The Perfect Machine

by Ronald Florence

  The only damage the structures suffered in the trip from the Philadelphia factory to the observatory floor was some minor denting, nicks, gouges, and defacing with green paint by the stevedores in transit. One piece was dropped from the rigging and made a dent in the concrete floor of the observatory. It would have no effect on the telescope, but Frank Fredericks raged for days at anyone within earshot. There weren’t supposed to be dents in the floor of a perfect machine.

  The shipment of the big Westinghouse components got enough publicity to attract carloads of tourists up the new road. There were no facilities at the top, not even a paved parking lot. The trees and brush had been cleared from the area around the observatory, leaving a barren moonscape; there were no pathways or signs; and only the intrepid could find their way to the door of the dome. Still tourists showed up, eager to see the great machine.

  The area that had been designed as a gallery for visitors, just inside the main entrance, wasn’t enclosed yet, except for a hastily erected chicken-wire partition. Visitors crowded into the area, watching the workmen and asking enough questions to distract them from the complex assembly procedure. Finally, after he had been interrupted with questions so many times that he couldn’t concentrate, Ben Traxler put a neatly printed sign on a pipe standard outside the chicken-wire enclosure:

  DON’T TALK TO THE PRISONERS ASK THE GUARD

  It took Byron Hill a few days to figure out why the tourists had suddenly quieted down. Hill was capable of laughing at a dry joke, but he could imagine the reaction when rumors reached Pasadena that the telescope was being built by chain gangs. He tore the sign down.

  Frank Fredericks had been at the Westinghouse plant in South Philadelphia when the alignment and bolt-holes for the components of the mount were bored. He came to Palomar to supervise the assembly. Some of the holes had to be rebored because of dents in the structures through shipping. Hill argued against relying totally on the alignment pins. He was a Caltech engineering graduate but, with a mechanic’s sense of a machine, skeptical about paper specifications.

  Once the three sections of the great horseshoe were aligned and bolted, the mounting began to take shape. The balance of the telescope depended on the weight of the primary mirror at one end of the tube, so for testing purposes Hill cast a dummy blank of eighteen tons of concrete. The final balancing depended on adjustments, but the design process, sometimes jumping from astronomers to draftsmen, had skipped the stage of rote mechanical engineering, when the mechanical assemblies would be checked for ease of adjustment, lubrication, and servicing. The adjustment studs and screw assemblies were an afterthought; the counterweights were on four-inch Acme screws with unlubricated nuts. The south bearing, a ball suspended on oil pressure pads, was brilliantly engineered, but the only access to align the bearing was through manholes. By the time the telescope was assembled, Hill’s knees were gone from too many trips down the manholes, and he had tasted enough dripping oil for a lifetime.

  Llewellyn Carlson, a salesman for Mobil industrial products, had pitched Mobil’s experience in developing special oils to get the contract to develop the oil for the pressure bearings. The oil had to maintain a film thickness of three one-thousandths of an inch, in a temperature range from 20 to 100°F. Mobil made a publicity fuss about the special work of the General Petroleum Corp. engineers in the Vernon Lab and the Research and Development Department of the Socony Vacuum Laboratories in New Jersey in developing Mobiloil 95 for use in the telescope. Later they marketed Flying Horse Telescope Oil for other observatories.

  Mobil and Westinghouse, in their publicity releases, bragged that the structures were designed and machined so carefully, and the oil bearings worked so perfectly, that 1/160,000 horsepower would be sufficient to move the telescope at the sidereal tracking rate of one revolution per day. As an example of the ease of movement, the engineers claimed that a milk bottle on top of one end of the perfectly balanced great horseshoe would be enough force to move five hundred thousand pounds of telescope. The skeptical “zombies” and Byron Hill couldn’t wait to try that experiment.

