The method to be described covers manufacture of test plates and optical components, centring and spacing, and it produces completed systems of the utmost precision in agreement. with the computed design. Some systems successfully built and used for research purposes will be mentioned.

Multi-element catadioptric systems are aligned after assembly by adjusting a few elements using data from interferograms taken on-axis and at four full-field locations. This method, which is essentially aberration balancing, has proven to be superior to the more traditional method where the alignment of each component is verified as it is added to the system.

A conventional laser beam expander consisting of confocal parabolic mirrors is quite sensitive to misalignment of its components. Beam steering by pivoting the small mirror about its "neutral point" gives a rather limited steering range. The ideal beam expander would be insensitive to misalignment of its components and would also provide substantial amounts of beam steering. At the same time it would be simple and have a substantial field view. This paper describes a two-element system with all these properties, and compares its features with a confocal parabola system counterpart.

An important feature of the diamond-machining process is the ability to accomplish both high-precision machining and optical finishing on the same fixture and to like tolerances. Thus highly accurate mechanical relationships can be built into an optical component which, while increasing part and fixturing complexity, provide a much simpler overall system design. It is this reduction in system complexity as it relates to system assembly and alignment, that provides the subject for this paper. Types of features that can be introduced, their typical accuracies and examples of their use will be presented.

The misalignment of the secondary and primary mirrors, in a full aperture Cassegrain test, introduces coma and astigmatism into the wavefront. Equations were derived which relate the magnitude and orientations of these two aberrations to the tilt and decenter of one mirror with respect to the other when tested in an autocollimating mode. By determining the coma and astigmatism, introduced into the wavefront, using Zernike circle polynomial fitting techniques, the tilt and decenter correction can be computed which would optically align the secondary to primary. These alignment equations were verified by using a Space Telescope optical' formula, misaligning the secondary mirror, and calculating the resulting coma and astigmatism in the wavefront. The tilt and decenter, necessary to realign the two mirrors of the Space Telescope were calculated from the derived equations, and found to agree with the perturbed secondary mirror misalignment. An autocollimation test procedure will be described for obtaining the full aperture wavefront of a Cassegrain-type system, both on-axis and in the field.

A method for aligning off-axis aspheric surfaces has been developed. Normally, the vertex of a rotationally symmetric surface is used in the alignment procedure. This point, however, is not always found on the off-axis aspheric surface. The alignment procedure employs mechanical techniques to center this virtual point and optical techniques to remove the surface tilt. A mechanical jig, alignment telescope, and wire test device are all the equipment required. Expected alignment accuracies are discussed. A procedure for aligning an off-axis aspheric with unknown parameters is outlined.

The ability to measure the absolute effective focal length (EFL) of high resolution tele-scopic systems is limited by errors in the alignment of the test equipment used for the measurement. The types of errors are usually dependent upon the specific test equipment used and the way in which it is set up. While these time independent (D.C.) misalignments are of serious concern, other environmentally-dependent misalignments can be equally dis-turbing; the periods of these dynamic misalignments range from 30 minutes or more to a millisecond. In an effort to measure EFL to better than one part in 104, an extensive study of all known sources of misalignment has been carried out in addition to simply evalu-ating the statistical error associated with repeated EFL determinations. This investiga-tion utilized strain gauges, precision knife edges, assorted lasers, mechanical fixtures, accelerometers, seismometers, position sensing detectors, interferometers, and a PDP 11/70 computer. Reported here are The real constraints to be observed in setting up a laser interferometer How support fixtures should be manufactured The effects of test-table bending in thermally unstable environments The ways to minimize beam steering effects over long optical paths The inadequacies of "vibration-isolated" tables How precisely one can expect to measure EFL absolutely

An autocollimator is an optical instrument which uses its own collimated light to detect small angular displacements of a mirror. Electronic autocollimators provide a continuous output voltage, the sign and amplitude of which are proportional to the mirror displacement. Design and performance considerations of electronic autocollimators are discussed with emphasis on internal and external noise, the trade-off between bandwidth and sensitivity, selection of light sources, and uses. Included is a review of the different types of autocollimators and the criteria for selection of accessories.

