Effects of primary

primary mirror
mirror segmentation
segmentation on
on telescope
telescope image quality

Terry S.
Terry S. Mast, Jerry E.E. Nelson
Lawrence Berkeley
Berkeley Laboratory, Space Sciences Laboratory,
Laboratory, University of California, Berkeley, 94720
Berkeley, California 94720
William J. Welch
J.Welch
Radio Astronomy
Radio Astronomy Laboratory,
Laboratory, University
University of
of California,
California, Berkeley, California94720
Berkeley,California 94720

Abstract
A number
number of designs
designs for future
future astronomical telescopes
astronomical telescopes A large
A number of effects
large number effects contribute
contribute to the the final image
final image
call for
call for large primary mirrors
large primary mosaics of
are mosaics
mirrors that are of smaller mir-
smaller mir- in a segmented
segmented mirror telescope. These
mirror telescope. include the
These include geometry
the geometry
rors. We
rors. We describe
describe herehere aa study
study ofofsome characteristics of the
some characteristics of the segmentation,
segmentation, the the precision
precision of of the phasing con-
and phasing
tilt and
the tilt con-
images expected
images expected from a telescope
telescope with primary mirror
with aa primary com--
mirrorcorn trol
trol of the
the segments,
segments, the the quality
quality of of the
the optical surfaces, the
optical surfaces,
posed of 36
posed 36 hexagonal
hexagonal segments.
segments. Various
Various effects caused by
effects caused by alignment
alignment of the optics,optics, the effectseffects of atmospheric seeing,
of atmospheric seeing, and and
the segmentation
the segmentation and imperfections
imperfections in the the segment
segment fabrica-
fabrica- the characteristics
characteristics of of the
the detector.
detector. The The time time dependence
dependence of of
tion and control have been
tion analyzed using
been analyzed optics. The
physical optics.
using physical these effects alsoalso becomes
becomes important for some some types types of obser-
of obser-
diffraction -limitedimage
diffraction-limited imagedistribution segmentation
from the segmentation
distributionfrom vations,
vations, for example,
example, speckle interferometry. In
speckle interferometry. In this
this paper
paper
geometry of
geometry of the primary is derived, and and the spikes
diffraction spikes
the diffraction we
we discuss
discuss qualitatively
qualitatively and, and, where
where possible,
possible, quantitatively
quantitatively
are shown
are shown to be similar caused by
those caused
similar to those by secondary support
secondary support some
some of these effects.
effects. We address those
We address those whichwhich are useful useful for
struts in existing
struts existing telescopes.
telescopes. A A general
general relation
relation between
between sur- setting
setting the parameters
parameters of of the
the segmented
segmented mirror fabrication
mirror fabrication
face quality
face quality and
and image
image quality given, and the
quality isis given, implications
the implications control.
and control.
for surface
for surface quality
quality tolerances
tolerances are discussed. The optical
are discussed. optical The geometry
The geometry of segmentation
segmentation will will cause diffraction
cause aa diffraction
effects of
effects of segment phasing errors
segment phasing are derived,
errors are derived, showing that at
showing that pattern
pattern different
different fromfrom that that of a continuous
continuous mirror. mirror. The The sizesize
least for
least visible light
for visible observations the phasing
light observations unimportant.
phasing is unimportant. and shape
shape of the segments
segments and the size of
the size of the cracks between
the cracks
For observations
For observations at 10 10 µm
fimnear
neardiffraction -limited perfor-
diffraction-limited perfor- them define
them define the scalescale and structure of of the pattern.
diffraction pattern.
the diffraction
mance can
mance can bebe achieved
achieved with with aa 1010 meter
meter aperture
aperture requiring
requiring A common initial
A concern with
initial concern segmented design
with aa segmented design isis that the
segments be
that the segments phased correctly
be phased at this
correctly at wavelength.
this wavelength. cracks between the
cracks between the segments
segments might might give give rise rise to veryvery large,
large,
Contents bright
bright diffraction spikes on
diffraction spikes on stellar
stellar images.
images. We describe in
We describe in
section 22 the
section the diffraction
diffractionpatternpattern fromfrom the primary mirror.
the primary mirror. Our
foremost conclusion
conclusion is that the the effects qualitatively like
will be qualitatively
effects will like
1. Introduction spiders in
support spiders in existing telescopes,
existing telescopes,
2.
2. Segmented
Segmented Aperture Image
Diffraction Image
ApertureDiffraction those caused
those caused by by mirror support
Surface Roughness and quantitatively
and quantitatively set set by fractional area
by the fractional area of of the cracks. If
the cracks.
3. Surface
3.
area
4. Segment
4. Segment Phasing
Phasing the cracks
the cracks between
between the the segments
segments occupy occupy aa fractional
fractional area
equal to
equal to that of spiders, the
supportspiders,
the support
of the diffractioneffects
thediffraction effects willwill
5.
5. Atmospheric
Atmospheric Seeing
Seeing
6. Speckle
6. Speckle Interferometry be essentially
be equivalent, and
essentially equivalent, and thus the the segment
segment diffraction
diffraction
effects will
effects willbebe nono more
more troublesome
troublesome than than those on existing
on existing
telescopes.
1.
1. Introduction
Introduction In practice,
In practice, for for visible
visible light, large telescopes
light, large telescopes do do not
achieve the angular resolution corresponding
achieve corresponding to diffraction
the diffraction
to the
Scaling lawsfor
Scaling laws for the
the materials
materials and the of large
costs of
the costs opti-
large opti- patterns of their primary
patterns mirrors. The
primary mirrors. The resolution limited by
resolution isis limited by
cal
cal telescopes clearly show
telescopes clearly show that increasing
increasing the area
collecting area
the collecting the random
the random errors
errors in the the primary
primary surface
surface and, and, more impor-
more impor-
beyond
beyond that of of existing telescopes will
existing telescopes will be without
impossible without
be impossible tantly, by
tantly, by the random atmospheric
atmospheric phase phase errors that the rays
the rays
major
major innovations
innovations in in the design. In the
the design. the last
last decade
decade several
several of light suffer before
light suffer before entering
entering the telescope.
telescope. The segmented
The segmented
different ways of achieving
different ways achieving a large collecting area
large collecting been
area have been primary with
primary with unphased
unphased elements elements has has in in addition
addition aa specialspecial
explored
explored by by many
many groups.
groups. These
These include
include thethe use of an an array
array form of random
form random surface error. In section
surface error. section 33 we we discuss
discuss the the
of individual telescopes,aa single
individual telescopes, single large
large thin
thin mirror,
mirror, and
and a effects of
effects of surface
surface roughness.
roughness. Three sources sources of roughnessroughness are
mosaic of mirror segments.
mosaic considered; the
considered; the fine
fine scale
scale polished
polished surface,
surface, the spatial
high spatial
the high
This
This paper is part of aa larger
paper is larger design study of a new opti-
new opti- frequency polishing
frequency polishingaberrations
aberrationsinin each each segment,
segment, and and the the
cal
cal telescope proposed for
telescope proposed for the University
University of of California (Nel-
California (Nel- lower spatial
lower spatial frequency
frequency segment segment aberrations
aberrations caused caused by seg-
by seg-
son,1980).
son, 1980). The mirrormirror is is aa mosaic
mosaic of of 36 36 hexagonal
hexagonal mirror
mirror ment misalignment or tilt. tilt.
segments
segments and and the geometry
geometry of the segmentssegments is shown in Fig-
is shown Fig- Another concern
Another concern has been been the the possible
possible adverse effects,
adverse effects,
ure 1.1. The
The area
area of
of the
the primary
primary mirror
mirror is equivalent to
is equivalent to that of particularlyfor
particularly for photometry,
photometry, of partially partially phasedphased or unphased
or unphased
a 10
10 meter
meter diameter
diameter disk.
disk. Each segment isis 0.9
Each segment 0.9 meters
meters on on a segments. Initially
Initially itit was was thought
thought that the the lacklack of ofperfect
perfect
side
side andand the spacing between segments
spacing between segmentsisisabout about 11 cm.
cm. An An phasing might
phasing might noticeably
noticeably alter alter thethe flux
flux at at relatively
relatively large large
active system of electronic
active system sensors and
electronic sensors and motor driven actuators angles. The
angles. The effects
effects of of phasing
phasing errors
errors are explored in
are explored Section
in Section
will maintain the
will maintain the relative
relative orientations against
segments against
orientations of the segments 4 and they are shown to be small for
be small visible light.
for visible
perturbations from
perturbations from wind,
wind, thermal
thermal changes,
changes, and and changes
changes in in There is is aa substantial
substantial literature
literature discussing
discussing the scattering
the scattering
gravitational loading (Gabor,1980).
gravitational loading (Gabor,1980). A A study has has been
been made
made of waves
waves from irregular
irregular surfaces.
surfaces. The monograph by
The monograph Beck-
by Beck-
of the
of the segment
segment control
control system
system performance,
performance, largelylargely using
using mann
mann and and Spizzichino,
Spizzichino, "The "The Scattering Electromagnetic
Scattering ofofElectromagnetic
geometrical optics
geometrical optics (Mast Nelson, 1982). In
(Mast and Nelson,1982). In the
the present
present Waves from
Waves from RoughRough Surfaces,"
Surfaces," (Beckmann
(Beckmann and and Spizzi-
paper we
paper we develop
develop somesome tools
tools useful
useful for for aaphysical
physical optics
optics chino,1963),
chino, contains an
1963), contains an excellent summary of
excellent summary various
of the various
understanding of
understanding of segmented mirrors in
primary mirrors
segmented primary general, and
in general, approachestoto this
approaches this problem
problemthat that have
have beenbeen used. used. We We havehave
the Ten Meter Telescope
Telescope in in particular.
particular. We believe tools
believe these tools drawn heavily
heavily fromfrom this monograph in sections 33 and
in sections and 4. 4.
will be
will be useful
useful because
because they span span the difficult region
the difficult between
region between
the limiting diffraction and
regions of pure diffraction
limiting regions optics.
geometrical optics.
and geometrical

