63 N 4.73 W
-
