ATM [Fwd: BOUNCE atm@shore.net: 3 4 spider vanes

[Mel Bartels <mbartels@efn.org>]

Date: Mon, 27 Dec 1999 13:24:21 -0500
From: "J.Everett Cairns" <ecairns@mcmaster.ca>
Organization: McMaster University

Christopher J.R. Lord wrote:
> 
> > Bryan Greer wrote:
> 
> >
> It is a myth that wire thin vanes produce fainter, longer spikes. The
> spike's brightness is dependent on the source intensity and the aperture.
> Whereas the obstruction causes the diffraction it is easier to regard the
> diffraction occuring at the edge of the vane.

        I know a few Physics profs who would cringe at this last statement.They
have a thing about suggesting that diffraction is an edge effect only.
It is like 
saying that the rust on your car is due only to what it was exposed to
last Thursday.
The reality is that the rust is the result of your car's total life
history. In diffraction
, the pattern seen is the result of ALL the light from the aperture
added
together at the observer.
If we talk about the SAME scope and ONE star I would have expected the
following:
As I think I  mentioned earlier, the diffraction pattern from an
obstruction in 
your scope's path is the same as that of an open aperture the same shape
as the
obstruction, with the addition of the unobstructed star image.The latter
is just
 the diffraction image of the full scope unobstructed aperture. Also,it
then 
follows that the intensity in the spider diffraction will be directly
proportional
 to the edge-on area of the vane. Most of us are aware that thin slits
produce 
wide diffraction patterns. From these considerations, which I have
checked out 
more than once under controlled conditions in the lab, I would have
expected that 
thinner vanes produce fainter, longer spikes. HOWEVER as they are
fainter, they 
may not appear longer! By longer in the body of this paragraph, I meant
longer
to the first monochromatic diffraction minimum, not to the edge of
visual 
threshold. Intensity falls ,in a linear diffraction pattern of a slit
object
 roughly as the distance (measured in "to the first diffraction minimum"
units 
for the slit) squared. So the distance to threshold limit would go as
1/vane
 thickness squared were it not for the fact that the overall intensity
will also 
go up directly as the vane thickness. Combining the two effects we find
that the 
distance to threshold goes crudely as 1/vane thickness. I am puzzled as
this still
 means wide vane--short spike.What Have I overlooked? Just as long as
the intensity 
falls faster than linear with off axis distance,it would seem that the
arguement holds.

 Suiter's and Zmek's
> diffraction modelling  are in monochromatic light and are unrealistic.
> Modelling in monochromatic light is useful in portraying the ideal Airy
> disc, and the efffects of obstruction, turned edge and Seidel errors, but
> unhelpful and misleading when modelling spider diffraction, particularly
> spider diffraction effected by turbulence.

        Monochromatic modelling should also be useful to determine what happens 
in white light. We just have to remember that the various colours will
show patterns 
scaled in size proportional to their wavelength. This will give the
rainbow effect
seen on a very bright spike from a bright star.




 T
> It has been my long expereince (over 30 years), and that of fellow ATM's of
> considerable experience in making and using classically designed Newtonian
> and Cassegrain reflectors ( such as R. N. Irving, J.D. Greenwood, and Phil
> Horrocks) that spiders with arms that intersect at the support spindle
> produce noticeably brighter and longer spikes than arms that meet the
> secondary holder tangentially, and are offset, as in the Hargreaves design.

        This perception may be complicated by a tendency to change several 
variables at once. As you point out,tangentially rather than radially
arranged 
vanes have much better torsional stiffness. On the average then it
appears that
radial vanes may be made heavier (wider) to compensate.This would
produce longer or 
shorter spikes depending on how width affects the outcome.


                                                Everett