SPIE Vol.
SPIE Vol 332 Advanced Technology Optical Telescopes
Technology Optical Telescopes(1982) 123
(1982)// 123
 The analyses presented
The analyses presented of effects of the
of the effects the segmenta-
segmenta- first
first two
two are 0.01 and 0.07
are 0.01 0.07 arc
arc seconds
seconds respectively.
respectively. This fine fine
tion
tion geometry,
geometry, the surface roughness, and and the
the phasing
phasing are
are for
for structure is integrated
integrated out out by by atmospheric
atmospheric seeing.seeing. The
monochromatic
monochromatic light
light and
and without effects of
without the effects of atmospheric
atmospheric remaining structure
remaining structure then results from
then results from the the 11 cm
cm crack
crack width
width
seeing. These assumptions
seeing. These assumptions are
are useful
useful totodetermine
determinefabrica-
fabrica- distribution along
and the distribution along a spike
spike isis given
given by by the
the simple
simple rec-
rec-
tion tolerances which
tion tolerances which will
will not
not degrade
degrade the image
image quality
quality even
even tangular aperture diffraction distribution; (4a2 (4a2bzb* 1).2)
/X2)
with the best seeing
seeing conditions
conditions andand with
with very
very narrow
narrow (sin(kax)/kax)
(sin(kax) /kax)2.2 . Here
Herex xis isthetheangular
angularimage
imageplaneplane coordi-
coordi-
bandpass
bandpass observations.
observations.InIn Section
Section 55 wewe discuss
discuss briefly
briefly the
the nate,
nate, 2a 2 a isisthe
thespider
spiderororcrack
crackwidth,
width, 22bb isis the
thecrack
crack length,
length,
effects
effects of
of atmospheric
atmosphericseeing.
seeing. The
The constraints
constraints on on the
the seg-
seg- and
and k = 277/k.27T/X.
mented
mented mirror control required for speckle
speckle interferometry
interferometry are
discussed
discussed in Section 6. Note that
Note that asas the
thecrack
crack width
width decreases,
decreases, the the energy
energy
diffracted into the central part part of the the diffraction spike
2. Segmented Aperture
2. Segmented ApertureDiffraction
Diffraction Image
Image decreases, and
decreases, and the
the angular
angular scale
scale ofofthe
thediffraction
diffraction effects
effects
increase. An
increase. An interesting
interesting result
result occurs
occurs for
for angles
angles larger
larger than
than
The geometry
The geometry of the the primary
primary mirror
mirror segmentation
segmentation will will characteristic diffraction
the characteristic diffraction angle O 0=X/2a.
=X /2 a. There aver-
There the aver-
cause a specific
cause specific diffraction
diffraction pattern
pattern that can be calculated
that can calculated in age intensity in the diffraction spikes spikes becomes
becomes roughly
roughly
detail using Fraunhofer diffraction theory. At visible independent of the crack
independent crack width
width and only dependsdepends on the total
wavelengths the
wavelengths fine structure of
the fine of the
thediffraction
diffraction image
image will
will bebe length of the the cracks.
cracks.
averaged by
averaged effects of
by the effects of atmospheric
atmospheric seeing
seeing andand mirror
mirror
roughness. Thus Thus onlyonly a qualitative description of of the the 2.2 Diffraction images
2.2 Diffraction images at infrared wavelengths
wavelengths
diffraction image
diffraction imagefor for visible
visible light
light isis warranted
warranted and and this
this is
given section 2.1
given in section 2.1 below.
below. At infrared wavelengths
At infrared wavelengths aa 10- 10- At infrared wavelengths
wavelengths the size
size of
of the
the diffraction
diffraction image
image
telescope can
meter telescope can approach
approachdiffraction-
diffraction-limited performance
limited performance isis larger
larger and
and the
the effects
effects of
of atmospheric
atmospheric turbulence
turbulence reduced.
reduced.
and
and thus aa quantitative
quantitative description
description of
of diffractive
diffractive effects
effects isis of scale structure is not
The fine scale not necessarily
necessarily integrated
integrated over
over and
interest. In section
section 2.2
2.2 we
we describe
describe thethe ideal
ideal diffraction
diffraction pat-
pat- thus aa detailed
thus detailed description
description ofof the
the diffraction
diffraction image
image isis of
the segmented
tern for the segmented aperture
aperture and
and discuss
discuss the pattern
pattern asas itit interest.
interest. The
The amplitude for the image is given by by
might appear in in practice.
practice.
I'(u,v)) _= ~J e ik(ux+^dxdy
J J'thee~ rk(ux+vy)dxdy (1)
(1)
aperture
2.1 Diffraction
2.1 Diffraction images
images for visible
visible light
light
In existing telescopes
telescopes bright
bright stellar
stellar images
images typically
typically show
where (x,y) are
where are the
theaperture
aperture coordinates,
coordinates, (u,
(w, v)
v) the
the image
image
angular coordinates,
angular coordinates, andand kk=27r/\ (Born and
=271. /k (Born and Wolf,
Wolf, 1980).
1980).
aa four
four spike
spike image
image pattern
pattern resulting
resulting fromfrom thethe diffraction
diffraction Since
Since the aperture is
the aperture is aa regular
regular array
array of
ofhexagonal
hexagonal apertures
apertures
effects
effects of secondary
secondary support
support trusses
trusses (spiders).
(spiders). With
With aa seg-
seg- this
this can
can be written
mented primary
mented primary mirror
mirror an additional
additional contribution
contribution to to the
the
diffraction 36
diffraction spikes
spikes will
willcome
comefrom
from the
the cracks
cracks between
between the seg-
the seg- (u,V) = hex (u,V)
ments.
ments. For For aamosaic
mosaic of of hexagonal
hexagonal segments
segments (Figure
(Figure 1) there , v) e- lk(x.,u + yw)1I (2)
will
will be
be aa pattern
pattern ofof six
six spikes
spikes instead
instead ofof the
the usual
usual four. r =1 1

Two parameters can be used to characterize the

Two parameters where
where xy ,yy are
xj,yi are the
the coordinates
coordinates of of the
the segment
segment centers,
centers, and
diffraction
diffraction image: the summation is
the summation is over thethe 3636 segments.
segments. "V 41 hex
hex is ampli-
the ampli-
tude for aa single
tude for single hexagonal
hexagonal segment
segment and and isis calculated
calculated by by
1.1. The
The fraction
fraction ofof the energy
energy incident
incident on the primary
primary integrating
integrating Equation
Equation 11 over
over the area of aa hexagon.
hexagon.
that is diffracted
diffracted byby the spiders or cracks:
cracks:
areaofofspiders
f = area spiders or
or cracks
cracks // area
area of
ofprimary
primary hex(u, V)-
3 a sin(3ß+a)sin(ß-a) +sin(3ß-a)sin(ß+a) )
2.
2. The angular scale scale of the diffraction
diffraction pattern: 4X ß(ß ) ß(ß+ )
9B -= wavelength
wavelength / (thickness
(thickness of support structures where
where aWJ a =f kau
/caw/4,/4,/3=fazv/4,
ß= kav /4, andand a^thea =the segment
segment edge edge
or width of of cracks)
cracks) (0.9m). The
length (0.9m). Thecoordinate
coordinatesystem
system isis defined
defined in in Figure
Figure 1.1.
We
We list
list these characteristic
characteristic parameters
parameters for some existing
existing large The irradiance distribution for aa single single hexagon,
hexagon, "*hex l^/^l122 ,
telescopes. is plotted
is plotted along
along aa spike
spike (on (on the
the uu-axis)
-axis) in Figure
Figure 22 and and
between
between the spikes
spikes (on the vv-axis)-axis) in Figure
Figure 3. 3. The
The distribu-
distribu-
telescope diameter f(%)
f( %) 0(arc
B (arc seconds)
seconds) tion
tion is plotted over the ranges ranges 00 to to 5,5, 50,
50, and
and500500 arcarc seconds
seconds
Palomar 5.0m 0.5 88
to illustrate the structure
to illustrate structure on on various
various scales.
scales. It is normalized
normalized
KPNO 4.0m 1.0 55
to 1.0 at
to 1.0 at 00=0.
=0. The distributions
distributions are areplotted
plottedforforaa -= 0.9 0.9 mm
Lick 3.0m 0.9 13
13 and for aa wavelength
and wavelength of of 10 µm. For
10 /im. Forotherotherwavelengths
wavelengths the
Las
Las Campanas 2.5m 0.9 88
angular scale
angular scale changes proportionally. In the distribution
changes proportionally. distribution alongalong
the spike
the spike there
there isis structure
structure on on the
the scale
scale of of 1.3
1.3 arc
arc seconds
seconds
(kax 3 /2 =ar) and
(/raW3/2=7r) and the
theaverage
average intensity
intensity falls falls off
offasas1192.
1/02.
With
With 11 cmcm widewide cracks
cracks between
between the the segments
segments these
these Between the
Between spikes the average intensity
the spikes intensity fallsfalls as 1/04. AA cir-
as 1/94. cir-
parameters for
parameters for the
the geometry
geometry of of figure
figure 11 areare f =^ 1.0%
1.0% and
and O9 cular aperture
cular aperture has aa diffraction
diffraction intensity
intensity which which falls
falls off asas
=* 10 arc
= 10 arc seconds,
seconds, i.e.
i.e. comparable
comparable to to those
those from
from struts
struts in 1/03
1/93..
existing telescopes.
existing telescopes. For narrower cracks
For narrower there is aa propor-
cracks there propor-
tional decrease
tional decrease in fractional energy
in the fractional energy and and aa corresponding
corresponding The irradiance
The irradiance distribution
distribution for for thethefullfullarray,
array,11112 ,, iss
increase in the characteristic
increase characteristic angular scale.
scale. shown along
shown along thethe uu-axis
-axis (direction
(direction of largestlargest angular
angular scale
scale
spike)
spike) in in Figure
Figure 4, 4, again
again forfor X^IO
X =10 /im. Figure 55 shows
µm. Figure shows the
Thus the diffraction
diffraction images from the the cracks
cracks will
will be corn-
com- intensity
intensity distribution
distribution along
along the vv-axis, between the
-axis, between the spikes.
spikes.
parable in
parable in intensity
intensity and
and size
size to those seen
to those seen on on existing
existing tele-
tele- Figure
Figure 4a 4a covers
covers the distribution out
the distribution out to to 55 arc
arc seconds
seconds and
scopes with
scopes with the scattered
scattered light
light distributed
distributed among
among sixsix spikes
spikes shows the
shows 0.25 arc
the 0.25 arc second
second structure
structure from from the overall
overall 10 10 m
instead of four. aperture. The first maximum
The first maximum away away from from the the central
central max-max-
A diffraction
A diffraction image distribution calculated for the imum is 2.1%
imum 2.1% as bright as the
as bright the central
central maximum.
maximum. AAcircular circular
geometry
geometry of Figure
Figure 11 contains
contains structure
structure on threethree angular
angular 10-meter
10 -meteraperture
aperture with
with no cracks gives the usual usual Bessel
Bessel func-
func-
scales
scales coming
corningfromfromthethethree
three scales
scalesof of the
the mirror;
mirror; the 10 m
the 10 tion diffraction
diffraction pattern which this maximum is 1.8%
pattern in which 1.8% of the
overall
overall aperture,
aperture, the 1.8 m segment
the 1.8 segment size, and the the 11 cm
cm crack
crack central
central brightness.
brightness. The The next highest maximum
next highest maximum is is then down
down
size.
size. For visible light
For visible light the angular scales
the angular resulting from
scales resulting from the
the in brightness by
in brightness by more
more than aa factor
factor of of ten,
ten, falling
falling faster than

