hertz to 17.8 hertz. This is equivalent to increasing the stiffness of the beam by a factor of about 5.

A further improvement can be obtained by increasing the stiffness-to-weight ratio of the beam. In the case being reviewed, this was done by casting a 30-inch-diameter cylindrical void through the length of the beam (fig. 18). This increased the fundamental frequency of the beam another 14 percent to about 20.3 cycles per second while decreasing the weight to be isolated by 36,800 pounds, a 27-percent decrease in weight.

To illustrate the improvement in performance of the isolated hollow block over the solid, simply supported block, let us assume that the floor has a predominant disturbing frequency of 10 hertz (which is not uncommon). A block of this construction would be moderately well damped, having a transmissibility at resonance of about 5. Using transmissibility curves for viscously damped systems (fig. 2), it is found that the 10-hertz disturbance would be amplified by the simply supported, 7.8-hertz beam by about 50 percent.

The isolated hollow beam has to be considered in two steps: (1) the response of the beam to inputs at its support points and (2) the attenuation of the disturbing frequency by the isolators. With regard to Step 1, the beam now has a fundamental frequency of about 20 hertz, and, instead of amplifying 10-hertz inputs, it will attenuate them by about 60 percent. This difference alone accounts for an improvement of approximately 4 to 1 over the simply supported solid beam.

If the isolation system is assumed to have a 1.6-hertz natural frequency (which is typical), the transmissibility of the 10-hertz disturbing frequency will be about 0.1 (attenuation of about 90 percent before it gets to the block). The combined effects will result in approximately a 40:1 improvement with regard to the relative transmissibility of the 10-hertz disturbance.

Secondary benefits of the hollow block are that the floor loadings will be lower and that an isolation system with lower load capacity can be used. Both of these will result in a lower installation cost.

Development and Trends in the Vibration Isolation of Optical Systems

The isolation systems that have been discussed in the two case studies represent state-of-the-art techniques now available to solve the more critical vibration isolation problems. The solutions are not always perfect, however, and one wonders what is being done or can be done to make better solutions available to science and industry.

Actually, there is a continuous effort being exerted in this field. The now famous Rolamite, developed at Sandia, has been cited as the only significant mechanical invention of the twentieth century. It is a spring that has infinitely variable force-deflection characteristics. Research and development by other companies is being carried on with other objectives; for example, to obtain systems with very low natural frequencies

without paying the price of large static deflections. The pneumatic systems now available were the first step in this direction; the adverse effects of static deflection have been eliminated by using servo-valves to vary pressure with load. These systems have literally zero static deflection. The properties of these systems, however, are still space-limited in a sense. The natural frequency depends on the volume of air in the isolators.

A large portion of the research and development in this field is being directed at active servocontrolled isolation systems. With the use of electronic control systems in conjunction with hydraulics, pneumatics, and mechanisms, the relationship between the static deflection of a spring and its isolation properties has been eliminated. Electrohydraulic systems that have been made and have been successfully demonstrated occupy a space of about one cubic foot and have a natural frequency of 0.1 hertz (refs. 7, 8, 9, and 10). If a passive, conventional spring were used, such a system would have a static displacement of over 80 feet, which would require a spring at least as long as that and probably 200 feet or more in length. Furthermore, these systems can be made with unilateral properties so that forces applied to one end "see" a very soft spring, but, when viewed from the other end, the system appears to be very, very stiff. In certain applications, this is a tremendous advantage over passive systems.

Another development, recently revealed, is an electromechanical feedback control system, which, when used with a pneumatic isolation system, will lower its natural frequency by an order of magnitude. These, too, can be designed with unilateral properties; however, most of the advantages obtained by the active feedback systems occur in the frequency range from very low frequencies (a fraction of 1 hertz) to about 30 hertz. Although this is frequently a very important frequency range, the optical researcher is affected as much or more by structural resonances within or around his equipment; and these resonances are usually in the approximate range of 20 to 200 hertz.

In this range, isolation efficiencies of 95 percent or better are now attainable by using well designed and well applied systems.

It has been the experience of the author that the isolation goals are often thwarted by so-called "peripheral effects." Probably he most common and most disturbing of these peripheral effects is acoustic coupling between the isolated equipment and the surroundings. This is especially true with vacuum chambers. Most vacuum chambers are excellent gongs; they have a multitude of resonant frequencies (most of them in the audible range) and extremely little damping. Another problem often encountered results from the tendency of experimenters to build their equipment "á la Erector set." Lens and mirror makers have long recognized the necessity of designing good supports for their optics, but they often overlook the same requirement when designing auxiliary structures, brackets, and instrumentation. Consequently, structural resonances in this equipment often cause trouble.

If the adage "an ounce of prevention is worth a pound of cure" is applied, then new research facilities should be planned with the vibration problem in mind. This includes selection of the site as well as the architecture of the building. Vibration and noise transmissibility should be included in the criteria for selection of building materials and techniques. The mechanical equipment in a building (air compressors, pumps, etc.) should be remotely located, if possible, as should be any auxiliary support functions, such as machine shops, printing facilities, and any others that generate acoustical noise or mechanical vibration. Effort in these areas will greatly reduce the isolation requirements and also the problems to be solved for the sensitive experiments to be conducted in the building.

Another area that definitely requires more study and development is the design of vacuum chambers. Materials and designs to date have been selected primarily on the basis of strength and outgassing properties. Stainless steel cylinders and spheres satisfy these requirements but are frustrating to deal with from the standpoint of vibration control.

Solid (homogeneous) and laminated materials have been developed that have excellent damping properties, but these materials have been developed for quite different applications (primarily for chassis and mounting bases for airborne electronics); hence, the materials and techniques are not suitable for vacuum chamber applications. On the other hand, the principles used would apply to vacuum chambers. This is a field that forward-looking manufacturers of vacuum systems might be well advised to investigate.

