The Perfect Machine

by Ronald Florence

  The Westinghouse workmen welded a brace across the middle of the temporarily assembled horseshoe, and a huge turnbuckle across the horns to apply the needed force. The engineers tried a test run and discovered that the horseshoe expanded as it turned on the boring mill. By late afternoon sections of the horseshoe had expanded as much as 13/1000 of an inch, more than two and one-half times the permitted tolerance of 0.005 inches. And the expansion wasn’t even. Part of the horseshoe expanded by only 7/1000 inch in the course of the day.

  An engineer watching the test figured out that the problem was sunlight through the overhead skylights heating the horseshoe unevenly. The Westinghouse engineers filled reams of paper trying to chart the expansion of the steel so they could adjust the grinding wheels of the mill to compensate. When the figures weren’t reliable, they painted the skylights with dark blue paint. That helped, but to keep the temperature even enough they finally had to build a forty-six-foot-diameter sunbonnet a few inches over the horseshoe. The milling went on for weeks. Stewart Way, a Westinghouse research engineer, studied the surface of the bearing with a microscope to search out ridges and valleys until the entire surface was within the 0.005-inch tolerance. When they pulled the 318,000-pound structure off the boring mill and removed the bracing, the horseshoe “sprung” open a few hundredths of an inch to the shape that would compensate for sag. They would know if it worked when the telescope was assembled on the mountain.

  Westinghouse completed the tube for the telescope first. In press releases Westinghouse liked to point out that turbines of that size—fifty feet long and twenty-two feet in diameter—were routine stuff for the South Philadelphia Works. But this product was different. In April 1937 Westinghouse sent out invitations and press releases to announce a ceremony marking the completion of the tube. William Ladley, an employee of Westinghouse for forty-eight years, was to have the honor of tightening the last bolt. The guest of honor would be none other than Albert Einstein.

  For the great day, April 30, 1937, rows of chairs were set up on the factory floor, and a dais was erected at one side of the tube, high enough so speakers could reach out and touch the immense structure. Ormondroyd’s celluloid model of the telescope shared the front of the dais with a microphone. An audience of dignitaries and reporters were joined by newsreel cameras to hear A. W. Robertson, the chairman of Westinghouse, trumpet the potential of the great telescope: “Sometimes one almost thinks the Good Book was right when it said that man was made to be only a little lower than the angels.” The New York Times reported that Ladley “skillfully inserted the bolt and tightened it.”

  Einstein, with his familiar halo of hair and an old-fashioned stiff collar, looked uncomfortable on the dais. His remarks that day have been lost. Afterward, at a reception, he met with several of the Westinghouse engineers. Einstein was clearly awed by the sheer size of the structure. “What happens if somebody makes a mistake in manufacturing?” he asked Rein Kroon.

  Kroon, awed by the presence of the great scientist, shrugged. The answer seemed so obvious. “We build it over again.”

  Einstein winked. “My work is so much simpler. When I make a mistake I just tear up the paper I wrote on.”

  Hale had said from the beginning that the need for the big telescope was so acute that it should be built as rapidly as possible, though not by sacrificing the capabilities of the instrument. He had made that speech during Sandy McDowell’s initial interview in Max Mason’s office. McDowell, who liked nothing better than to manage a complex project against a demanding schedule, had taken Hale’s words literally. Even before he arrived in Pasadena, he had begun to organize the telescope project as if he were preparing a warship for battle. John Anderson remained the executive director of the project, but his interest was optics, and with the arrival of the mirror, he had his hands full. The construction of the observatory and the telescope was McDowell’s bailiwick, and he flung himself into it, pressing the project as if to make up the lost years of preparation. He had been there only a few days when he reported, “I have been quite upset at the slowness of things getting under way.”

  Because he had not been there for the long, frustrating days of mirror work at GE, the disappointment of the first mirror at Corning, or even the long process of designing the mounting, McDowell tended to be impatient with the astronomers and engineers. As an outsider he believed he could see beyond the parochialism of astronomers who thought their machine unique. He had supervised the construction of large gun turrets and sighting devices for the Bureau of Ships; if the telescope was larger and required higher precision, to McDowell the changes were incremental. From his years of supervising underwater research for the navy in New London during the war, he believed he knew the style and foibles of scientists. He was a great advocate of research and had publicly advocated a much-expanded research program for the navy. But he also believed that scientists needed the guidance of a man like himself to get a project finished.

  McDowell got on well with the engineers and sales managers at Westinghouse. They were accustomed to production schedules, flowcharts, organizational diagrams, and other management aids that were second nature in the navy or in industry. He got along with Byron Hill, not the easiest of men, if the two allowed each other room. Between McDowell’s acknowledgments of every sentence, which triggered the voice-controlled transmitter at the Pasadena end of the link, and the static from cars on the street outside, communications between Pasadena and Palomar were fairly imprecise during the first months. Ben Traxler later built one-hundred-watt amplifiers for both ends of the link that made radio communications reliable. But by then Hill had already implemented many of his own ideas regarding the site work.