124 //SP
124 /E Vol
SPIE Vol. 332
332 Advanced
AdvancedTechnology
Technology Optical
OpticalTelescopes
Telescopes (1982)
 diffraction pattern
the diffraction pattern for circular aperture.
for a circular aperture. FigureFigure 4b 4b cov-
cov-
the distribution
ers the distribution out to to 5050 arc
arc seconds
seconds and and shows
shows 2.6 2.6 arc
arc f z(x,y)z(x,y,T)dxdy
~
second
second structure
structure from from the the segment
segment spacing.
spacing. The The segment
segment C(r)
spacing in
spacing in the xx-axis
-axis direction
direction isis effectively
effectively 0.78 0.78 m m because
because of z2 (x,y)dxdy
fJ z2(x,y)dxdy
the staggered setting
the staggered setting ofof the
the hexagons
hexagons in in the array (see (see Figure Jz(x',y')dx'<fy'
r» e \ *fz(x
——?",y')dx'dy'
Finally Figure
1). Finally Figure 4c 4c gives
gives the distribution along
the distribution along the spike spike Z(x9y,r)
Z(x,y,T) -
out toto 500
500 arc
ture from the
2.6 arc
the 2.6
arc seconds
the 1.0
arc second
seconds and
1.0 cm crack
second scale
and shows
scale which
shows the
crack width.
the 200
width. There
which if itit were
200 arc
were brighter
arc second
There isis sharp
second struc-
sharp structure
struc-
structure on
brighter could
could possi-
on
possi- where Z(x,^,r)
where
J dx'dy'
dx'
Z(x,y,r) isisaaratio
f
ratio of
of line
line integrals
integrals and
(5)

andr T =t*tx - jT[

bly be
bly mistaken for
be mistaken for stellar
stellar structure. However, it is
structure. However, is 44 orders
orders Note
Note that
that C(0) =- 11 andand C(co)
C0»)=-O.O.The Theshapeshape ofofthis
this func-
func-
of magnitude
magnitude or 10 10 stellar
stellar magnitudes
magnitudes reducedreduced in in brightness.
brightness. tion depends
depends on the details
details of the
the origin
origin of
of the
the surface
surface irregu-
irregu-
addition, this structure is
In addition, is only
only sharp
sharp in in one
one dimension
dimension and and larities.
larities. AA simple
simple reasonable
reasonable model
model for the the correlation
correlation func-
func-
thus does not produce
thus does produce pointpoint-like
-like images.
images. Given the dynamic dynamic tion, which is also useful for for calculations
calculations isis aa gaussian.
gaussian.
range of infrared
range infrared detectors
detectors thisthis structure is is unlikely
unlikely to cause C(T) = ee~^
C(T)- z2'T2T2 (6)
(6)
any problems.
any problems. Figure shows the spike
Figure 6 shows spike intensity
intensity distribution
distribution
an array
for an array with
with no no cracks.
cracks. In this this case
case the structure
structure in the the T
T is correlation length along
is the correlation along the
the surface
surface averaged
averaged over
over all
all
image results
image results from overall 10m
from the overall 10m aperture
aperture and and thethe angular
angular directions. The
directions. The correlation
correlation length
length roughly
roughly defines
defines the
the spatial
spatial
edge of
outer edge of the
the array.
array. frequency of
frequency surface variations.
of the surface variations. Two Two different
different surfaces
surfaces
The discussion
The discussion and equations above
and equations above and and thethe associated
associated might have
might have the same
same rmsrms height
height variation
variation but
but very
very different
different
figures are
figures monochromatic light.
are based on monochromatic light. In practice
practice obser-
obser- spatial frequencies
spatial frequencies andand thus correlation
correlation lengths.
lengths.
vations
vations maymay be made made with
with aasignificant
significant bandwidth
bandwidth which which willwill Beckmann and Spizzichino
Beckmann Spizzichino give the physical optics
average
average overover the fine scalescale structure in in the
the irradiance
irradiance distribu-
distribu- diffracted
diffracted field
fieldfrom
from aa point
point source
source for
for the
the random rough sur-
tion. In addition the fine
In addition fine structure
structure willwill be be further
further averaged
averaged face described above.
face described above. For smallsmall scattering
scattering angles
angles the irradi-
irradi-
over
over by by anyany aberrations
aberrations in in the optical
optical surfaces
surfaces and and by by the
the ance
ance as a function of of angle
angle isis given
given byby
atmospheric
atmospheric seeing.seeing. We We conclude
conclude that that there
there will not be
will not be II (0)I2 = é gl o(0)I2 + D (7)
artifacts
artifacts in in the
the infrared image due to diffraction
infrared image diffraction from the seg- seg-
mentation geometry. Finally
mentation geometry. Finally we we note
note thatthat thethe large
large scale
scale Here |^0 (0)| 2 isisthe
(0) 12 the irradiance
irradiance ofof the
the aperture was cal-
aperture that was cal-
spikes
spikes from
from the cracks arise from the the very
very special
special relationship
relationship culated in
culated in section
section 22 above and gg and D
above and D are
are given
given by
by
between
between the the phases
phases of of the segments
segments given given by phases of
by the phases
the vectors in Equation
the unit vectors Equation 2. 2. Any
Any phase
phase errors
errors in the posi-posi- g=
tioning of
tioning of the
the segments
segments will will alter
alter these
these phases
phases and and thus
thus X
strongly effect the size
diffraction
diffractionspikes.
spikes. These
size andand even
These phasing
even the
phasing errors
the existence
errors and
existence of the the
effects are
and their effects are D rTég
x
2 00
gm C
-k20 2 T2/4m
kZBZT2/4m
(8)
discussed
discussed in Section 4. m-1 m!m
The fraction, F, of
fraction, F, ofthe
thetotal
totalscattered
scatteredenergy
energy contained
contained within
within
a cone of
of half
half-angle 0' is
-angle O"
9'
3. Segment Surface
3. Segment Surface Roughness
Roughness
F(9")3= e g f 1410 (0) I29 dd dcA
The
The surface
surface roughness
roughness of the the primary
primary mirror
mirror should
should o
occur on the
occur on the three
three scales. On the
scales. On the finest
finest scale there is
scale there is aa
roughness from the
roughness the polishing process. The
polishing process. The scale
scale of these
these +1-C g(1 +i -CJL-.
(kTBe-(kTO'/2) 2/m
"l2)2 /m (9)
irregularities is
irregularities typically1010toto 30
is typically 30 Angstroms.
Angstroms. On On aa larger
larger l Ifl-l
m =1
m!
scale there
scale is a roughness
there is roughness due due to to high
highspatial
spatial frequency
frequency pol-pol-
ishing aberrations
ishing (high order Zernike
aberrations (high Zernike polynomials).
polynomials). BasedBased We have been
We been unable
unable totocalculate
calculate analytically
analytically the integral
integral in
on experience
experience with polishing, we expect these residual
with mirror polishing, residual first term for
the first for aa hexagonal
hexagonal array.
array. However,
However, we we can
can getget aa
aberrations
aberrations to surface height
to have rms surface height of of order
order XX/20/20 == 0.025
0.025 feel for the
feel for the relative
relative size
size of
of these
these terms
terms byby using
using aacircular
circular
/Ltm. Finally the lower
µm. Finally lower order aberrations
aberrations in each each mirror
mirror (pis-
(pis- aperture,
G/X)(2/1
GI X) (2 J1
for
(ka0)/ka0).
which
(ka0)/ka0).The
the
The integral is
amplitude
is then
then 1.0
V0
''Yo (9)
(8)
1.0 --Jp/<?(ka0')
-
Oca0')
-
tilt, focus,
ton, tilt, focus, astigmatism,
astigmatism, coma), although not random for
coma), although
an individual segment,
an individual segment, perhaps
perhaps can can be
be treated
treated as aa random
random -Ji/!2(ka0').
- (ka0').
roughness
roughness on on the
the scale
scale ofof the
the full
full primary.
primary. To To the
the extent that In general the
In general full series
the full series for D D needs
needs totocalculated.
calculated. In In
the number of of segments
segments isis large
large this
this isisaauseful
usefuldescription.
description. the
the limiting
limiting cases casesofofg«l g<<1and andg»lg» 1 the expression
expression
These
These variations
variations may
may be be of
of order
order 0.1 0.1 to
to 0.2
0.2 /u,m.
µm. follows.
simplifies as follows.
A useful model for a rough rough surface
surface is is given
given by the "nor-
"nor- 1) For aa "smooth"
1) For "smooth"surface,
surface,(47rcr/X)
(4iro-/X)2«1,
2 « 1, andand
mally distributed" surface.
mally distributed" surface. Scattering
Scattering from from this
this surface
surface isis 2
described
described by by Beckmann
Beckmann and and Spizzichino.
Spizzichino. An isotropic,isotropic, ran- r
e-(rTo/x)
4QI*(0I2= Zl
dom
dom surface
surface isis defined
defined by by two
two functions.
functions. The first is the dis-
Yo0)12- Y 1I [I,o(9)I2
i*u v/|
T2
2
dis- X A.2
/ ^
tribution
tribution of surface heights
of the surface heights above above and and below
below the mean
mean II

surface
surface z^O.
z =0. This
This distribution
distribution isis assumed
assumed to be normal
to be normal with
with The first term is
The first is the
the specular
specular diffraction
diffraction of aa perfect
perfect mirror,
mirror,
the probability
probability density
density of height z given by and the second
second isis the
the diffuse
diffuse scattering
scattering from
from the
the irregularities.
irregularities.
11 - 22/2Q2
2/2<r 2 We
We can integrate this
can integrate this expression
expression toto find
find the fraction of
the fraction of the
the
w(z) - e (4) energy
energy enclosed
enclosed inin a circle
circle of angular
angular radius 0".
0'.
N/27TO-

