figure 14. The facility is located in a threestory, class-10,000 clean room in the test complex of the Test and Evaluation Division at GSFC and has been in use for experiment calibration since late 1964.

Figure 14. Vacuum optical bench (VOB).

The principal elements of the VOB are a vacuum chamber, 7 feet in diameter by 25 feet high, and a removable vertical optical bench. The bench has its own integral starsimulator system, which includes light sources, a vacuum ultraviolet monochromator, a 38-inch aperture Cassegrainian collimator, and a movable beam probe.

The bench is adequately isolated from building vibrations; it stands on a set of kinematic piers mounted on a 9-foot deep concrete block; the chamber rests on the building floor. With the use of an overheadtraveling bridge crane and a special bench up-ender, the entire bench, with an experiment installed, can be positioned vertically

with either end up (for ± 1 g testing) and in or out of the chamber as desired.

The optical experiment to be calibrated is mounted within the optical bench structure and is rigidly held in place on mounting lugs that are geometrically identical to those on the spacecraft. The bench is capable of accepting experiment packages up to 40 inches in diameter by 120 inches long and weighing up to 1000 pounds.

The Star-Simulator System (SSS) is the key system of the VOB with regard to the confidence level that can be established in calibrating a stellar experiment. The SSS can be described by using the optical schematic shown in figure 15. Radiation from a light source, S, is focused by a toroidal mirror, M1, onto the entrance slit, A,, of the monochromator. Passing through the monochromator, the light is diffracted and focused on a circular exit aperture, A2. The exit aperture serves also as the entrance aperture of the main collimator. The collimator optics are an f/10 Cassegrainian system with a convex hyperboloid, M4, and an f/2.03 concave parboloid, Ms. The 38-inch diameter output beam of the Cassegrainian collimator is used to irradiate the experiment being calibrated.

The performance requirements of the SSS are unusual. Not only are narrow bandpass, small beam divergence, and both high and low flux levels desired, but also a high degree of uniformity across the collimated beam is necessary. In addition, a high degree of stability with time is needed with regard to the flux level and pointing direction.

In general, the VOB operates over the wavelength range from 1100 to 5000 angstroms. The bandpass can be varied from 2 to 4 angstroms while under vacuum. The total decollimation of the beam at a 2-angstrom bandpass may be as small as 2 arc seconds. A dynamic range of flux of 103 can be achieved and reliably measured without resorting to the use of pulse-counting techniques.

Although enclosed light sources are used for some specific purposes, a free-flowing gas discharge source is used for most calibrations.

design and additional details on the VOB are given in reference 10. The original monochromator design consisted of a plane mirror, which rotated about the center of the Rowland circle, and a circle, and a concave tripartite grating. In 1962, when that first design was selected, no one had ever coated a 38-inchdiameter mirror with aluminum plus a magnesium fluoride overcoating; hence, the reflectance that would be achieved at wavelengths near 1216 angstroms was unknown but was expected to be low. Consequently, the minimum possible number of reflecting surfaces was used throughout the VOB. This was the reason for selecting that particular monochromator design. Due principally to the tripartite grating, large intensity variations existed across the output cone from the monochromator. After the facility was built, it was found that sufficient reflectance was achieved to permit the use of additional reflections. At that point, the monochromator was redesigned. A plane grating was used, and more efficient use of the light source was made by incorporating the toroidal mirror. A dramatic improvement in uniformity in the 38-inch beam occurred, and the flux levels actually increased even though two additional reflecting surfaces were involved.

A differential pumping system minimizes the pressure rise in the main vacuum chamber. The type of gas, its flow rate, and the light source current can be varied while under vacuum. In addition, a filter changer allows the placement of color and neutral density filters in front of the monochromator entrance slit. Optimization of the source materials and the range of currents has resulted in stable operation for periods in excess of 100 hours.

Specific details of the present monochromator design are given in reference 9. A discussion of the results achieved with this

The Cassegrainian collimator has the focal point located inside the hole in the primary mirror. This is a most restricting choice, one to be avoided in any future facility if at all possible. The aluminium-plusmagnesium-fluoride coating techniques have been so refined during the past few years that in December 1968, when the 38-inch diameter primary mirror was last coated, the reflectance obtained at 1216 angstroms was 82.5 percent with only a ±2 percent variation over the entire surface. Nearly the same degree of uniformity has been achieved using lithium fluoride overcoatings, but a compromise must be made between the reflectance desired at Lyman a and Lyman ẞ. The technique used is described in reference 11.

