(Sanitized) QUARTERLY STATUS REPORT NO. 5 JUNE 10, 1964 - JANUARY 1 1966

• Document Type: CREST
• Collection: NGA Records (Formerly NIMA)
• Document Number (FOIA) /ESDN (CREST): CIA-RDP78B04770A001500060005-6
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 19
• Document Creation Date: December 28, 2016
• Document Release Date: September 26, 2005
• Sequence Number: 5
• Publication Date: January 1, 1966
• Content Type: PERRPT

Approved For R&ease 2005/11/21 / Cl
Quwterly status Report No a 5
June 10, 196 , a Januwary 1, 1966
.ASpH IC OPTICAL SYSTEMS
rove Revieapemp a eor Release 2005/11/21 : CIA-RDP78BO477OA001500060005-6
~
770' 1~50
66U0
 Approved For Pease 2005/11/21 : CIA-RDP78B047704&01500060005-6
During this period we discontinued work for several
months for the following reasons.
(1) Our preliminary studies of aspheric optical designs
were disappointing. It became clear to us that our optical
design programs were inadequate to cope with aspheric optical
systems. We did not wish to run the chance of drawing
incorrect conclusions, so we have spent our time trying to
improve our program. We now have a new program and have
been able to use a new program written by
Design
work has now been resumed. This report describr-s our latest
efforts.
(2) We carried on an investigation of using thin films
to coat aspheric surfaces. This work was carried far eziough
to show praetite but it was clear that we could not, under
this contract, make much more of a contribution without
extensive expenditures for automatic control systems, for
coating. We have therefore discontinued further work in
this area until we are-able to show more positive gain
in using aspheric surfaces. Part II of this report is a
au mary of the work done on coating.
Approved For Release 2005/11/21 : CIA-RDP78B04770A001500060005-6
 Approved For F gJease 2005/11/21 : CIA-RDP78B04770,Q 01500060005-6
P art I
The Design of Aspheric Surfaces
Aspheric surfaces have been used extensively in optical
instruments with varying degrees of success. They have
been used most successfully in astroncseical systems. The
aspherio is usually placed on a single surface close to the
aperture stop and is used primarily to correct spherical
aberration. Few attempts have been made to study the use
of several aspheric surfaces in optical systems which
must cover substantial fields of view. There have been in
the past two good reasons why these studies have not boon
made. First, aspheric surfaces are difficult to make, and
second, the design problem becomes nauah more difficult,
Now that we have large computers available it should be
possible to study the value of using aspherie surfaces
more extensively.
We have attempted to do this by selecting a well
known. photographic lens and introducing aspheric surfaces.
In order to evaluate the gain in using aspheric su "
faces we have designed an optimum series of all spherical
lenses to use as comparison.
The preliminary results of this utudy were reported
to the Tokyo meeting of the International Commission on
Optics. A copy of this paper is included in Appendix 1.
In the paper we described six lenses which were-all
designed to the same specifications. A triplet objective
(See Table I and Fig. 1 of Appendix 1) was compared with
Approved For Release 2005/11/21 : CIA-RDP78B04770A001500060005-6
 Approved For Rease 2005/11/21 : CIA-RDP78B04770WWp1500060005-6
a triplet with four *spheric surfaces (See Table V and Fig.
5). The objective with the aspheric surfaces was disappoint-
ing, for the imagery was not as good as the all spherical
lens. The only advantage in the use of the aspherics a-
peared to be in the size of the lens. With the aspheric
surfaces the objective was reduced to an overall length of
4.79 inches, while without aspheric surfaces the overall.
length was 8.76 inches. In this paper we commented that we
were not confident that we had optimised the aspheric triplet.
The design problem was much more difficult than we had
anticipated. We would like to comment an the problem of
designing lenses with aspheric surfaces using our present
techniques.
Design Problems in Aspheric Lens Systems.
We designed the aspheric lens system using the semi-
automatic design program called "Ordeals". A manual for
this program is included in Appendix II. Ordeals designs
lenses using third and fifth order aberrations, and eval.u-
ates by ray tracing. It was found that when general aspheric
surfaces were introduced that the program could find a large
variety of solutions which were corrected for third and
fifth order aberrations but most of the solutions balanced
out large fifth order aberrations thereby introducing higher
order aberrations. In order to reduce the high order aber-
rations it was necessary to not only balance out the total
fifth order aberrations but the individual fifth order our-
~' ~" ~e ~0tQ/1 ~1 ~3- r-F W7`s'B? 7 b'PS0060 t6-d'
 Approved For Re,W'ase 2005/11/21 : CIA-RDP78BO477OAQ01500060005-6
this we decided to introduce a feature in Ordeals called
RAYDEV which would enable it to correct a fan of meridional
rays. This was not a completely sufficient procedure but
it was about the only procedure that could be fitted into
the 10R storage of the 7071j. machine. With this feature we
were able to design the lens described in the paper above.