The second function

F(0')= Fo(0')(1 - g) + g(1 - e (rre'la)2) (11)
The second function that must bebe specified
specified is the
the correlation
correlation
of
of surface heights at
surface heights at two
two separated points as
separated points as aa function
function of where F0
where FQ (0)
(0) is the
the fractional
fractional enclosed
enclosed energy
energy for the
the perfect
perfect
their separation.
separation. This
This is independent
independent of the the surface
surface height
height mirror.
distribution. If we
we take
take the
the roughness
roughness to to be
be isotropic,
isotropic, the
the 2)
correlation depends
correlation dependson on only
only one
one variable,
variable, T,
r, the
the distance
distance 2) For a "rough"
"rough" surface,
surface, (4iro-/X) 2 »l,
(4aro- /X)2» and
1, and
between
between anyany two points (x,y)
two points (x,y) and
and (x',y "). This
(x'jO. This correlation T2
4(0)12 T2 p-(T0/4a) 2
(TB/40)2.
function is given by 167TO- 2
(12)
16/ro-2

SPIE
SPIE Vol.
Vol 332 Advanced Technology Optical Telescopes
Technology Optical Telescopes(1982)
(1982)// 125
125
 In this
this limit
limit only
only the diffuse
diffuse field contributes. This result is
field contributes. is goal. TheThe lower
lower spatial
spatial frequency
frequency aberrations
aberrations with
with larger
larger
same as
the same as that
thatforforgeometrical
geometrical scattering
scattering from
from aagaussian
gaussian correlation
correlation lengths
lengthswill
willneed
needtoto be
be less
less than
than about
about 0.10.1 pmµm oror
distribution
distribution ofof slopes.
slopes. DefineDefine aa slope
slope 00 =V2 a/T
00=v2 cr/T (For
(For X/5. polishing aberrations
X /5. The polishing expected to be
aberrations are expected be within
within this
this
2<r«
2Q« T,71,Bp00isisthe
therms
rms slope.)
slope.) and then limit.
1l ,(918)2 Figure
Figure 88 shows
shows the limiting
limiting regions
regions for a wavelength
wavelength of
1`Y(B) I2 = (13) 10 /im
10 µm andand an an angle
angle ofof0.2
0.2 arc
arc seconds.
seconds. TheThe first
first integral
integral inin
8riOct02
87T0Q Equation 99 is
Equation equal to
is not equal to 1.0
1.0 for
for these
these conditions
conditions and and the
the
The fraction,
fraction, F,/% of
ofthe
thetotal
totalscattered
scatteredenergy
energy contained
contained within
within lower part
lower part of curves are altered
of the curves altered from
from those
those inin Figure
Figure 7. 7.
of half
a cone of half-angle
-angle 9"9' isis given low
At a given low crQ aa larger
larger value
value ofof TT is needed
needed at 10 10 fjum
µm to to
achieve the same
achieve same fraction
fraction since
since the
the perfect
perfect mirror
mirrordiffraction
diffraction
F(0")= i - e 8
1
--
070)2
(14)
pattern is
pattern is contributing
contributing at at 0.2
0.2 arc
arc seconds.
seconds. TheThe axes
axes in in Figure
Figure
88 have
have been
been scaled
scaled appropriately
appropriately soso thatthat in
in the regions
regions of
large
large crv andand TTthey they match
match those
those inin Figure
Figure 7.7. Based
Based on
Now armed
Now armed with with the the general
general result result in in Equations
Equations 7, 7, 8, assumed noise
assumed noise levels
levels inin the
the sensors
sensors and and actuators
actuators of of 5050
and 9 and
and and the the limiting
limiting casescases we we can can apply
apply them them to to a seg-seg- nanometers
nanometers we we have shown the rms surface
have shown surface error for the the Ten
Ten
mented primary mirror. Meter Telescope design
Meter Telescope 0.21 pm
design isis 0.21 (Mast and
µm (Mast andNelson,1982).
Nelson, 1982).
For surface roughness
For surface roughness due to to polishing
polishing we we can assume assume This surface
This surface error
error is almost
almost entirely
entirely due
due to pistoning
pistoning of the
vcr=30
=30 Angstroms.
Angstroms. For For X= X 8* 0.50.5 µm/im(g(#=0.006)
=0.006) the the"smooth"
"smooth" segment
segment centers
centers and
and thus correlation length
thus has a correlation length ofof about
about 22
formula
formula above above indicates
indicates the the intensity
intensity at at 0B = 00 is is down
down by less Figure 88 indicates
meters. Figure indicates that 10 pm
that atat 10 µm about
about 80%80% of the
than 11%, %, aa negligible
negligible effect.effect. The fractional
fractional energy energy contained
contained will be enclosed in aa 0.3
energy will 0.3 arc
arc second
second radius
radius circle.
circle.
in
in aa given
given circle
circle alsoalsochanges negligibly. For
changesnegligibly. For the the longer
longer
infrared wavelengths
wavelengths the effect is is even smaller.
For surface aberrations
For the surface aberrations resulting
resulting from from the the polishing
polishing
we
we might
might expect
expectaacrcrof ofabout
about 0.0250.025/xm, µm, giving
giving #=0.4. g =0.4. Thus 4. Segment
4. Segment Phasing
Phasing
g# is
is about
about 11 and and the the full
full series
series for for DD isis required.
required. Surface
aberrations
aberrations from polishing are largely
from polishing largely described
described by low order
by low primary mirror active
The primary active control
control system
system maintains
maintains the the
Zernike
Zernike polynomials.
polynomials. For 1.8m diameter
For 1.8m diameter segments,
segments, they give give phasing of the segments
tilt and phasing segments in the the face
face of
of thermal,
thermal, gravi-
gravi-
correlation lengths, T,
correlation lengths, T, in the range range from from 0.2 0.2 toto 0.40.4 meters.
meters. tational, and wind perturbations.
tational,
Rather
Rather than than calculate
calculate the the irradiance
irradiance we we can can use use the the The desired tiltstilts of the segments
segments can be determined dur-
corresponding
corresponding expression expressionfor forFF (Equation
(Equation 9) 9) to see see which
which ing a calibration
ing procedure using
calibration procedure using aa star,
star, and
and thenthen will
will bebe
values of
values of cr
Q andand TTwill will allow
allow various
various optical
optical criteria
criteria to to be maintained
maintained by by the control system. This Thiscalibration
calibration procedure
procedure
satisfied. Figures
Figures 77 and and 88 show
show the v- cr- TT plane
plane and and show
show the will
will bebe automated,
automated, under computer control, and and require
require only
only
regions
regions in in this plane where various various opticaloptical goals
goals are achieved
achieved aa few minutes. Thus
few minutes. Thus itit can
can be be done
done atat least
least once
once per per night
night
for a 10 10-meter
-meter diameter
diameter mirror.mirror. The The optical
optical goals
goals are defineddefined and
and possibly
possibly moremore often.
often. The The tilt
tilt degradation
degradation of the the image
image
by the
by the angle
angle B" 0' andand the the fraction
fraction of of thetheenergy
energy contained
contained due to noise
due noise inin the
the sensors
sensors and and actuators
actuators hashas been
been studied
studied
within aa circle
within circle ofof that
that radius.
radius. In In Figure
Figure 77 the the limits
limits havehave been been using
using geometrical
geometrical optics
optics (Mast
(Mast and and Nelson, 1982). Assuming
Nelson, 1982). Assuming
calculated for aa wavelength
calculated wavelength of of 0.5
0.5 µm. /im. The acceptable acceptable 50 nanometer
50 nanometer noise noise levels
levels in the the sensors
sensors andand actuators
actuators we we
regions are
regions are toto the lower
lower right of the curves. The Theoptical
optical goals
goals showed that
showed that 80%
80% of of the
the energy
energy is contained
contained in in a circle
circle of
have been
have been calculated
calculated for for anan angle
angle 0' B" =0.1
= 0.1 arc arc seconds
seconds and and radius
radius 0.100.10 arc
arc seconds.
seconds. Prototype
Prototype actuators
actuators and sensors sensors
for the fractions
for fractions 0.5 0.5 through
through 0.9. 0.9. For small small #g (small (small cr),cr), the which
which have have been
been built
built forfor thethe TenTenMeter
MeterTelescope
Telescope are are
dependence on
dependence on the correlation length
the correlation length disappears.
disappears. Thus Thus for operating with
operating with substantially
substantiallyless lessnoise
noise than
than this. Thus we
this. Thus we do do
surface variations
surface variations which which are are much
much less less than X, X, thethe scattering
scattering expect the component
not expect component noise noise contribution
contribution to degradedegrade the
is independent
is independent of of the slope of the surface
the slope surface irregularities.
irregularities. For image beyond
image beyond that caused
caused by by atmospheric
atmospheric seeing. addition,
seeing. In addition,
X s" 0.5
X= 0.5 microns,
microns, a= a- 55meters,
meters,and andB" 9' ==0.10.1 arc arcseconds
seconds the measured stability
the measured stability of the the prototype
prototype sensors
sensors suggests
suggests thatthat
quantity
quantity kaO" katf is is large
large and the the first
first integral
integral in in Equation
Equation 99 recalibration need only be done
recalibration done infrequently.
infrequently.
equals
equals 1.0.1.0. For large large values
values of kaO" ka9' the the scaling
scaling to to different
different phasing of the mirrors
mirrors may may be be aa more
more time
time consum-
v,
cr, T,X
T,X is is apparent
apparent from from Equation
Equation 9. 9. ThusThus figure
figure 77 can can be The phasing consum-
scaled
scaled to to other wavelengths
wavelengths in in the region around 11 micron, micron, by by ing
ing procedure
procedure and thus for for practical
practical reasons
reasons oneone would
would like
like itit
simplyscaling
simply scalingthe thecrcrand and T axes. axes. The Thelimits
limits for fordifferent
different required infrequently.
to be required infrequently. Calculation
Calculation of ofknown
known effects
effects and
and
values
values of 0' 9" can alsoalso be obtained by scaling scaling the the TT axis.
axis. drifts indicate itit is probable
drifts indicate probable that that the
the initial
initial phasing
phasing (accurate
(accurate
to a small
small fraction
fraction of of visible
visible wavelengths)
wavelengths) will will be
be maintained
maintained
Note Figure 77 that
Note in Figure thatthethegeometrical
geometricaloptics optics approxima-
approxima- by
by the control
control system.
system. However,
However, even even though
though longlong term
term sta-
sta-
tion
tion using
using slope
slopeerrors
errorstotocalculate
calculateimage image blurblur(#>>!)
(g» 1) worksworks bility
bility ofof the phasing
phasing is expected,
expected, itit isis worthwhile
worthwhile to to establish
establish
quite
quite well
welldowndowntotoabout aboutcrcr*==0.03 0.03/u,m µm(#= (g= 0.4), and thus thus itit the consequences of partially
the consequences partially or or completely
completely unphased
unphased seg- seg-
will
will be adequate approximation
be an adequate approximation for for many
many optical
optical calcula-
calcula- ments.
tions,
tions, even
even when
when #»1 g>>1 isisnot notsatisfied.
satisfied. The amplitude given in Equation 22 isis modified modified to include
include
The optical goal
The optical goalfor for the
the primary
primary mirror mirror of of the
the Ten Meter the effect
effect of the segment tilt tilt and
and phasing
phasing as as follows
follows
Telescope
Telescope calls callsforfor 80%80%ofofthe the energy
energy to to be
be within
within a circlecircle of 36
radius 0.16 arc
radius 0.16 seconds (for
arc seconds (for XX= 0.5 0.5 microns).
microns). The The limit
limit for for %Y(u,v) _ E*heX(u +2aj,v +2ßj)e-ik(x u+ v; v +2 Y',) (15)
this optical goal
this optical goal isis given
given by by Figure
Figure 77 with with the the TT-axis
-axis scalescale j7=1
-1
expanded
expanded by byaa factor
factorof of 1.6.
1.6. At largerlarger values
values of of cr, this opti-
v, this opti- where a,/3, andyyare aredefined
definedfor foreach
eachsegment
segment by by the
the error
error
cal goal
cal goal becomes
becomesaa limit limiton on thethe rmsrms slopeslope of the the surface,
surface, where a,ß, and
900<
0 < 0.045
0.045 arc-
arc-seconds.
seconds.
surface describing
surface describing the deviation of the
the deviation the surface
surface from
from its
its ideal
ideal
position.
Figure
Figure 77 shows
shows that that for the the high
high spatial
spatial frequency
frequency aber- aber-
rations we
rations we need
need the the rmsrms surface
surface height heighttoto be be <0.025µm
^0.025/im 8z = aa(x-y)+p(y-y)+y
Sz j (x- yj) +ßj (y- yj) +yj (16)
(«X/20).
<X /20). A A new
new technique
technique for polishing polishing the the mirror
mirror seg- seg- To understand the the effects
effects of
of phasing
phasing alone
alone we we can
can make
make the the
ments is
ments is being
being developed
developed for for the Ten Meter Meter Telescope
Telescope (Lub- (Lub- assumption that
assumption that the
the tilt errors OLa and
tilt errors and ßft are
are small.
small. For
Forvisible
visible
liner and
liner and Nelson,
Nelson, 1980, 1980, and and Nelson
Nelson etal, etal, 1980.
1980. An An initial
initial light this is not aa good
good assumption
assumption since
since the
thediffraction
diffraction limit
limit isis
demonstration
demonstrationof of the
the technique
technique on on aa 3535 cmcm mirror
mirror met this 0.012 arc
about 0.012 arc seconds
seconds andand the
thegeometric
geometricoptics
opticscalculation
calculation