Externally applied materials to absorb vibration energy would seem to be a practical alternate approach, but, to the best of the author's knowledge, little or no effort has been spent on developing this idea. Studies of the effect of externally applied materials need not involve drastic redesign of the chambers; thus, the studies would be easier and less expensive. Summary

Although the theory of vibration isolation is well developed and has been successfully applied for many years, the everincreasing demand for better isolation forces a much closer scrutiny of the environment, of the equipment to be isolated, and of the isolation systems and their controls.

Electronic feedback control systems are being developed which make many heretofore unattainable isolation characteristics feasible. These are concerned primarily with attitude and position control and with improvements in low-frequency (up to approximately 30 hertz) isolation. Although these are often matters of concern to optics experimenters, isolation of higher frequencies is just as important, or more so.

No great breakthroughs are currently envisioned that will make a dramatic improvement in the isolation of vibration in the middle to high frequency ranges; however, significant improvement can be achieved by taking the "systems approach." The vibration sources (including the natural environment) must be considered, and steps must be taken, where possible, to control the source. In

many cases, vibration transmission paths, which include the building structure and the air within it, can be improved at moderate cost. Finally, the laboratory equipment, including the instrumentation and its supports, can usually be designed with high resonant frequencies and good damping properties. Such design will greatly reduce the vibration problem.

Vibration exists and, in all likelihood, will continue to cause problems as long as man continues to make things that go. The most effective and economical approach to solving these problems is to search them out at the earliest possible stage, preferably in the conceptual phase of a program, and to consider the isolation as part of the system and not merely an accessory to it.

References

1. Jacobsen, L. S.; Ayre, R. S.: Engineering Vibrations. McGraw-Hill Book Co., 1958.

2. Crede, C. E.: Vibration and Shock Isolation. John Wiley & Sons, Inc., 1951.

3. Den Hartog, J. P.: Mechanical Vibrations. McGraw-Hill Book Co., 1956.

4. Timoshenko, S.: Vibration Problems in Engineering. D. Van Nostrand Co., Inc., 1928.

5. Crede, C. E.; Harris, C. M.: Shock and Vibration Handbook. vol. 2, chap. 33. McGraw-Hill Book Co., 1961.

6. Krach, F. G.: Reinforced Concrete Beam Resonances. Shock and Vibration Bulletin, No. 38, pt. 2, August 1968.

7. Ruzicka, J. E.: Active Vibration and Shock Isolation. SAE Paper No. 680747. October 1968.

8. Calcaterra, P. C.: Performance Characteristics of Active Systems for Low Frequency Vibration Isolation. SM Thesis, MIT, June 1967.

9. Calcaterra, P. C.; Schubert, D. W.: Research on Active Vibration Isolation Techniques for Aircraft Pilot Protection. Air Force Report AMRL-TR-67-138, (DDC Report AD 664090), October 1967.

10. Pepi, J. S.: Vibration Isolation of Optical Reconnaissance Sensors. Air Force Report AFAL-TR-67-277, (DDC Report AD 822872), October 1967.

Contamination rather than degradation is the principal concern of the discussion within this paper. If we devote our attention to the optical system, the type of degradation to be expected is a weakening or loss of the optical signal. In many cases, this will be wavelength-dependent; hence, some distortion of the spectrum will occur. Other types of degradation also occur. For example, a stray bit of lint or a dust particle, when illuminated by sunlight, may look enough like a star to confuse a tracking system. Solid contaminants may degrade the performance of a mechanical system. Bits of electrically conductive material may short out an electrical circuit element, thereby causing loss of power, or failure of a logic circuit, or some other failure.

In the long process from the start of fabrication to the launch into space, there are, however, a great many opportunities for contamination. Let us assume for this discussion that the fabricator takes all the precautions necessary to avoid contamination in fabrication and assembly. (Sometimes he does not take these precautions, and his failure often becomes apparent during test and evaluation.) In addition to the measures used to avoid contamination in fabrication, telescope systems are customarily protected in transportation and storage by shock mountings to prevent mechanical damage and by bagging in an above-atmosphere pressure of dry nitrogen and at a controlled temperature. These precautions are generally effective, but an occasional failure does occur.

When an item is ready to be shipped to the Goddard Space Flight Center (GSFC) for testing, we perform the first inspection at the supplier's plant and a similar inspection after

receipt at GSFC. These inspections help us to determine whether the item is in a flightreadiness state as supplied and whether contamination has occurred in shipment. They also provide a baseline for determining whether contamination has occurred during the test cycle. The detailed procedure that has been developed for the Goddard Experiment Package of the Orbiting Astronomical Observatory (OAO) is attached to this paper as Appendix A. Briefly, the procedure is to subject the equipment to careful visual scrutiny with enhanced light and with ultraviolet to bring out any fluorescence. In addition, a vacuum-sampling procedure is used; all particulate matter thus collected is examined with a microscope and identified. This vacuum sampling is done by passing the end of a tube through which air is being drawn over critical or representative parts of the apparatus. The most frequently found particles are epithelial cells, hair, eyelashes, cosmetics, food, smoke particles, dust, plastics, carborundum, glass, paint, and metal. Wipe tests are also made when appropriate. The material collected in the wipe test is ordinarily analyzed by spectroscopy or chromatography. The things we find most frequently are fingerprints, grease, oil, hydrocarbons, monomers or plasticizers, and paint residues.

We have been concerned about the possibility of contamination in the test-andevaluation building by particulate matter and by gaseous components of the air such as water, hydrocarbons, and corrosive vapors. We attack the particulate problem by maintaining a class-10,000 clean room and meticulous clean-room procedure and by