Each time the mirrors have to be changed, the remounting and realignment of

them is a very time-consuming process. In this f/10 system, the best focus position is very sensitive to intervertex spacing. Adjustments must be made to locate the best focus within the dimensional range available for positioning the monochromator exit aperture. Except for the time required for these initial alignments, the method presently in use is adequate for the non-diffraction-limited optics concurrently on hand. This method involves autocollimating from a large flat, scanning along the optical axis to find the best image position for various intervertex settings, then measuring and photographically recording the return image. This cut-and-try method would not be accurate enough to make effective use of diffraction-limited optics. (Interferometric and other suitable techniques currently available will be discussed in other workshop papers.) Careful consideration must also be given to the adaptability of such techniques for use in a thermal-vacuum environment. This is certainly a "must" for any future calibration facility.

Because it is time-consuming to remove, recoat, and reinstall the VOB collimator optics, degradation of coatings due to any type of contaminant is a matter of continual concern. Regardless of the care exercised, eventually the problem of cleaning the mirrors in place must be faced. This type of cleaning has been successfully done for the magnesium-fluoride overcoated mirrors by flooding the 38-inch-diameter primary mirror with freon TF and immediately drying it with clean, dry, gaseous nitrogen. That statement, however, should be viewed with a great deal of caution because success is largely dependent upon the actual technique used and several other factors, such as the type of tubing, the spray nozzle, the handling, the source of the freon, etc. Cleaning this 38-inch mirror when overcoated with lithium fluoride has not as yet been attempted because to determine the degree of success would require that the chamber be available for three months. Such cleaning, however, has been done on small mirrors by Dr. George Hass at Fort Belvoir.

The intended use of the SSS is to perform an absolute radiometric calibration of stellar experiment packages. This requires a detailed knowledge of the absolute flux level in the 38-inch beam and any spatial or temporal variations of the flux level. To obtain such data in the VOB, an r-0 drive, beam probe is used to transport calibrated multiplier phototubes to scan the 38-inch beam. In a normal scan in the VOB, 3000 data points are obtained. This requires the availability of high-speed data collection, processing, and display equipment.

The calibration of these scanning multiplier phototubes, in order to use them as secondary standards, requires on-site laboratory equipment for performing accurate calibrations in the visible, near ultraviolet, and vacuum ultraviolet. Years of learning time are needed to use such equipment knowledgeably. The on-site requirement stems from two sources. First, there is evidence that such tubes change absolute calibration with time, which makes us to want to calibrate them immediately before and after each use period. Second, only a very few organizations are equipped to perform the vacuum ultraviolet portion of the calibrations, and their laboratories are fully utilized in handling their own work. Much more needs to be done in this particular field.

Consideration of the beam-probe drive system and the electronic and electrical equipment involved in the measurement of the flux in the beam opens another door to a whole spectrum of problems. Avoidance of contamination serves as a major constraint on material selection in each case. Some key areas of concern are with motors, lubricants, nonreflecting finishes and coatings, highvoltage cables, connectors and feedthroughs, electrical noise and magnetic field generation. Although it seems that enough is known in these areas to survive today, each time the need arises to stretch the capabilities of the VOB, substantial exploratory work has to be done. The list of usable materials and equip ments in these areas must be expanded to meet the needs of future observatories.

Low Temperature Optical Facility

The Low Temperature Optical Facility (LTOF), which is a companion facility to the VOB, was built to determine any misalignments that may occur in the OAO experiment packages when they are subjected to orbital temperatures. Ever since the first test of an experiment was run in late 1964, this facility has been found to be far more useful than was originally envisioned. Consequently, some broad-bandpass calibration capability has been incorporated into the facility. A conceptual drawing of the LTOF is shown in figure 16.

Figure 16. Low temperature optical facility.

The LTOF consists of a controlledtemperature test chamber, 13 feet wide by 43 feet long by 72 feet high. Inside the chamber, but mounted on a vibration-isolating block via penetrations in the chamber floor, is a 39-inch clear-aperture collimator and a special handling device for experiments. The chamber is provided with a thermal lock for personnel access at low temperatures. The chamber and anteroom meet class-10,000 clean-room requirements.