126 //SPIE
725 SPIEVol.
Vol. 332
332Advanced
AdvancedTechnology
TechnologyOptical
OpticalTelescopes
Telescopes (1982)
(1982)
of segment tilt errors mentioned mentioned aboveabove gives
gives 0.10
0.10 arc
arc second
second the contribution from
the contribution from thethe phased
phased array
array isis negligible.
negligible. Thus,
images. At 10
images. 10 µmju,m ititisisbetter
bettersince
sincethethediffraction
diffraction limit
limit is
is for this phase
for this phase error wewe expect
expect the diffraction
diffraction spikes
spikes from
from thethe
about
about 0.24 0.24 arc
arc seconds.
seconds. The The calculation
calculation of the the irradiance
irradiance dis- cracks to disappear.
cracks disappear. The diffraction image
image will
will be
be that due to to aa
tribution
tribution from Equations 15
from Equations 15 and
and 1616 for thethe hexagonal
hexagonal arrayarray single hexagon.
single hexagon. At At aa wavelength
wavelength of 10 10 um o2 - 0.063
ttm cr^2 0.063 and
and
with cracks
with and with
cracks and with thethe expected
expected tilttilt and
and phasing
phasing noise
noise the phased
the phased array
array makes
makes the dominant
dominant contribution
contribution to to the
the
requires
requires a monte carlo carlo computer
computer calculation.
calculation. ThisThis calculation
calculation diffraction image. We We have
have found
found no no tractable
tractable analytic
has
has been
been done and the the results
results are
are described
described in in Mast
Mast andand Nel-
Nel- expression
expression forfor the
the fractions
fractions FF for the array
array or or for
for an
an indivi-
indivi-
son (1982). The
son (1982). The error
error surface
surface created
created by by sensor
sensor noise
noise isis dual
dual hexagon.
hexagon. So So again
again to get
get aa qualitative
qualitative understanding
understanding of of
dominated
dominated by by pistoning
pistoning of of the segment centers,
the segment centers, i.e.
i.e. phase
phase the unphased system
the unphased system wewe use thethe fraction
fraction forfor aa circular
circular aper-
aper-
errors.
errors. With With the
the assumed
assumed sensorsensor noises
noises the rms rms surface
surface error ture. If we assume a circular
circular aperture of radius RR we
of radius wecan
can cal-
cal-
isis 0.21
0.21 /Ltm.
µm. WeWe calculate
calculate that
that at 1010 µm
/im thethe assumed
assumed sensor
sensor culate the
culate functions F(0)
the functions FQ (0)(Born
and F0(8)
F(0) and (Bornand andWolf):
Wolf):
noises give
noises give anan 80%
80% imageimage radius
radius ofof 0.27
0.27 arcarc seconds.
seconds. We We
describe
describe here here an analytic calculation
an analytic calculation ofof the
the effects
effects of
of only
only the F(6) = 1 -Ji (kR0) -Jl (kRe)
phasing
phasing errorserrors in
in order
order to give a qualitative
qualitative understanding
understanding of of F0(0) = 1 -Ja (kao)-/2 (ka9).
-J? (kao). (23)
their effects.
their effects. We We therefore
therefore makemake thethe assumption
assumption here here that
that
the
the tilt
tilt errors
errors are
are small. irradiance then reduces
small. The irradiance reduces to Figure
Figure 99 shows
showsthe
theexcluded
excludedenergies 1 1 FQ
energies -F0and F, as
and1 1 -F, as
function of
a function of image
image angle
angle for
for apertures
apertures corresponding
corresponding to the the
I' (u, v) 12
j =1
é rk(x a +y +2y) 2, (17) full primary
full primaryand
Now
Now using
segment,R—
andaasegment,
using Equations
Equations22
R= 5m
22andand 23
and a=
5m and
23 for
5/6 m.
a= 5/6 m.
given F^
for aa given F° we we can
can
where the
where the final
final factor
factor isis the
the amplitude
amplitude squared squared of of aa sum
sum of relate kRkR9'
B" to to Qy,
cry , and this is
and this is plotted
plotted in Figure Figure 10 10 for
for
unit vectors with
unit vectors with random
random phases.
phases. The statistics of
The statistics of the
the sum of F^O.S
F° 0.9. For
=0.8 and 0.9. For kX/50
/50 oror less
less there
thereisisonlyonly aa small
small degra-
degra-
random
random vectors
vectors is treated in
is treated in wonderful
wonderful detail detail by
by Beckmann
Beckmann dation of
dation of the
thediffraction
diffraction-limited
-limited image.image. For For rms rms surface
surface
and
and Spizzichino.
Spizzichino. We We have
have used
used their
their treatment
treatment to derive the
to derive the larger than kX/50
errors larger however the
/50 however the image
image size (9' )) contain-
size (8" contain-
irradiance
irradiance including
includingthe the effects
effects of of the cracks
cracks between the seg- seg- ing 90%
ing 90% ofof the energy in particular
the energy particular grows
grows quitequite rapidly.
rapidly. If the
ments.
ments. We We derive
derive thetheirradiance
irradiance for fortwo twodifferent
different possible
possible segments
segments are arephased
phasedtotoanan rms rms of of 0.2 0.2 µm,
/im, then
then near near
distributions of
distributions of phases.
phases. The The first
first isis with
with the
the segment
segment heights
heights diffraction-limited
diffraction -limitedperformance
performancecan can be be achieved
achieved at 10 pm.
at 10 µm.
random
random over over a range
range of of heights
heights that that exceeds
exceeds aa wavelength,
wavelength, The
The image
image diameter
diameter containing
containing 80% 80% of of the
the energy
energy is
all phases
i.e. all phases inin the
the interval
interval (- (-77,77- ) are equally
ir,ar) equally probable.
probable. shown
shown as as a function
function of of wavelength
wavelength in Figure Figure 11 11 for
for various
various
The net irradiance
irradiance isis then given by by values
values of of rms
rms piston
piston error. The The change
change from from segment
segment
111%.(u, 012 = NIT h,(u,V)I2. (t/,v)P. (18) diffraction limit (the (the upper
upper straight
straight line) line) to fullfull primary
primary
diffraction limit (the
diffraction limit (the lower
lower straight
straight line)
line) as as the
the wavelength
wavelength
It is
is noteworthy
noteworthy that the the parameters
parameters of the the array
array areare absent
absent increases isis illustrated.
increases wiggles in
illustrated. The wiggles in the curves come come from from
from this
from this expression;
expression; that that isis itit doesn't depend
depend on the the crack
crack the
the zeros
zeros of thethe Bessel
Bessel functions
functions in in Equation
Equation 23. 23. Again
Again an
width. It only only depends
depends on the the dimensions
dimensions of of aa single
single hexa-
hexa- surface phasing
rms surface phasing of of0.2
0.2p.mfjumgives
givesnearneardiffraction
diffraction-limited
-limited
gon. wavelength of
performance at a wavelength of 10
10 p.m.
//.m.
A second
A second possibility
possibility isis a gaussian
gaussian distribution
distribution of phasephase
errors.
errors. IfIfthe
the phase
phase errors
errors have
have the thegaussian
gaussian distribution
distribution
-"^
w(Ch) - l1 e ml2Q,
r- e~^'2(T \ (19)
(19)
5. Atmospheric
5. Atmospheric Seeing
Seeing
Typically, the
Typically, the irregularities
irregularitiesof of the
the index refraction of
index of refraction
where 04) = 47ry
where /k. The
47T <y/X. Therelationship
relationship between
between the rms
rms sur-
sur- atmosphere limit
the atmosphere limit the
the achievable
achievable resolution
resolution of of aa ground
ground--
face
face height,
height,ocry , and the rms phase,v0,
phase, cr^, isissimply
simply based optical
based optical telescope
telescopetotoangles anglesofof the the order
order of one one arc arc
second. Parallel light
second. Parallel light rays from a distant star star that
that are
are incident
incident
<r+ -= 47rcry/X,
(To 4irvy /k, (20) upon the
upon the telescope
telescope mirror
mirror suffer
suffer different
different delays,
delays, so so what
what
which
which includes
includesaafactor
factorof
of 22 from
from the reflection from
the reflection from the
the sur- started as plane wave
as a plane wave has
has been distorted before reaching reaching the
face. Again
Again using results of
using the results ofBeckmann
Beckmann and andSpizzichino
Spizzichino telescope. JustJust as as with
with thethecalculations
calculations aboveabove forfor surface
surface
the irradiance is given by roughness, ifif we
roughness, we have
have a model
model of of the
the wavefront
wavefront distortions
distortions
we
we can
can calculate
calculate thethe effect
effect onon the image.
image. We We briefly
briefly describe
describe
ICY °(u,0I2= 1111(u,y)I2e °L +NI1I hex (u,y)I2(1 -é °g)(21) the standard description
the standard description of phase errors
of the phase errors introduced
introduced by by
The first term contains
contains thethe irradiance
irradiance ofof the
the perfectly
perfectly phased
phased atmospheric turbulence.
array and is
array and is thus defined
defined by the the array
array parameter
arameterss. including
including Many
Many analytical
analytical and and experimental
experimental studiesstudies have
have beenbeen
crack width. The total
the crack total irradiance
irradiance 111%.(u,
pP^d/, v) v)PI` isis just aa made of
made of the
the effects
effects of atmospheric
atmospheric turbulence
turbulence on on light
light
fraction of the image
fraction image from aa perfectly
perfectly phased mirror added added to to transmission.
transmission. We We refer
refer primarily
primarily toto the
thetwo
two excellent
excellent articles
articles
fraction of the
a fraction the image
image from
from aacompletely
completely randomly
randomly phasedphased of Hufnagel
Hufnagel and and Stanley
Stanley (1964)
(1964)andandFried
Fried (1966).
(1966). The model model
mirror. Thus
mirror. Thus the transition from
the transition from one
one extreme
extreme to other is
to the other is used characterizes the wavefront
used there characterizes wavefront distortions
distortions by by aa phase
phase
smooth. structure function
fractional energy
The fractional energy contained
contained in in aa circle
circle of
of radius
radius 0' B' is D(r)
D(r) = <i(fit) - Ot3c')J2> ,, r = tx -z i (24)
given by
where ^ is is the wavefront
wavefront phasephase (in
(in radians), vectors Hz
the vectors
radians), the
Fv (ff)- (e(e'tr
F °(9')= ^F(0')+(1(l-e-^F^tf),
-'4)F(61')+ -e °m5Fp(O'), (22) and ~x refertotopoints
z refer pointsininthetheaperture
apertureplane,
plane,andandthe
theaveraging
averaging
where F(Q')
where F(0') and
andF0FQ(9')
(9')are
arethe
thefractional
fractional energies
energies for
for the
the is
is over
over both
both thethe high
high frequency
frequency of the the light
light itself
itself and
and thethe
atmospheric change
atmospheric change times.
times. The phasephase 4$ hashasaagaussian
gaussian distri-
distri-
phased
phased array
array and individual segment respectively.
respectively. with a mean of zero.
bution with zero. This
Thisstructure
structurefunction
functionassumes
assumes
The
The fractional
fractional energy
energy for the phased
for the array, F(9'
phased array, ),
F(0 "), the
the average
average wavefront
wavefront distortions
distortions are are isotropic
isotropic and and thus
thus
includes
includes the
the effects
effects of large diffraction
of the large diffraction spikes
spikes from
from the
the depends only
depends upon r.r. ItItdescribes
only upon describes thethedifferential
differential rather
rather than
than
cracks
cracks between
betweenthe the segments.
segments. Based
Based on on the study made
the study made of absolute optical pathlengths.
absolute optical pathlengths. For For the phenomena
phenomena of interestinterest
the sensor and actuator
the sensor actuator noise
noise we
we expect
expect the phase error
the rms phase this
this is sufficient.
sufficient. On On the
the basis
basis of
ofindex
index ofofrefraction
refraction varia-
varia-
ay
Qy toto be
be about
about 0.2
0.2 µm.
/u,m. tions described
tions described by by the
the Kolmolgoroff
Kolmolgoroff theory theory of ofatmospheric
atmospheric
For
For visible lightthis
visible light thisimplies
impliescr^2
02 =25
=25 andand