Two instruments have been developed, the Bi-Directional Telescope (BDT) and the Variable Offset Periscope, that will aid the alignment of large aperture, long pathlength optical systems. The BDT has been designed to view two ends of a complex long pathlength optical system along a common optical axis. The BDT provides a beam whose maximum offset between beams approaches .001" and maximum angular misalipment approaches 6.0 microradians. The other instrument, the Variable Offset Periscope, has been developed to align multiple mirror or annular optical systems to a com-mon axis, where the optical axis is obscured. The periscope is capable of offsetting a beam over most of a 23.0" diameter aperture with an angular error of less than 2.0 microradians between input and output beams.

A shearing interferometer block consisting of a beamsplitter cube, a right angle prism and a corner cube is used in conjunction with a dual detector to measure when an input beam is normal to the cube to better than 0.001 arc seconds. The two arms of the interferometer have optical path differences of X/4. The wavefronts of the exit beams, one from the corner cube and the other from the right angle prism are tilted in opposite directions from a moving input beam. Zero-crossings of the detector pair output occur when they each receive the same optical power. The output maximum and minimum occur just before and after the zero-crossing associated with normal in-cidence of the input beam. Analysis shows the zero-crossing to be detectable to an accuracy of better than 0.0001 arc seconds for 2.5 cm optics, 1 mW optical power and 1 MHz electronic bandwith with a dual p-i-n photodetector. Laboratory tests confirm this result. The interferometer block has many precision measurement applications. Those discussed herein include: a precision angular reference point; a building block to define a precision reference beam; a sensing element for high precision active control to maintain alignment between two surfaces; a star tracker sensing element. All of the above instruments should have a precision better than 0.001 arc seconds.

The design of a heterodyne interferometer to align the two elements of a reflaxicon is described. The discussion covers the optical, electro-optical, and mechanical design. In this application the interferometer is required to measure the phase variations across the wavefront of a probe beam that returns from the test optics. The measurement must have high resolution and be completed within about one second. The test wavefront must be corrected to remove system figure errors, alignment optics errors, and dynamic disturbance effects. The corrected wavefront is then processed to separate effects due to each degree of freedom of misalignment of the reflaxicon. Then each degree of freedom can be corrected by commands to mirror mount actuators from an alignment computer.

This paper discusses a common reflective collimator approach to the alignment and testing of laser ranger/designators and thermal imaging systems mounted on common carriers. All test measurements are made at a common focal plane. A large aperture reflective common collimator is capable of testing multispectral systems independent of configuration as long as the total aperture of the system to be tested is less than the aperture of the collimator. An analysis of the test system capabilities with respect to measurement of imaging system performance, boresight of imaging system or laser receiver to laser transmitter and measurement of laser transmitter parameters is outlined. A shearing plate technique for accurate focusing of the test collimator is discussed. The shearing plate technique is suitable for field use and can provide a focus accuracy on the order of tens of microradians.

Several techniques are described for boresighting 1.06 pm designator lasers to visual, TV, and forward looking infrared (FLIR) sights. One concept can be used to boresight all three types of sights, even in flight with relative motion between the laser designator system and the boresight module. Special factory hardware incorporating a CO2 laser, a reflective beam expander telescope, and a phosphor thermal imaging disk was developed for precision alignment of the laser input and FLIR output lines of sight of the boresight module. The operation and effectiveness of the boresight module is illustrated by airborne sensor video imagery and missile firing test results.

The accuracy and precision of optical alignment ultimately depend on the accuracy and precision of estimating the position of the pattern of light radiation. When the light radiation is sensed with an array of electronic photodetectors, the position is estimated from the photoelectric counts registered by the detectors. The probability distribution of the counts is used to establish the most elementary bound (Cramer-Rao) for variance of a set of position estimates. The corresponding minimum number of signal photons necessary to perform the measurement (estimation) with specified accuracy is established and its dependence on background noise and random fluctuations in signal intensity is examined. The results are of interest in all those applications for which the accuracy of the measurements has to be much better than the diffraction spread introduced by the collecting aperture.

Remote sensing, high resolution FTS instruments often contain three primary optical sub-systems: Fore-Optics, Interferometer Optics, and Post, or Detector Optics. We discuss the alignment of a double-pass FTS containing a cat's-eye retroreflector. Also, the alignment of fore-optics containing confocal paraboloids with a reflecting field stop which relays a field image onto a camera is discussed.