  The drive and control system was almost but not quite complete when Sinclair Smith died. His use of servo motors and generators was almost a decade ahead of its time, and not in general use for complex control systems until late in World War II. Smith’s design for the drive and control system placed every control for the telescope at duplicate consoles for the night assistants. Wherever the astronomer was working, the prime-focus cage at one end of the telescope tube, the Cassegrain focus at the other, or the Coudé room below and south of the south pedestal, there was a console nearby. The consoles themselves were meant to be foolproof, even at the end of a long session in a dark observatory.

  The first task of the drive and control mechanism was to point the telescope accurately, rotating the telescope around the right ascension axis, parallel to the axis of the earth, in sidereal (star) time, so the objects in the focus of the telescope would appear to stand still. The telescope would be moved by huge worm gears. The machinists in the astrophysics machine shop had already polished the gears for a year, achieving the precision of a chronograph on the largest gears ever cut. But no amount of polishing could eliminate every trace of backlash, and no bearings could be built without some measurable play. To achieve pinpoint exposures on the faintest possible objects, at the optical resolution that was anticipated for the telescope, the drive mechanism had to compensate for those mechanical aberrations with minuscule corrections in the motion of the telescope. That was the first task.

  As the telescope turned, the dome would also have to turn, its motion synchronized so that the shuttered opening was in line with the field of view of the telescope. It seems simple until you realize that the dome turns in azimuth, rotating on a vertical axis perpendicular to the earth, while the telescope is turning in right ascension, on a polar axis (tilted thirty-four degrees from the horizontal at Palomar). Curve the index finger and thumb of one hand into a C and hold it parallel to the floor. Hold one finger of the other hand, angled at approximately thirty-five degrees from the floor inside the C and slowly twist and lift it. Now try to keep the opening of the C aligned with the end of the tilted and turning finger. That is the task of the dome drive mechanism.

  Wind blowing into the shutter opening of the dome can disrupt the motion of the telescope and change the thermal equilibrium of the optics. A heavy canvas windscreen was designed to fit the shutter opening. On windy nights the screen would be raised high enough to block the wind yet left low enough not to block the telescope’s field of view. The motion of the windscreen is in altitude, perpendicular to the ground; the motion of the telescope it is protecting is in both altitude and azimuth, rising and falling on its equatorial axis. You would need three hands and Houdini’s coordination to imitate the motions with your fingers.

  Finally, one major design goal of the telescope had been that various focal points would be switchable, by the night assistant alone, without the loss of a day or two of engineering time. The ability to shift from prime-focus work to the Coudé room in the course of an evening meant that if the weather or seeing were not good enough for deep-space work, the observer could switch and spend the evening on spectrographic work on nearby stars. The tube, prime-focus cage, and auxiliary mirror placement had all been designed to permit the relatively quick switches. The control system needed to include motorized controls for the auxiliary mirrors that would swing the mirrors into place and remotely lock them in precise alignment.

  Sinclair Smith had worked at a furious pace to finish the drive controls before his death. He left behind drawings and schematics, some complete, some in sketches. The system in its entirety was so complex that Sandy McDowell turned to Vannevar Bush at MIT for help. Bush, famous for his analog computer systems, recommended his best student, Edward Poitras, as the man to take over the work. McDowell, still eager for Hannibal Ford to build the drive system, suggested that the remaining work on the control system be shifted to Ford�
��s company, with Poitras, who was then with the Lombard Governor Corporation in Ashland, Massachusetts, serving as eastern liaison. McDowell assured Ford he could get security releases from the Navy Bureau of Ordnance and individual approval from the resident naval inspector for any work Ford did.

  Ed Poitras was intrigued by the idea of working on the telescope, but before he agreed, he went to talk with his mentor Vannevar Bush. He found Bush, on a steamy hot July day, at his farm in New Hampshire, perched on his tractor, mowing a field of hay. Bush took a break, got a pitcher of lemonade, and the two men talked for hours about the problems of controlling a telescope. When the day was finished, Bush’s hay wasn’t mowed, but the project had a new man to take over the work on the drive and control system.