  All power, telephone, and propane lines at Palomar were to be installed underground. The original specifications didn’t take the highly acidic soil into account. Within weeks of installation, the outer coverings of the lines rotted and the wires reacted electrolytically with one another and with the DC phone lines. The auto dialer in the powerhouse would signal the trouble by suddenly rotating without stopping, phone communication between sites on the mountain would fail, and Hill would rant about the “fool engineers” as he led another party to dig up the ground searching for the new break. Hill replaced the cables with heavy-duty neoprene-sheathed cable, but the gophers liked the shield. He ultimately switched to lead-covered cables, installed with a Byron Hill-designed machine that extruded concrete around the cable.

  Work on the mountain had to follow a lockstep flowchart, because each stage of work depended on the previous one. The million-gallon water tank was a priority, because Hill needed water to mix concrete for the footings of the observatory. When the footings were in place, Consolidated Steel, from Los Angeles, began work on the base frame for the telescope and the dome supports. Much of the work on the mountain was routine, complicated only by the remoteness of the site. San Diego County work crews were building the new road, but it was slow going, especially in bad weather. Some of the road was usable for deliveries at night, when the road crews weren’t working, but much of the heavy steel still arrived up the old Nate Harrison Grade. One load of heavy construction steel, overhanging the back of a tractor-trailer rig, tipped the trailer backward on a steep grade. A crew of workmen with the Caterpillar tractor and some chain righted the truck, turned it around, and backed it up the slope. The “zombies” prided themselves on resourcefulness.

  The dome of the observatory, another construction job for Consolidated Steel, had been the subject of a good deal of engineering study in Pasadena. Russell Porter had drawn the original dimensions for the observatory. The relatively fast focal ratio of the telescope meant that the height of the dome could be the same as the diameter, and the finished dimensions, quite accidentally, are almost identical to the dimensions of the Pantheon in Rome. Porter had strong ideas about the design of the observatory. He wasn’t a writer, preferring to articulate his ideas in sketches, but for the observatory he made an exception: “Aside from the princi
ple that a building should express the functions of the mechanism that it covers,” he wrote in a memo to the Observatory Council, “… I have felt the importance of expressing extreme simplicity along with the appearance of permanency.” He included large base moldings on the building, to make it appear that the building was solidly rooted. He had noted, as many visitors do when it is pointed out, that the dome of the one-hundred-inch telescope on Mount Wilson appears to stop a foot above the ground, which creates the odd and unpleasant sensation that it is floating. The only decorations Porter included on his design were a simple ribbing on the dome, and moldings around the entrance. “I wish to call attention to the absence of superfluous decorations. Any attempt—to me—to embellish the wall surfaces with flutings, panels, medallions, etcs. so prevalent at the present time, will be as obsolete fifty years hence as the hoop skirt and bustle are today.”

  As the telescope rotated, or slewed, to different points, the dome of the building had to rotate in perfect alignment, and with no measurable vibration. It also had to provide sufficient insulation to maintain the temperature inside during the day, so that precious hours wouldn’t be wasted each evening waiting for the telescope to reach thermal equilibrium. The engineering of the dome was assigned to Mark Serrurier. Serrurier elected a welded monocoque construction for strength and light weight, and turned to his own professor at Caltech, Romeo Martel, and Theodor von Karmann of the Aeronautics Department, for ideas on the design of what would constitute the largest welded structure ever built.

  Von Karmann, then much sought after by the aircraft industry for his engineering work, noticed the similarity of the shape of the dome to the end of a zeppelin, and turned to the Goodyear Zeppelin Company, a forerunner of the blimp company, for information on their own designs. Von Karmann added some calculations, Martel added additional ideas, and Serrurier came up with a design for a model to be fabricated by the machine shop. The model consisted of a copper hemisphere floated in mercury, with measured forces applied by levers, to test the strength of the dome structure and the shutter doors that opened on one side against the potential loading of earthquakes, wind, and snow.

  The lower portion of the dome was designed with double walls of reinforced concrete, thirty feet high, built on the load-bearing steel framework. A twelve-inch airspace between the walls would allow heated air to rise and escape from a row of vents. The steel dome above the walls was also designed with vents at the top and aluminum-foil-filled insulation panels that fit over the grid work to form a four-foot airspace. Ventilation equipment could be used to purge heated air from the spaces between the layers to speed the cooling process in the early evening.

  Consolidated Steel, working from Serrurier’s drawings, raced to get the dome supports in place before winter. Scaffolding, then steel, rose quickly, as a circle of columns went up. Byron Hill watched the first columns go up and protested. They should start with an absolutely level circle at the bottom, he argued, then machine the columns to length before they put them up, so the tracks for the dome would be exactly level. The Consolidated Steel workers pointed to the plans. Serrurier had specified push-pull bolts at the top to level the support for the dome. Consolidated’s job was to get the steel up. Someone else could level the top.

  That job fell to Hill. He had to set up a transit at the top, in midwinter. They could only align and level the bolts at night, because direct sunlight would heat the steel unevenly. One storm blew the 2-×-12-inch scaffolding planks down the mountainside. Hill cursed the desk engineers.