A novel method using a Dynamic Hartmann Sensor approach is described that allows measurement of the "composite slope" of an optical system as a function of spatial coordinates across its aperture. The technique allows these measurements to be made either in transmission or in reflection (roundtrip). The present technique may be used as a sensor for aligning components of a multi-element optical cavity. One of the important features of this approach is the ability to measure the total error of the measuring system and to subtract instrumentation error automatically from subsequent measurements. Many of the desired features of an electronic interferometer may be incorporated in the system, since the two-dimensional data - spatially correlated with the optical system's aperture to be measured - is available instantaneously in analog or digital format. This latter feature also makes this technique a significant sensor concept for servo control tasks and in real time systems diagnostics. This paper will describe specific applications.

An interferometric technique for aligning a reflaxicon consisting of two separate conical mirrors is presented with emphasis on the techniques for analyzing the interferometer output to extract the required alignment corrections for five degrees of freedom. Error sources for the technique are delineated and analyzed. A computer simulation is described for estimating system performance.

This paper will cover a few of the alignment problems solved by conventional optical tooling techniques at the Los Alamos Meson Physics Facility (LAMPF). The LAMPF accelerator is a large linear proton accelerator used in physics experiments. The alignment techniques and expertise developed on the LAMPF project have been utilized on several other Los Alamos Scientific Laboratory projects including alignment of laser systems, normal physics experiments, machine tools and normal mechanical inspection jobs. The topics to be discussed are conventional optical tooling setups, non-conventional setups, special and/or unique alignment targets and the utilization of an optical tooling dock. The relationship between alignment tolerances, component design, support structure and/or mounting base design and the specific optical tooling setups will be the major theme in the paper.

Two beam interferometers have been proposed for space applications such as sensing the shape of a large antenna. Since alignment and adjustment of interferometers have long been considered difficult laboratory tasks, the question of making their operation sufficiently automatic for space applications is a serious one. As a first step in addressing this question certain manual procedures, which may not be well known, have been collected from widely scattered sources. These techniques are illustrated by two examples: (1) the alignment of a Mach-Zehnder interferometer and the adjustment of fringe location. (2) The adjustment of a Michelson interferometer for zero path difference (white light fringes). .

This report describes one technique used to obtain orecise alignment of small diameter lasers. This procedure may be useful in the alignment of other lasers, but is especially valuable when aligning lasers that have a small diameter active medium and/or a curved mirror at one end of the laser cavity. The technique described in this report uses a He-ATP laser at one end of the laser being aligned and an autocollimator at the opposite end. These instruments are used to generate and observe the diffraction pattern and interference fringes caused by the limiting aperture of the lasing medium and the end mirrors of the cavity, respectively. These Patterns and fringes are used both to establish a common optical axis between the active volume of the laser being aligned and the aligning instruments, and to set the end mirrors of the cavity normal to this axis.

A reflaxicon that consists of an inner and outer axicon generates an annular bear from a compact beam. After the reflaxicon has been internally aligned, a linear cone must be centered in the annular leg to the optical axis of the reflaxicon. A technique for centration of the linear cone will be discussed.

This paper describes the engineering steps taken to obtain best imagery across the field at the R-C focus of a 4m f/2.65 primary in combination with a secondary giving a final f/8 beam. Real life constraints, such as limited initial information, time available for testing, and manpower available, point to the importance of some factors which tend to be overlooked during the early mechanical design and optical manufacturing stages, as well as during the initial assembly of the telescope. Some of these factors are listed so that others may succeed in an operation which most likely will be performed under adverse conditions.

It is shown that, for each field-dependent aberration type, the effect of tilts and decentrations is to produce a more complicated field dependence and local orientation of the aberration without otherwise changing its character. Particular attention is given to astigmatism.

This paper describes the coalignment system for the six 1.8 m astronomical telescopes of the Multiple Mirror Telescope. Emphasis is on individual component alignment and the achievement and maintenance of particular image configurations. The system employs a HeNe laser as an artificial star with its light colli mated by a 0.7 m guide telescope and trasferred to the 1.8 m telescopes via periscopes and retroreflectors. Silicon detectors sense laser image position deviations due to mirror tilts or displacements, and generate signals to actively correct pointing errors via 3 axis motion of the tel Cassegrain secondaries. Procedures are given for aligning the laser beamsplitters, collimating the beams, and aligning the detectors sequentially to arrive at a desired configuration. Keys to the operation are the use of a pinhole in the guide telescope's focal plane as a defining point for the laser source, the active stabilization of the guide telescope's focus, the stability of the periscopes, and the use of remotely-controlled thin wedge prisms to change each laser beam's position when reconfiguring any one telescope. Tests of star tracking show typical coalignment to <1.0 arc second RMS. A discussion of problems encountered is included.