SPIE Vol.
SPIE Technology Optical Telescopes
Vol 332 Advanced Technology Telescopes (1982)
(1982)// 127
127
 turbulence
turbulence the structure function
function isis given
given by
by 6. Speckle Interferometry
6. Speckle
D(r)- 6.88(r/r0)5/3. (25) A number of of studies
studies have have been
been made
made of of the
the features
features of of
where r0,
where r0, the
the transverse
transverse coherence
coherence length,
length, depends
depends on the
the large
large telescopes
telescopes that that are required to to do
do speckle
speckle interferometry
interferometry
wavelength XX and
wavelength and zenith angle a
zenith angle as follows
a as follows and speckle
speckle imaging
imaging (Dainty
(Dainty 1977,1977, Nisenson 1980). 1980).
re--X615(sec a) -3)5. Analysis
Analysis of photon and
of photon and speckle
speckle statistics
statistics showshow that for for
(26) most applications
applications an an improved
improved signal signal to to noise
noise ratio
ratio and reso- reso-
The angular size size of the image image blur varies
varies as asyX"
X-1/5.1/5 . lution indeed achieved
lution are indeed achieved with with aa larger
larger aperture
aperture telescope.
telescope.
Various
Various aspects
aspects of of this
this model atmospheric distor-
model of the atmospheric Thus the the Ten
Ten MeterMeterTelescope
Telescope will will bebe anan improved
improved instru- instru-
tions
tions have been been confirmed
confirmed experimentally
experimentally using using stellar
stellar seeing
seeing ment for doing doing bothboth speckle
speckle interferometry
interferometry and andspeckle
speckle imag-
imag-
data stellar interferometry.
data and stellar interferometry. In one study
In one study ofofnight night-time
-time ing.
conditions
conditions atat mountain
mountain sites sites the coherence cell
the coherence cell size,
size, ro,r0 , was
was The effect of telescope
The effect telescope aberrations
aberrations on specklespeckle work work has
measured
measured to normally distributed
to be normally distributed from night to night night with investigated both
been investigated both analytically
analytically and and experimentally.
experimentally. The The
aa mean
mean of about 99 cm cm and and changing
changing by by a factor
factor of 1.4 1.4 with
with overall conclusion
overall conclusion isis that the the optical
optical quality
quality of of the
thetelescope
telescope
one standard deviation
one standard deviation occurrence
occurrence (Walters
(Waltersetal, etal, 1979).
1979). An An does not
does not have
have totobebeexceptional
exceptionaltotoachieve achievenear neardiffraction
diffraction--
excellent study
excellent study of ro r0 and the atmospheric
atmospheric modulation transfer limited results
limited results and the telescope
telescope aberrations
aberrations do do not have to be
function as a function of geographic
function geographic location,
location, time time of day, day, andand calibrated. The
known or calibrated. The aberrations
aberrationsininthe theoptics
opticswill will have
have aa
season has been
season been mademade by byWalters
Waltersand andco- co-workers
workers (Walters (Walters small effect
small effect provided
provided the geometrical
geometrical image image spread
spread from from the
1981a,
1981a, 1981b). aberrations isis less
aberrations less than or equal equal to to the
the spread
spread due due totoseeing.
seeing.
We
We can can compare
compare this this model
model of ofwavefront
wavefront distortions
distortions Over small
Over small sections
sections of of the
the aperture,
aperture, of size size r0,
r0, thethe wavefront
wavefront
with
with the "normally
"normally distributed surface" model
distributed surface" model (with (with a normal
normal aberration needs
aberration needs to be less than X/4. Across Across aa 10 10 meter
meter aper-
aper-
phase
phase distribution
distribution and and gaussian
gaussian correlation
correlation function)
function) used used in in ture thethe atmosphere
atmosphere introduces
introduces pathlength
pathlength fluctuations
fluctuations of 5 to
the sections above.
the sections Squaring the phase
above. Squaring phase difference
difference in EquationEquation 10 jjim
10 µm for for r07*0 between
between 20 20 and
and 10 10cm,cm,asascalculated
calculated from from
24
24 we we can
can write
write thethe averages
averages in in terms
terms of the the parameters
parameters o- or Equation 25.
Equation 25. Thus Thus aa surface
surface quality
quality better
better than than aa few few
and TT defined
defined for for the
the normally
normally distributed
distributed surface.
surface. microns will
microns will still
still allow
allow diffraction
diffraction limited
limited speckle
speckle analysis.
analysis.
This aberration criterion
This aberration criterionisis readily
readily met met byby thethe Ten Ten Meter
Meter
D(r) = 2o-2(1 -é r2I T2). (27) Telescope design.
The functional form
The functional form of of this
this"normally
"normally distributed
distributed surface"surface" The
The exposure
exposure times times used used ininspeckle
speckleinterferometry
interferometry
structure
structure function
function isis very very different
different from from the the atmospheric
atmospheric should
should be be short
short enough
enough to to freeze
freeze the theatmosphere
atmosphere associatedassociated
structure
structure function
function of EquationEquation 25. 25. As As aa consequence
consequence we we can-
can- image
image structure changes changes (typically
(typically 10 10-20
-20 milliseconds).
milliseconds). The The
not
not use the the equations
equations describing
describing the image image for for thethe normally
normally maximum exposure
maximum exposure time time isis order
order of the time for the wind wind to to
distributed
distributed surface
surface to to describe
describe the effects
effects of the atmosphere. move
move the the turbulence
turbulence pattern pattern aa distance
distance r0, r0, and
and thus thus it is
As alternative we
As an alternative we cancan gain
gain somesome qualitative
qualitative insightinsight independent of of telescope
telescope size.size. If the actuator
actuator movements are
into
into the effects
effects of the atmosphere by modeling modeling the wavefront wavefront small compared to
small compared the atmospheric
to the atmospheric fluctuations
fluctuations during during the the
distortions
distortions as as an array of small
an array small "segments"
"segments" of of size
size about
about r0 r0 same
same time, time, then the the active
active control
control of of the
the primary
primary will will not
with phase
with phase errors.
errors. Then we we can
can use the effectseffects of phase phase errors limit the resolution of of the
the speckle
speckle imaging.
imaging.
described by
described by Equations
Equations 21 21 and
and 2222 above.
above. We We firstfirst consider
consider Except during strong
Except during strong windswinds we we expect
expect the the actuator
actuator
visible light.
visible light. For distances separated
For distances separatedby by more
more than than r0 r0 the motion corrections on on the
the 10 10-20 millisecond time
-20 millisecond time scalescale toto be
be a
phase
phase differences
differencesare arelarge
largeenough
enoughthat that thethe first
first terms
terms of small
small fraction
fraction of of aa wavelength.
wavelength. With strong strong winds
winds the the faster
faster
Equations
Equations 21 21 and 22 22 are negligible.
negligible. Only Only the the diffraction
diffraction image image motion
motion of the the turbulence
turbulence patternpattern demands
demands shorter
shorter exposure
exposure
of the "segments"
"segments" described
described by second term
by the second term contributes.
contributes. times allowing higher
times thus allowing higher velocity
velocity actuator
actuator motions
motions (( as as are
are
Thus, as
Thus, as isis well
well known,
known, the the image
image size size is given by
is given by thethe anticipated
anticipated in in periods
periods of strong winds).
diffraction
diffractionfrom from cells
cells of size r0.
of size /Q. Therefore
Therefore the diffraction diffraction The maximum allowed allowed bandwidth
bandwidth is is given
given by by the atmo-
spikes
spikes on on the
the images
images of large large telescopes
telescopes do do not
not comecome from sphere induced phase
sphere induced phase difference
difference across
across the aperture, and so so itit
the overall aperture.
the overall aperture. They They instead
instead arisearise fromfrom the the fact
fact that isis related
related to telescope
telescope size. Thisjnaximum
size. This maximum bandwidthbandwidth is is pro-
pro-
some
some of the small small "segments"
"segments" have have one of the the secondary
secondary sup- portional to (telescopediameter)~5
portional (telescopediameter) -5 ' 6 and and for the the Ten Ten Meter
Meter
port struts
port struts across
across them.them. The The image
image isis approximately
approximately an an Telescope it is 56%
Telescope 56% of that allowedallowed withwith a 5 meter telescope.telescope.
incoherent sum of of "segments"
"segments" not crossed crossed by by struts
struts and and"seg-"seg-
ments" that are.
ments" are. ForFor the the hexagonal
hexagonal segmented
segmented geometry geometry of
Figure 11 the struts will
Figure will be at threethree angles
angles and and thusthus givegive aa six
six
spike diffraction image. image. Acknowledgments
At a wavelength
At wavelength of of 1010 //,m
µm thethe cell
cell size
size r0 r0 isis 36
36 times
times We gratefully
We gratefully acknowledge
acknowledge the very helpful
the very helpful comments
comments
larger than
larger than at pm. Thus
0.5 µm.
at 0.5 Thus for forconditions
conditions with with visible
visible light
light and suggestions of David Cudaback, Sandra Faber, and
and Roger
Roger
cell
cell sizes
sizes ofof 10 10 -- 20
20 cm the 10µm 10/xm cell cell sizes
sizes will
will be 3.6 3.6 -- 7.2
7.2 Ulrich.
meters.
meters. We We expect
expect the the image
image willwill be a result of of thethe combined
combined
effects
effects of of the mirror segment
the mirror segment pistonpiston errors
errors and the the atmos-
atmos-
pheric
pheric cell cell phase
phase errors.
errors. However,
However, the the optical
optical pathlength
pathlength
errors due
errors due to to piston
piston errors
errors (about
(about 0.42 0.42 µm)/xm) will will be be small
small
compared to
compared to the atmospheric
atmospheric pathlength
pathlength errors errors (about 4.2 /xm µm References
if wewe evaluate
evaluate Equation
Equation 25 25atat r=r$).
r= r0). For For thetheexcellent
excellent seeing
seeing
conditions at
conditions at the high altitude locations of large telescopes telescopes the Beckmann, P.
Beckmann, and Spizzichino,
P. and Spizzichino, A. The Scattering
A. ""The Scattering of
of Elec-
Elec­
coherence cell
coherence cell size
size may
may be be 20
20 cmcm or more more for for visible
visible light,
light, tromagnetic Waves
tromagnetic Waves from
fromRough
Rough Surfaces.
Surfaces."" Macmillan, 1963.
1963.
and
and thusthus 77 meters
meters or or more
more at at 10
10 µm,
/u,m, i.e.
i.e. close
close to the the full
full
aperture. For those conditions
For those conditions the the full
full aperture
aperturediffraction
diffraction Born, M.
Born, M. and Wolf, E. Principles
Wolf, E. Principles of Optics, 6th
of Optics, 6th edition
edition Per
Per--
pattern described in in section
section 22 above
abovewill willagain
againapply.
apply. gamon Press (1980).
(1980).
Brown,
Brown, D. S. "" The
D. S. The Imaging Potential of
Imaging Potential ofLarge
LargeGround
Ground-based
-based
Infrared Telescopes " "Proceedings
Infrared Telescopes Proceedingsofof the Peak National
the Kitt Peak National