The temperature in the chamber can be varied from +23° to -60°C at a predetermined rate. Test temperature is held to ±1°C throughout the testing area. The dew point of the air is maintained below -60°C in order to prevent any condensation on components in the chamber.

The collimator is a Newtonian mirror system, which directs the light from several interchangeable light sources in a 39-inch collimated beam to the optical experiment under test. The optical system of the facility is thermally compensated, and means are provided for initial alignment and for checking this alignment at the test temperature.

The special handling device is provided to receive an OAO experiment in the vertical position, to rotate it to the horizontal, to transport it into the chamber, and to serve as a mounting structure during tests. The experiment can be rotated in pitch and yaw ±15 degrees about the geometrical center and 360 degrees about the roll axis for ±1 g testing. The handling device will accept experiment packages up to 40 inches in diameter by 120 inches long and weighing up to 1000 pounds.

An X-Y beam-probe scanning device is provided; it carries a multiplier phototube. The use of a thermoelectric cooler makes it possible to maintain this tube at -20°C at any chamber temperature. The flux level and uniformity in the 39-inch beam are measured by using this scanner. High-speed data collection techniques are employed in much the same fashion as in the VOB. An array of light sources, imaging optics, broad bandpass, and neutral density filters are located in an insulated light-source housing that is maintained at +23°C by a separate air-handling system.

Numerous boresighting and anglemeasuring instruments are available, but each presents a unique set of problems because of the low temperatures in the chamber. The proper application of low temperature lubricants and the selection of suitable electrical insulation are most perplexing until one has gained some experience in their use at these low temperatures. An accuracy of 1 arc second in making angle measurements has been achieved and can be improved upon under certain conditions.

It should be pointed out that, since air is used in the LTOF, tests are generally run with all of the elements of an experiment package at a uniform temperature as opposed to a

gradient condition, which would prevail in orbit. Consequently, the information obtained in such tests falls into the category of diagnostic and relative data rather than absolute calibration data. By the same token, since man can enter the chamber, tests such as image quality and resolution tests can be performed easily whereas these would be very difficult, if not impossible, to do in a vacuum.

For the calibration and evaluation of future observatories, a combined thermalvacuum-optical facility is needed. It has been found to be most difficult to combine the results obtained in the VOB (vacuum only) with those of the LTOF (low temperature only). From an operational standpoint, however, the unique operating modes of each facility should be retained to aid in improving the confidence level one can achieve with regard to the launch readiness of a future observatory.

Princeton Optical Bench

A partial step toward a combined thermal-vacuum-optical facility has been taken with the Princeton Optical Bench (POB), shown diagrammatically in figure 17. The bench and a special optical dome are mounted on a 12-foot-diameter by 15-foothigh thermal-vacuum chamber. Several optical measuring instruments are mounted on an isolated table outside and above the chamber. These instruments consist of a 4-inchdiameter star simulator and several autocollimators and theodolites. Electronic levels and a pendulous mirror are used to adjust the lines of sight of these instruments parallel to each other. Hanging from the optical table are three servo-driven columns, which penetrate the dome through vacuum-tight bellows. Inside the chamber, the actual bench structure is connected to the columns. At the lower end of the bench is a rigid support ring in which the experiment package and a multimode thermal control shroud are mounted.

As the thermal gradients on the experiment package are changed, through the control shroud, to simulate various orbital conditions, the angular relationships among

the experiment's fine guidance sensor, telescope axis, spectrometer slit location, and several reference mirrors located on critical elements can be measured.

The optical instrument table, from which all other bench components and the experiment are suspended, is isolated from chamber vibrations through air mounts and is also servo-driven to preserve its orientation to gravity. In addition, as thermal gradients change the lengths of the columns, these columns are servo-driven to retain parallelism between the experiment support ring and the optical table.

Acceptance tests using a model of the OAO Princeton Experiment were completed during March 1969. Angle-measuring accuracies of approximately 2 arc seconds were obtained. With this design, isolation from vibrations and other distortions is sufficiently good to keep the jitter in the sub-arc-second region.

Note that all of the optical instruments are mounted external to the thermal-vacuum