Room temperature alignment and evaluation techniques for the Infrared Astronomical Satellite (IRAS) telescope which has a primary mirror figured to correct for surface distortions at the 2K operating temperature will be discussed. Interferometric cryogenic testing of the 0.6 meter, f/1.5 lightweighted beryllium primary mirror at its intended operating temperature revealed surface distortions that could be modelled with Zernike polynomials. With this model, it was possible to derive the "inverse" of the cryo wavefront error (ideal cryo mirror) and figure the cryo correction into the primary mirror using Perkin-Elmer's Computer Controlled Polisher. It was recognized that during room temperature assembly of the system, misalignment of the secondary mirror could introduce additional unwanted aberrations that would cancel or distort the wavefront errors purposely introduced by the cryo figuring. To avoid this possible degradation and to ensure optimum telescope performance, the system Zernike polynomial co-efficients and wavefront maps generated from the in-process alignment interferograms were monitored and compared to Zernike coefficients and wavefront maps for the cryo corrected primary mirror. Performance specifications for the telescope included encircled energy on-axis and at the edge of the field of view. The predicted Zernike coefficients for the telescope at cryo temperatures were determined by subtracting the room temperature coefficients for the aligned system from the ideal cryo mirror. Encircled energy was calculated for on- and off-axis by adding the predicted Zernike coefficients to the nominal optical design prescription in the Perkin-Elmer Optical Design Computer Program. Using Zernike polynomials to iiionitor the optical quality of the telescope enabled figure and alignment errors t© be monitored, and demonstrated that the alignment tolerance was achieved. Describing the telescope's wavefront in terms of Zernike polynomials also per mitted prediction of the telescope performance for various field positions and detector or

This paper describes the alignment approach for the infrared astronomical satellite (IRAS) optical subsystem from initial design to acceptance testing. The constraints imposed by the requirement of maintaining alignment at 300K and 2K, in a 1-g and 0-g gravitational field, during warm and cold vibration, and during various stages of assembly, are discussed. The paper concludes with the methodology of applying NASTRAN finite element analyses to the alignment design, followed by the verification of the accuracy of the design with the test results.

An electro-optical attitude transfer system was developed to monitor the angular orientation of magnetometers deployed at the end of a 20-foot boom extending outboard from the Magsat spacecraft. One autocollimator monitors pitch and yaw attitude, cooperating with a plane mirror at the end of the boom; a second monitors roll (twist) from an offset look-angle, using one dihedral reflector at the boom end and a second on the spacecraft. RMS errors due to all causes including linearity, 0-forces, cross-coupling and translation are estimated to be 3.9 arc-seconds over ±180 arc-sec. excursion in pitch and yaw, and 5.3 to 7.5 arcseconds over ±300 arc-sec roll. Design and fabrication problems relative to the remote dihedral reflector proved to be the most challenging, and solutions to these problems will be described.

The low-flying MAGSAT spacecraft, launched October 30, 1979, included a Vector Magnetometer to accurately map the magnitude and direction of the magnetic field of the Earth. Calibration of the magnetometer included arc-second precision determination of the relative orientations of the three sensor axes in a coordinate system defined by optical references. This determination began with laboratory measurements of the relative alignments of optical components mounted with the magnetometer. The actual calibration procedure then consisted basically of accurate and repeatable positioning of the Vector Magnetometer within a unique magnetic test facility which nulls the earth's magnetic field, then generates magnetic fields of various orientations and strengths. Analysis of the magnetometer sensor outputs together with the position and alignment data then gave the axes orientations. We used precision theodolites and methods re-lated to surveying techniques to achieve the accurate positioning and optical component alignment measurements. The final calibration accuracy exceeded results previously achieved in the facility.