128 //SPIE
SPIEVol.
Vol. 332
332Advanced
AdvancedTechnology
TechnologyOptical
OpticalTelescopes
Telescopes (1982)
Observatory Conference
Observatory Conference on Optical
Optical and Infrared
Infrared Telescopes
Telescopes
for the 1990's
1990's (May,
(May, 1980,
1980, Tucson, Arizona)
Arizona)
Dainty, J.J. C. " TelescopeRequirements
C."Telescope Requirementsfor
forSpeckle
SpeckleInterferometry
Interferometry
ESO Conference:
"" ESO Telescopes of
Conference: Optical Telescopes of the
the Future,
(Geneva, 12 12-15
-15 December 1977 ed. Pacini,
December 1977 Pacini, et
et al.
al.
Fried, D.L. "Optical
Fried, D.L. "Optical Resolution Through aa Randomly
Resolution Through Randomly Homo-
Homo­
geneous
geneous Medium
Mediumfor Very Long
for Very Long and
and Very
Very Short
Short Exposures
Exposures ""
J.O.S.A. 56, 1372
J.O.S.A. 56, 1372 (1966)
(1966)
Gabor, G. "DisplacementSensors
G. "Displacement Sensorsand
andActuators
Actuators Needed
Needed to
to Con-
Con­
trol
trol aa Segmented
SegmentedPrimary
PrimaryMirror " Proceedings
Mirror" Proceedingsofofthe
the Kitt
Kitt Peak
Peak
National Observatory
National Observatory Conference
Conference on on Optical
Optical and Infrared
Infrared
Telescopes
Telescopes for
for the 1990's (May,
the 1990's (May, 1980,
1980, Tucson, Arizona)
Tucson, Arizona)
Hufnagel, R.E. and Stanley,
Hufnagel, R.E. Stanley, N.R. "ModulationTransfer
N.R. "Modulation Transfer Func-
Func­
tion Associated
tion Associated with
with Image Transmission through
Image Transmission through Turbulent
Turbulent Media
Media
""J.O.S.A. 54,52
J.O.S.A. 54, (1954)
52 (1954)
Mast, T.S.
Mast, T.S. and
and Nelson,
Nelson, J.E. "Figure Control
J.E. "Figure Control for
for aa Fully
Fully Seg-
Seg­
mented
mented Telescope
TelescopeMirror"
Mirror"(1982)
(1982)Ten
Ten Meter
Meter Telescope
Telescope Report
#64,
#64, submitted
submitted to
to Applied
Applied Optics
Optics
Nelson, J.J. E. "Segmented
Nelson, "Segmented Mirror
Mirror Design
Design for
for aa 10
10-Meter Tele­
-Meter Tele-
scope
scope"" Society
Society of Photo -Optical Instrumentation
Photo-Optical Instrumentation Engineers
Engineers
Proceedings, Volume 172
Proceedings, 172 (Jan
(Jan 1979)
1979)
Nelson, J.
Nelson, J. E. "" The
The University
Universityof
of California Ten Meter
California Ten Meter Telescope
Telescope
Project
Project -- The
The Segmented
Segmented Design
Design" "Proceedings
Proceedings ofof Kitt Peak
Peak x,uu
x,
National Observatory
National Observatory Conference
Conference on on Optical
Optical and
and Infrared
Infrared
Telescopes for
Telescopes for the (May, 1980,
the 1990's (May, 1980, Tucson, Arizona)
Arizona)
Nisenson, P.,
P., R.
R. Stachnik,
Stachnik,and
andC.C.Papaliolios "HighResolution
Papaliolios"High Resolution
Imaging at
Imaging Large Telescopes
at Large Telescopes " "Kitt
Kitt Peak
Peak National
National Observatory
Conference
Conference on Optical and Infrared Telescopes
Optical and for the 1990s
Telescopes for 1990s
(May,
(May, 1980, Tucson, Arizona)
1980, Tucson,
Walters, D.L., D.
Walters, D.L., D. L.
L. Favier,
Favier, and
and J.J.R.R.Hines " Vertical Path
Mines" Vertical
AtmosphericMTFMeasurements"
Atmospheric MTF Measurements"J.O.S.A.
J.O.S.A. 69, 828 (1979)
(1979)
Walters, D.L. "" The
Walters, The Atmospheric
AtmosphericMTF
MTFforfor Desert
Desert and Mountain
and Mountain XBL 822-145

Locations
Locations-- r0r0 Measurements"
Measurements"J.O.S.A. (1981a)
J.O.S.A. 71, 397 (1981a) 1.
1. The
The geometry
geometry of the primary
primary mirror showing
showing the
the 36
36 hexag-
hexag-
onal
onal segments.
segments. The
The mirror
mirror plane
plane coordinates are (x,y)
coordinates are (x,y) and
and
Walters, and K.
Walters, D.L. and K. E.E.Kunkel "Atmospheric MTF
Kunkel"Atmospheric MTF for
for Desert
Desert the
the image
image plane
plane coordinates are cartesian
coordinates are cartesian (u,v)
(u,v) and
and polar
polar
and Mountain Locations
and Mountain Locations -- The Atmospheric Effects
The Atmospheric Effects on /Q ""
on r0 (B,¢).
J.O.S.A. 71, 406
J.O.S.A. 71, 406 (1981b)
(1981b)

SPIE Vol. 332

SPIE Vol. 332 Advanced
Advanced Technology
Technology Optical
Optical Telescopes
Telescopes(1982)
(1982)// 129
129
 o
a)

N
hexagon, <f>-0°
Single hexagon,
Single 4 =0° Single hexagon, sb=30° a)

a=10P.m
-2 -2 X=10µm
a)
-3 c -3
a
u 'o
g
o -4
ó
g

O
t
-4
-4

-5
-5 Vo -5 . /

Jó -6
,

g2 -6
s.,
J
Ó
-7 -7 _

-8
-8 -8
I

-9-
-9 -9
10
-100 0.5 1.0
1.0 0) 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
50
1.5
1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
0o 0
Single hexagon, =0° b) Single hexagon
hexagon,, 43 =30° b)
b)
-I

-2 a=IOµm -2
-2
-2 a=10µm
o> -3
-3
^ -33
~ CJ
C
^

O
g -44
_ -4
-4 "f
73
O '1,R11)

t -5 -5
- Y r
2 -6 . --6
6 ,
q -7
-7
-7 -7 TI n

Il I,
I

-8
-8 _
--8
Y I
8
1

-9
-9 --9
9

--100
|C\ 10
10 15
15 20 25 30 35 40 45 50
-10
IO
0D
.