The procedure was to first use only loth and 6th order deforma
tion terms to bring the solution into a region of solution
and then allow the 8th and 10th order terms to vary to clean
up small residuals. The procedure very definitely decreases
the effectiveness of using the high order terms. The final
meridional ray curves are shown in Pig. 1. The meaning of
these curves is discussed on pages 97-98-99 of the Ordeals
manual. These curves show that the wave surfaces are
smoothly varying because the third and fifth order aberra-
tions have been corrected to small values. The curves do
show, however, a high. order negative astigmatism (this is
indicated by noticing the downward slope of the curve on
the top left hand side of Fig. 1). We were not able to
improve the lens much further with Ordeals, so we deferred
further work on the problem until we could use our new
program called Flair.
Flair is a program written fo.' a much larger machine
and it corrects on the basis of ray tracing and third order
or exclusively with ray tracing. Up until very receftly
we have not had Flair working properly on spherical sur-
faces, so it was not possible to use it on aspherio sur-
faces. Now Flair is working well on spherical surfaces
Approved For Release 2005/11/21 : CIA-RDP78BO477OA001500060005-6
 Approved For ReLease 2005/11/21 : CIA-RDP78B04770AQ01500060005-6
MERIDIONAL FANS SKEW FANS
H=I
H'= O
FIG. I
0 10
Approved For Release 2005/11/21 : CIA-RDP78B04770A001500060005-6
 Approved For Rase 2005/11/21 : CIA-RDP78B047704 41500060005-6
so w3 intend to use it on the aspheric lens problem.
In the meantime we have had the opportunity to use a
program written by This program was prepared
for the I.B.M. 7094 model 2. We have been able to use no
Grey program on one of the machines at the White Sands
Missile range thru the courtesy of
We have not up the aspheric triplet
problem on tht machine and have found several very inter-
esting solutions. The meridional plots for one of the best
solutions is shown in Fig. 2. These curves are rippled but
the overall straightness is a great deal better than in Fig.
1. The ripples in these curves result in the balance of
large high order aberrations on the aspheric surfaces. This
balance is shown very clearly in the aberration curve for
the central image. This is shown enlarged in Fig. 3. This
curve looks bad but one must actually perform an optical
path calculation to evaluats it properly.
The energy distribution curves for this lens are shown
in Fig. 4. The energy distribution curves for the triplet
designed with the Ordeals program em shown in Fig. 5.
The new design is definitely better than the original
triplet.. It is clear that a large design program cam
do a better job of correcting a system with aspherlcs than
a semi automatic program like Ordeals. From the work we
have done so far we can see that there is a wide variety of
solutions possible and that we are by no means certain that
the solution shown is optimum. It is necessary to learn
Approved For Release 2005/11/21 : CIA-RDP78B04770A001500060005-6
 Approved For R?alease 2005/11/21 : CIA-RDP78BO477OA901500060005-6
MERIDIONAL RAYS
I-I ? 2/3
SKEW RAYS
H s O _?__~~~.~..',~~
FI G. 2
Approved For Release 2005/11/21 : ,CIA-RDP78B04770A001500060005-6
 Approved For Release 2005/11/21 : CIA-RDP78BO477OA001500060005-6
F.I G. 3
w
c -.002
w
U,
z
4
!-
6
10
SEMI APERTURE finches)
Approved For Release 2005/11/21 : CIA-RDP78BO4770AP01500060005-6
1 Vw1he
.15
2.0
 Approved For Release 2005/11/21 : CIA-RDP78BO4770A0001500060005-6
FIG. 4
4
8 12 16 20
SPOT DIAMETER (inches)
24
Approved For Release 2005/11/21 : CIA-RDP78BO477OA001500060005-6
 Approved For Release 2005/11/21: CIA-RDP78B04770SO1500060005-6
FIG. 5
100
80
C7
Cr.
z 60
w
F-
z
w 40
w
a.
20
5
10 15 20 25X IO-3
SPOT DIAMETER (inches)
(a)
Approved For Release 2005/11/21: CIA-RDP78BO477OA001500060005-6
 Approved For Rease 2005/11/21 : CIA-RDP78BO4770AW1500060005-6
how to design with these large programs. The solution s":gown
is one of many we have obtained. As long as we arrive at
different solutions depending on bow we weight aberratioas
and select rays there is the possibility that we are not at
an optimum.