A modular alignment system has been designed and built at the Institut d'Astrophysique, Lige. Based on the multiple-beam autocollimation principle, it is composed of two collimators, a series of beam splitters and mirrors and a large reference flat. All the generated laser beams are aligned with respect to the latter and are projected onto corresponding test mirrors. These can then be set perpendicular to the beams by shimming the payload on which they are attached. The autocollimation is controlled visually by a micrometer stage, or on a TV screen. Some twenty beams may be aligned simultaneously with a precision of the order of one second of arc, regardless of the floor stability.

The Shiva oscillator pulse is preamplified and divided into twenty beams. Each beam is then amplified, spatially filtered, directed, and focused onto a target a few hundred micrometers in size producing optical intensities up to 1016W/cm2. The laser was designed and built with three automatic alignment systems: the oscillator alignment system, which aligns each of the laser's three oscillators to a reference beamline; the chain input pointing system, which points each beam into its respective chain; and the chain output pointing, focusing and centering system which points, centers and focuses the beam onto the target. Recently the alignment of the laser's one hundred twenty spatial filter pinholes was also automated. This system uses digitized video images of back-illuminated pinholes and computer analysis to determine current positions. The offset of each current position from a desired center point is then translated into stepper motor commands and the pinhole is moved the proper distance. While motors for one pinhole are moving, the system can digitize, analyze, and send commands to other motors, allowing the system to efficiently align several pinholes in parallel.

The OMEGA laser irradiation facility consists of a 24 arm Neodymium glass laser coupled to a spherical target irradiation chamber. A computer network is connected to the laser and target systems through a control room where operators can access any of the devices or subsystems within the facility. The facility is designed to produce 12 terawatt laser pulses and irradiate various materials to study laser-matter interactions at extremely high peak power. The system is capable of sustaining one target shot every 30 minutes.

Antares is a 24-beam-line CO2 laser system for controlled fusion research, under construction at Los Alamos Scientific Laboratory (LASL). Rapid automatic alignment of this system is required prior to each experiment shot. The alignment requirements, operational constraints, and a developed prototype system are discussed. A visible avelength alignment technique is employed that uses a telescope/TV system to view point light sources appropriately located down the beamline. Auto ignment is accomplished by means of a video centroid tracker, which determines the off xis error of the point sources. The error is nulled by computer riven, movable mirrors in a closed op system. The light sources are fiber ptic terminations located at key points in the optics path, primarily at the center of large copper mirrors, and remotely illuminated to reduce heating effects. There are 2 power amplifiers, each containing 12 beams. One telescope/TV camera is needed per amplifier. The driver amplifier, power amplifier, and target system optics are coaligned to the telescope/TV camera optical axes. The final alignment to the target is accomplished with the use of a special fixture. The use of visible light for alignment, rather than infrared, requires less expensive components, gives a smaller diffraction blur, and permits more accurate pointing. The dispersion between 10.6 μm and visible light due to the NaC1 windows is measured and compensated by the control system.

Large inertial confinement fusion laser systems have many beams focusing on a small target. The Antares system is a 24�beam CO2 pulse laser. To produce uniform illumination, the 24 beams must be individually focused on (or near) the target's surface in a symmetric pattern. To assess the quality of a given beam, we will locate a Smartt (point diffraction) interferometer at the desired focal point and illuminate it with an alignment laser. The resulting fringe pattern shows defocus, lateral misalignment, and beam aberrations; all of which can be minimized by tilting and translating the focusing mirror and the preceding flat mirror. The device described in this paper will remotely translate the Smartt interferometer to any position in the target space and point it in any direction using a two axis gimbal. The fringes produced by the interferometer are relayed out of the target vacuum shell to a vidicon by a train of prisms. We are designing four separate "snap in" heads to mount on the gimbal; two of which are Smartt interferometers (for 10.6 pm and 633 nm) and two for pinholes, should we wish to put an alignment beam backwards through the system.

A design approach for an alignment system for an unstable resonator is presented. A description of the Air Force airborne gas-dynamic laser resonator is followed by the results of a paraxial ray trace, including aperture stops, of the images used in the alignment method. An analysis of these results led to the design of the resonator alignment system.

A multidither resonator alignment scheme has been developed which automatically aligns the optical components of a laser to optimize its performance and then maintains the alignment despite further disturbances to the optical components. This was demonstrated on a reflaxiconcone Half Symmetric Unstable Resonator with Internal Axicon (HSURIA) laser, controlling azimuth and elevation of the feedback mirror to maximize Far Field Power-in-the-Bucket.