5 10 15
15 20 25 30 35 40 45
V

50
50

0 0
Single hexagon,
Single hexagon, </>
0 ==0° c) Single hexagon , iiti =300 c)
-I

-2 -2 X=10µm
-3
a,

I
O
-5

-6
o+
-7 J -7
-8 -8

-9 -9
ti
-10 + ,1i' ..
-10 -10
0 50 100 150 200
200 250
250 300
300 350 400 450
350 4OO 450 500
500 0 50 100 150
150 200 250
200 250 300
300 350 400 450
350 4OO 450 500
500
Image radius 98 (arc
Image radius (arc seconds)
seconds) radius B
Image radius
Image 8 (arc seconds)
seconds)

XBL822 -4460
XBL822-4460 XBL822-4461
XBL822-446I

2. intensity of
2. The intensity of the
the diffraction
diffraction pattern
pattern along
along the
theu u-axis
-axis 3.
3. intensity of the diffraction
The intensity diffraction pattern
pattern along
along aa line
line at
at 30°
from single hexagon.
from a single hexagon is
hexagon. The hexagon is 0.9
0.9 m on aa side
side and
and from the uu-axis
from -axis for
for aa single
single hexagon. hexagon is
hexagon. The hexagon is 0.9
0.9 m
plotted for a wavelength
the pattern is plotted wavelength of of 10
10 /u,m.
µm. on
on a side
side and
and the
the pattern
pattern isis plotted
plotted for
for aawavelength
wavelength of 10 10
µm.

730 SPIE Vol.

130 //SPIE Vol. 332
332 Advanced
AdvancedTechnology
Technology Optical
OpticalTelescopes
Telescopes (1982)
(1982)
 u \
0 0
-1 \ Full array, 0.0°

\
<£=0° a) Full array, 0=30°
-2
-2 xX=10µm
= 0/xm -2 a=10µm
-3
c
r\ -3
o A
TD -4
o
ffi "I AA 1 'O
6 -4

o
-5
II
A
A if
I
o
-5
-6
-6 V
A A
i Aifl -6
Jo
If
I I n

-7
-7 ,
I
,I -7
-8
-8 _
1

1 I
-8
-9
-9
-9
-10 -10
100 111
0 0.5
0.5 .0
1.0 1.5
1.5 2.0 2.5 3.0 35
3.5 4.0 4.5 55.0 0.5 1.0 1.5
1.5 20
2.0 2.5
2.5 3.0 3.5
3.5 4.0 4.5 5.0
5.0
o 0

Full array, 0=0° b) Full array, ck =30° b)

-2 a=10µm
X = 10µm
v -3
c
v -4
°
o
-5
o -6
J -7
I

i ,

li
-8 ¡
I I

-9
- 10 L
0 5 10
10 15
15 20 25 30 35 40 45 50
5 5 10
10 15 20 25 30 35 40 45 50
0
'
I

- Full array, ck =0° c) Full array, 4 =30° C)

-2 X = 10µm
X = 10 µm
7L;
-3
Cl
c
o -4
vo

g -6
I

-8 I
i
II

-9
-10 . , .
-10
o 50 IOO
100 ISO
150 200 250
200 250 300
300 350 400 450
350 4OO 450 500
500 0 50 100 150 200
150 200 250
250 300
300 350 400 450
350 400 450 500
500
Image radius 9
Image radius (arc seconds)
8 (arc seconds) Image radius 9
Image radius B (arc
(arc seconds)
seconds)

X8L882 -4462
XBL882-4462
X8L882-4463
XBL882-4463
4. The
4. The intensity
intensity of the
the diffraction
diffraction pattern
pattern along
along the
the uu-axis
-axis
full array
array of hexagons
hexagons plotted
plotted for
for aa wavelength
wavelength ofof 10
10 5.
5. The intensity of the
the diffraction
diffraction pattern
pattern along
along aa line
line at 30°
at 30°
from the full from the
from the uu-axis
-axis for
for the full
full array
µm. array of hexagons plotted
of hexagons plotted for
for aa
wavelength
wavelength of 10 /x,m.
µm.

SPIE Vol.
SPIE Vol. 332
332 Advanced Technology Optical
Advanced Technology Telescopes (1982)
Optical Telescopes (1982)// 131
131
 o 0.15
Full array,/1)=0°
Full array,</> = 0° a)
without intersegment
without intersegment cracks
cracks X = 0.5/xm
-2 Q- O.I arc seconds,
a=10/1m
-3
,,

-4 I \r\(I/r\\\\(\1
0.10
-5 E

-6 "b

-7

-8 0.05

-9

10 I

0) 0.5 1.0
1.0 1.5
1.5 2.0 2.5 3.0 3.5
3. 5 4.0 4.5 5.0
5

0
Full array,
Full array,<£
0 == 0° b)
without intersegment
intersegment cracks
cracks 0.5 1.0 1.5
-2 X=10µm (meters)
T (meters!
-3 7.
7. The rms surface
rms surface roughness
roughness (o-)(cr) -- correlation length ((T)
correlation length T)
plane showing
plane showing the of the
the limits of the regions
regions which
which satisfy
satisfy various
various
-4 V

optical
optical goals
goalsatataawavelength
wavelengthofof0.5 0.5/x,m.
µm. Values
Values ofof cr
o- and
and T
-5 to lower right
to the lower right of the curves
curves give
give >^ the
theindicated
indicated fraction
fraction
of energy in a circle
circle of radius 0.1
0.1 arc seconds.
seconds.
I

1 I

-6 I '
I'

-7 1

-8 A == IOµm 0.5
e'= 0.2 arc
0'=0.2 arc seconds
seconds 0.6
-9
I l'
-10 5 10
IO 15
15 20 25 30 35 40 45 50
51 0.7
0
Full array,
Full array,sb= 0° c) Ì

0.8
without intersegment
without intersegment cracks
cracks
E
-2 a=10µm
-3 b
-4
-5
-6
-7
-8
-9

-10
"0 50 100 200 250
150 200 250 300
300 350 400 450
350 400 450 500
500
Image radius
Image radius BQ (arc seconds)
seconds) 5 10
10

X8L882 - 4464
XBL882-4464
meters)
T ((meters)
8. The
8. The rms
rms surface
surface roughness
roughness (Q)
(cr) -- correlation
correlation length
length (T)
6.
6. The intensity
intensity of the
the diffraction
diffraction pattern
pattern along
along the
the uu-axis
-axis for plane
plane showing
showing the limits of the
the limits the regions
regions which
which satisfy
satisfy various
various
full array with no cracks
the full cracks between the segments
segments plotted
plotted for
for optical goals
optical goalsatataawavelength
wavelengthofof10.0 10.0/xm.
µm. Values
Valuesof and TT
ofcra and
wavelength of 10
a wavelength 10 pm.
µm. the lower
to the lower right of the
the curves give >^the
curves give theindicated
indicated fraction
fraction
of energy in aa circle
circle of radius
radius 0.2
0.2 arc
arc seconds.
seconds.

SPIEVol.
732 //SPIE
132 Vol. 332
332Advanced
AdvancedTechnology
TechnologyOptical
OpticalTelescopes
Telescopes (1982)
(1982)
 1
1 I TÌT 1 I I 1 1 1 1 1 1_
- lO.Oc
10.0 1 i l 1 1 1 111 1 1 1 111111 1 I 1 111I`
-0
C
C
E ,-5/6 m o
N
o
Uv_
ó
J3
1.0µm
1 arc second
I arc second seeing
or
cr li
visible light
for visible
1.0
1.0 0.5p.m
0.1r-
w
2
Q
Q
Ó
LU
w 0.1p.m
o
2~
0.01
_ -r O.I
0.1
O.Op m

o
tr
a:
LU
w
- zw
ó
o
oo
co
0.01 1 I I 1111 1 1 111111 1 1 11111,
1 111_ 111 1.0
0.001
0.001 L I 1 1 1 1 1 1 1
O.I
0.1 1.0 10.0
0.001 0.01
0.01 O.I
0.1
X(Fm)
9e (arc seconds)
XBL 819-1867
seconds)
11. The
11. image diameter
The image diameter (80 %)for
(80%) various rms
for various piston errors
rms piston errors
X8 322 146
wavelength. As
as a function of wavelength. thewavelength
As the increases the
wavelength increases
image diameter
image changes from
diameter changes from the single segment limit
single segment (rms
limit (rms
9. excluded energy
9. The excluded radius for
image radius
energy versus image diffraction
for diffraction 0.0).
«* 0.0).
patterns from
patterns circularapertures
from circular 5/6 m
and 5/6
5mand
aperturesofof5m m (a 0.5
(X=0.5 = infinity) to the
infinity) to full primary
the full limit (rms
primary limit (rms =
/am).
am).

A/100
X/100 A/50
X/50 A/20
X/20 A/10 A/5 A/2
X/2

a-
yy
x a L 822.147

10. The
10. "radius" kRO
image "radius"
The image versus rms
kRO versus rms piston which
error which
piston error
80%and
contains 80%
contains 90%of
and 90% image energy.
the image
of the small piston
energy. For small
angles are
errors the angles
errors are those of the the diffraction limit for
diffraction limit for the full
primary.
primary. For large piston errors the
For large correspond to the
angles correspond
the angles
diffraction limit
diffraction single segment.
limit of aa single

SPIE
SP 332 Advanced
Vol.332
/EVol. Advanced Technology (1982) // 133
Telescopes (1982)
Optical Telescopes
Technology Optical 133