This study has shown that it is essential to have a
large computer program to adequately study the use of aspheric*b.
A designer simply can not cope with all the independent vari..
ables.
We have shown an improved design of a triplet objective
using aapherie surfaces. The objective is smaller in site
and is corrected better. It is doubtful if the improtement
is sufficient to justify the expense of the four aspherios,
but the design illustrates interesting possibilities for the
future. It is necessary to have a large computer to
thoroughly analyse these problems. We do not believe any
programs are yet quite adequate so we are actively-trying
to improve our program to be more effective on aspheric
surfaces.
Approved For Release 2005/11/21 : CIA-RDP78BO477OA001500060005-6
 Approved For Rase 2005/11/21 : CIA-RDP78B04770A1500060005-6
List of Illustrations
Fig. 1. Ray plots for Aspherio Triplet design on Ordeals.
Fig. 2. Ray plots for ga Triplet designed on the Program. STAT
Fig. 3. A sealed up plot of the axial spherical aberration.
Fig. 1i.. Energy Distributions for New Aspheric Triplet
Designed on the
Program.
Fig. 5. Energy Distribution for Aspheric Triplet Desigreii
on Ordeals.
Approved For Release 2005/11/21 : CIA-RDP78BO477OA001500060005-6
 Approved For Re ase 2005/11/21 : CIA-RDP78B04770AW1500060005-6
8
Part II
Summary of progress to date on project to aspherize optical
surfaces by evaporating films in a vacuum,
EVAPORATION OF
OXIDES OF SILICON: Although other workers have used films
of LiFl, ZnS2, and other materials for asperhizing, we. have
directed our efforts towards the evaporation of the oxides
of silicon. The advantages of such films are they are durable
and hard. They can be cleaned in water without injurin
theme
OPAQUE AND
TRANSPARENT
FILMS:
There are two possible users of such a
film: (a) It is used to aspherize a reflecting surface.
In this case, the film need not be transparent, but only
durable and smooth. (2) The film is employed to aspherise
a transparent optical surface. In this latter case two more
stringent requirements should be met:
(a)
The film should
have a negligible optical absorption.
(b)
The film should
have a refractive index close to that of the glass substrate.
Any appreciable "mismatch" between the refractive index of
the film and the refractive index of the substrate would re-
sult in an additional "Fresnel loss" due to the reflection
of light at the interface.
MEASUREMENT OF
OPTICAL PROPERTIES OF
SILICON OXIDE FILMS: The films were deposited by the evapo-
>, ~ti cxCppe~veo ~ixcc tea~g DO#d7 /2 h1@IA- 18 # I SOt T60b 6 eme t
 Approved For ReIase 2005/11/21 : CIA-RDP78BO4770AW1500060005-6
9
Company. As Ritter3 has shown, the evaporation of SiO ;yan
result in SiOA 5120. , or SiO 2. The refractive index of the
films which we evaporated was measured by the Abe ll s4 method.
The index of six different films was measured by this nv ans
and an index was obtained which varied from 1.49 to 1.51
for the individual films. The evaporation took place at a
pressure of approximately 10'*4 tors and the films had an.
optical thickness of approximately .75 waves at A5461A. Since
we ascertained that the refractive index could be reproduced
within reasonable limits, we proceeded to see how thick a
film we could deposit., Si,O films can be distinguished
G- 3
from SiO and Si02 by absorption bands near 12--""( in the in-
frared. Although we did not measure the infrared absorption
of the films to confirm that the films were indeed Si20`, ,
from the work of Bradford and Rass5, and Ritter3 it is
reasonably certain that this is indeed the composition of
the films which we are producing. In the remainder of this
paper we shall refer to tht! films of Si.103 as "silicon
sesquioxide".
IEPOSI TION OF THICK
FILMS: The thick films of silicon sesquioxide
were evaporated in a stainless steel vacuum chamber which was
pumped by 600 litre/second diffusion pump which is baffled by
liquid nitrogen cooled trap. A servo controlled needle valve
in the top of the chamber enables one to bleed in gas and
thus maintain the pressure of the chamber at any value from
Approved For Release 2005/11/21 : CIA-RDP78BO477OA001500060005-6
 Approved For R@,lpase 2005/11/21 : CIA-RDP78B04770AQ1500060005-6
1,0
10`5 torr to 5 x 10-4 torr. An optical monitoring system
was used to measure the optical thickness of the film as
it was deposited.