  It is difficult to determine where Smith’s work stopped and Poitras’s began. The drive system is a complex maze of gears, servos, motors, and controls, with hundreds of miles of wiring tying the various units to one another and to the consoles for the night assistants. The heart of the system Smith and Poitras created is a tiny phantom telescope. Models that trace the movement of large systems were a fascination of early-twentieth-century technology. The control room for the Panama Canal, a technical marvel of an earlier age, included a working model that duplicated each operation of the locks: Tiny aluminum fender chains rise and fall with the movement of the controls, aluminum pointers representing the gates swing over the blue marble that represented the lock chambers, upright indicators show the positions of the rising valve stems, and indexes show the level of water in the chambers.

  The phantom telescope at Palomar is less than 1/200 the size of the real telescope, small enough to fit under the counter of the console at the head of the stairway to the Coudé room. There are no optics in the phantom telescope, but the tube, yoke, dome slit, and windscreen of the phantom mimic every movement of the huge instrument above it. At the edges of the dome slit on the phantom sensitive microswitches open and close at a contact of the phantom tube. A small recorder next to the phantom, with a typewriter ribbon, could be engaged to record automatically the right ascension, declination, sidereal time, and the guide rates on both axes, giving the observer an automated log of the observing session.

  The clean, simple appearance of the consoles the night assistant would use to control the telescope belies the complexity of the systems they control. The stark black consoles, stripped of ornament, are both a reflection of late-1930s and 1940s design and an effort to minimize mistakes in the control of the telescope. Telescope domes are cold at night, and even the most rested of observers and night assistants can begin to make mistakes after eight or ten hours of concentration at subzero temperatures. Astronomers who had rare seeing conditions wasted because of a human error in the wee hours, or who had used up valuable telescope time battling a balky instrument, saw their dreams realized with the two-hundred-inch telescope. The controls were set up with simple spinner dials and readouts. When it was time to turn to another object, the night assistant would spin the controls to the new right ascension and declination, turn on the fast-slewing drive if necessary, then engage the slow drive until a second set of indicators showed that the telescope was pointed at the object. From there the drive systems would take over.

  With no further intervention the telescope would track the object, the dome would track the horizontal movements of the telescope, and the windscreen would track the vertical movements of the telescope. For fine control, to compensate for instantaneous seeing effects in the atmosphere, the observer would still have the familiar paddle with control buttons. The entire drive and control system requires some sixty-five motors and more than four hundred miles of wiring. To maintain the electrical connections as the dome turns required more than four miles of dome slip rings. Emergency trip buttons guard against the telescope moving too far in any direction, a drop of oil pressure on the bearings, excessive acceleration, or electrical failures. Much of the wiring is concealed, making the telescope look simple and sleek. The balky and sometimes mysterious controls of the Mount Wilson telescopes, which had experienced observers like Milton Humason or Walter Baade doing bumps and grinds against tubes and clock drives to get the telescope to behave, were gone.

  By 1939 enough of the mounting and control system had been assembled that the instrument inside the dome began to look like the telescope in Russell Porter’s drawings. Astronomers who came to visit, even Walter Adams, who had been director of the largest working telescopes in the world for many years, were astonished at the size of the machine. The proportions of the dome, more classical than the taller domes at Mount Wilson, minimized its sheer size. Inside, the immense battleship-gray machine made the Mount Wilson telescopes seem toylike. Sometimes Byron Hill would make the machine perform for visitors, firing up the oil bearing pumps and drive and control systems.

  The great machine looked and worked just like a telescope, if you didn’t know that in place of a primary mirror it had a disk of concrete.

  28

  Testing

  The first optical test of the mirror was in September 1938.

  It took most of a year to grind a spherical figure into the mirror. As the shape progressed, the coarse grades of carborundum were replaced with successively finer grades of carborundum and emery. When the thirty-six-inch spherometer, an instrument to rough measure the curvature in the glass, indicated that the concave shape was a sphere of the proper radius, the grinding, the first stage in making a mirror, was done. Brownie ordered another complete scrub of the optical lab. The skirt that surrounded the disk to facilitate washing was removed; the disk was moved off the grinding machine to its storage easel; and the entire lab was hosed, scrubbed, wiped, vacuumed, rescrubbed, revacuumed, and rewiped to remove every trace of ground glass or metal fragments. Men spent days on their hands and knees searching for a single errant grit of carborundum or glass. Before the disk went back on the machine, the walls were dressed with cedar oil to make them sticky enough to capture stray dust that escaped the filters in the air-conditioning system.