  Consolidated was supposed to bend and clamp the rails for the cars that would carry the dome to tolerances of one-sixteenth of an inch before they were welded. They couldn’t make the tolerances, so the rails emerged with rough welds at each joint. It would take months of hand-grinding, by workmen sitting in the cold at the top of the steel structure, to smooth the welds. When the welds were down to tolerances, the astronomers then insisted that the rails be ground all the way around until they were absolutely smooth and level. “What sort of tolerance do you mean?” Byron Hill asked. He expected the figure of one-one-thousandth of an inch, which had become their standard. For the rails, it wasn’t good enough. Any bump in the rail would affect the motion of the dome, which would in turn be picked up by the telescope. He was told to grind the rails so there was no measurable vibration when the car was moved. Jerry Dowd, at Mount Wilson, had worked out experiments to test the vibration and slippage of a dome truck on the rails.

  Johnny Kimple from the Mount Wilson shops was sent over to take charge of the grinding operation. A traveler built on an arm from a central pillar carried the belt-driven grinding machine. The crews ground the rails every day for six months before they were smooth enough. That same winter of 1936–37, the plans called for pouring the floor of the observatory. Work crews collected firewood for a makeshift furnace and coil to heat water for the concrete.

  Gradually, steadily, the building went up. Railroad-style carriages rode on the rails to carry the load of the dome. The sections of the dome, huge curved pie slices of sandwiched steel plates with lightweight girder sections between them, were fabricated in Los Angeles, trucked up the mountain, and hoisted into place. Consolidated Steel had built a crane in the center of the building to hoist the sections. Workmen aloft welded the sections together, creating the monocoque that Serrurier, von Karmann, and Martel had designed. Byron Hill and his workmen used the steel company crane and leftover materials to fabricate the walkway around the base of the dome. When it was finished, for a first trial, they hooked up the Caterpillar tractor to rotate the dome. The grinding had paid off. Workmen came from all over the mountain to stand inside the building and watch the dome; it turned so smoothly that it was hard to believe that it wasn’t the dome that was standing still and the building turning.

  Sandy McDowell, eager to find portions of the project where his own experience was relevant, dived into the dome project. The engineers and astronomers had gently dismissed his suggestions for guidance mechanisms and spotting scopes as far from the mark. Turning big structures, like a battleship gun turret, was an area he knew. He took the plans for the dome rotation to Westinghouse and GE to ask them to bid on motors to drive the dome. They studied the plans, with heavy, solid truck tires driven by DC motors turning the dome by friction, and turned him down. The draftsmen had put the motors under the gears. The motor engineers said that oil would drip onto the motors and short them. Hill agreed. He suggested that they turn the motors on their sides and use a right-angle gearbox. But McDowell liked the clean look of the motors mounted under the truck. The drawings had already been entered in a contest and won a prize.

  McDowell got the firm that built the gears for the trucks to supply DC motors with magnetic brakes. They looked like the drawings, clean, compact, a proper modern design for the greatest telescope in the world. When they were installed, the seals on the gearboxes leaked, oil dripped onto the motors, and caused arcing and charring. Using grease instead of oil on the bearings helped, but the prizewinning design was never quite right.

  John Anderson, Francis Pease, Russell Porter, and even Mark Serrurier and the other engineers at Caltech didn’t take to McDowell’s careful organizational charts with names and titles in boxes connected by lines of authority and reporting. They had worked together for years, in some cases for decades, without a formal structure. Hale had set up committees, and the more important of the committees held regular meetings, less as a forum to decide questions than as a means of keeping one another up to date on the progress of various aspects of the project. Serrurier would report on the latest tests of the tube structure while Pease reported on his work with bearings and mirror mounts. Most communication, before McDowell arrived, had been oral, with scant weekly or monthly minutes for the committee meetings.

  At a distance Sandy McDowell was a good employer. He was comfortable around men like Byron Hill, who acknowledged his authority and who spoke in the practical terms of engineering. But in Pasadena he w
as seen as a usurper, grabbing work and authority from his equals. John Anderson, still executive officer of the project, came out McDowell’s equal on the charts of authority McDowell sent out. The other scientists didn’t fare as well. Sinclair Smith, an experienced astronomer who was working full-time on the control system for the telescope, was treated as an errand boy; McDowell would send him off to the East Coast to look up contacts or to check on the progress of work at Westinghouse and Babcock & Wilcox. Francis Pease, who had as much experience with large telescopes as anyone in the world, but little respect for McDowell’s style or knowledge, was sentenced to an oblivion as the design work was gradually taken away from him and given to the engineers.

  George Hale and later Max Mason tried to smooth over the abrasions. When Mason arrived in Pasadena and took over the chairmanship of the Observatory Council, he found that he had to do “some organizing. It is necessary, I believe, in order that the peculiar abilities of some of the members of the group shall be used at their full capacity.” The real problem was that McDowell and the scientists didn’t understand one another.