OPTICAL MONITORING
SYSTEM: The optical monitoring system consisted
of a tunsten lamp which was imaged upon the glass monitoring
near the roof of the vacuum chamber. The reflected light
was thence imaged on a photomultiplier in the base of the
chamber. The photomultiplier was filtered with a silver-
dielectric-silver wedge interference filter which had been
previously calibrated with a mercury lamp. The output tf
the photomultiplier was fed into a RCA model microammeter.
The optical thickness of the film was measured by keeping,
track of the number of successive maxima and minima of the
photocell current.
CONTAINER FOR
EVAPORATING SiO: Two types of electrically-boated con-
tainers were used to evaporate the Sion One type is de.
signed by Drumheller and is manufactured by the Allen-Jones
Company. We found that this type was satisfactory, but that
eventually the central heater in the boat burns out. Another
type of container is the "baffled box type" manufactured by
the Mathis Company. Although this tantalum does not burn
out, the cover warps and when one applied pressure to remove
it, the tantalum would sometimes crack. We found the latter
type of boat preferable.
Approved For Release 2005/11/21 : CIA-RDP78BO477OA001500060005-6
 Approved For Rolpaase 2005/11/21 : CIA-RDP78B04770A~1500060005-6
CONTROL OF 02
PRESSURE: The refractive index and optical ab-
sorption of the silicon sesquioxide films depends markedly
on the partial pressure of the oxygen during evaporation and
to a lesser extent on the rate of evaporation. A servo.
controlled needle valve was used to regulate the partial
pressure of the oxygen during the evaporation. The entire
chamber was pumped to a pressure of less than 10`5 tort
in order to outgas the walls and remove the water vapor,
which is the predominate residual gas at such pressure"
Oxygen was then bled into the chamber, An oxygen partial
pressure of 2 x 10-' was the highest partial pressure which
was used. The oxygen was admitted at the top of the chamber.
No attempt was made to direct the oxygen at the substrate.
THICKEST FELM
DE?OSIT8D: The thickest film which has been dc.
posited to date has an optical thickness of ten waves at
5400A: The film is quite transparent, durable, and adheres
tenaciously to the substrate. The optical density of this
film is .08 which corresponds to a transmittance of 83%. This
transmittance would of course increase if both sides of the
glass plate were anti-reflected. However, this small residual
absorption is serious. It means that if the film were made
thicker, the transmittance would decrease below tolerable
limits. Thus we have initiated a program to bleach the films
Approved For Release 2005/11/21 : CIA-RDP78BO477OA001500060005-6
 Approved For Rglpase 2005/11/21 : CIA-RDP78B04770491500060005-6
BLEACHING OF
SiO FILMS: Bradford and Hass5 have shown that films
of S12O3 are bleached by exposing them to strong U.V. light
in an atmosphere of oxygen. The films they investigated
were a half-wave in optical thickness and were rendered
relatively transparent by this bleaching process in the
spectral region from 25OOA to 7000A. 'However, the films
which we are depositing are twenty times as thick as those
deposited by Bradford and Hass. In order to test the bleach-
ing effect on thick films, we deposited a film which was
12 waves in optical thickness at 51i.OOA. This film was
yellow looking in appearance due to the absorption in the
blue. This film was bleached for four hours under a Hanovia
Analytical Model Quartz Lamp, Model 7120, 435 watts. The
transmittance of the film in the blue spectral region im-
proved markedly. For example, the bleaching caused the
optical density at 4000A to decrease from 1.45 to .75. At
)4500A the optical density decreased from .52 to .3. Although
an improvement is achieved, the film is somewhat yellow
and additional bleaching does not produce any marked improve-
ment. Bradford and Hass conjecture that the exposure to the
U.V. removes dislocations and produces better-defined stoichio-
metric order in the film. It is also possible that oxygen
diffuses into the film. It is intended to investigate this
question by bleaching the film while it is vacuo.
Approved For Release 2005/11/21 : CIA-RDP78B04770A001500060005-6
 Approved For Ruse 2005/11/21 : CIA-RDP78B04770A1500060005-6
113
PERSONNEL: The evaporations of silicon sesquioxide
films described herein were done by
furnished assistance and guidance.
i~1NCI S:
1. L. G. Schultz, J. Opt. Soc. Am. , L.32 (1948)
2. J. A. Dobrowolski, Thesis, Imperial College of Science
and Technology, University of London, 195.5(unpublished)
3. 1. Ritter, Opt. Acta 9, 197 (1962)
ll.. F. Abele: Progress in 0 t~~ics__ (E. Wolf, :editor) Vol. 2
u
p. 257 North-Holland P
lisping Co. (1963)
A. P. Bradford and George Hasa, J. Opt. Soo. Am. , 1096
(1963)
Approved For Release 2005/11/21 : CIA-RDP78B04770A001500060005-6