  Grinding used progressively finer carborundum to remove glass from the disk. Polishing, which would carry the surface from a rough sphere to a fine one, and then slightly deeper to a paraboloid, would be done with rouge, mixed into a slurry with water. For polishing, the five-ton full-size tool was mounted on the grinding machine. The glass facing blocks on the tool were covered with pitch pressed to the shape of the curve in the disk. Optics shop workers who hadn’t ever polished a mirror discovered that the polishing, though less noisy, was even more tedious than the grinding. To prevent the tool from galling on the glass, one man and sometimes two stood on a bridge over the disk with squeegees on long poles. As the polishing tool slowly turned, they made sure the rouge stayed in suspension and that there was always a slurry of rouge and water under the polishing tool. Even an instant of dry pitch against the glass disk would leave scars. Hour after hour they would slosh the slurry under the polishing tool. They would then lift a hose, suck on the end to start the siphoning, and wash away the accumulated mud of fine abrasive and glass. Hour after hour a man on the scaffolding under the machine lubricated the main drive gear by hand.

  The full-size tool consumed fifty pounds of polishing rouge per hour, most of it washing away over the edges of the disk. The rouge was expensive. When he saw the trucks bringing barrels of rouge to the lab, John Anderson—who had learned pencil counting from George Hale—decided that they would polish with smaller tools and use the full-size tool only when they absolutely needed it.

  The routine was deadening. At least the rough grinding had showed some minuscule, visible progress. Workers who had been on the project during the grinding quit in frustration at the monotony of the polishing. Marcus Brown was fair, but he wasn’t an easy boss. Shaping the mirror was a compulsion for Brownie. He had no time for arguments. “I know what I want done,” he would say. “Do it!” By 1938 the depression was easing. There were other jobs.

  In September 1938 the d
isk was raised to a vertical position for the first optical test. Even for the largest mirror in the world, the testing procedure was straightforward. J. B. L. Foucault, who devised a pendulum experiment to demonstrate the rotation of the earth, had also developed a simple and reliable test for measuring whether a mirror had a true spherical shape. Foucault discovered that if a point source of light is shined at the center of curvature of a perfect spherical mirror, all the light is reflected back to a point. A knife edge moved so that it cuts off the light at the focus will cause the mirror to darken evenly all over. A slight movement of the knife edge across the point of focused light will cause no moving shadow in the mirror.

  For testing the two-hundred-inch mirror, a Foucault test station, with a micrometer-adjustable eyepiece, knife edge, and point source of light was set up at the far end of the room, 120 feet away. The point source focused the light through a pinhole 1/1000 inch in diameter onto the mirror. Brownie or Anderson would then move the knife edge until it just cut off the light. If the mirror was perfect, the entire diameter of the disk would suddenly go black. But no mirror is perfect. The secret to the testing procedure was to read the significance of the patterns of light and shadow created by the knife edge, to identify areas that required additional polishing. Good opticians read the shadows the way a sailor reads dark spots on the water for wind, seeing significance in patterns that escape the eye of the amateur.

  Brown and Anderson tested the mirror on Saturdays. The optical lab had been designed as much for testing the mirror as for polishing: The big five-hundred- and one-thousand-watt lights in the ceiling were set behind panels of heat-absorbing glass, and the vents in the walls and attic were oversize. The accuracy of the tests depended on the optical homogeneity of the air in the room. The air-conditioning and ventilation equipment would all be shut down one half hour before testing began, to allow the air in the room to steady itself. The heavy cork insulation on the walls and ceiling were sufficient to maintain the interior temperature, in summer or winter, while tests were in progress.