by Josh Dean
The first design submitted by Wenzel, according to CIA electrical engineer Dave Sharp, was “a disaster . . . a concept that embodied the type of structure that one might expect from a high-tech missiles and space company.” It was overly complicated, heavy, and “resembled a large birdcage.” On top of that, the projected costs were well above what the CIA had budgeted. Fearing lost months, Parangosky called in Kelly Johnson to troubleshoot, and Johnson dispatched his star engineer, Henry Combs, to Ocean Systems.
Combs was “an irascible genius” who led the structures team on the A-12 Oxcart and was the man responsible for making it the world’s first titanium airplane. Combs and Johnson, who sat in on the design reviews himself, helped steer Wenzel’s ocean guys toward a far simpler design, based on a huge spine—or strongback—made of HY-100 steel. The strongback would be the largest single weldment structure ever made, and it would support the grabber arms of the capture vehicle.
The final concept had giant arms consisting of davits (think forearms) and beams (upper arms) that looked like an enormous set of metal claws. That’s how it got its nickname “the claw,” a descriptor so accurate that Parangosky forbade anyone on the program from uttering it. Should anyone see the claw, it would blow the whole cover, because a device of that design was clearly not a miner; it could logically do only one job—pick up a huge object in the ocean.
The CV wasn’t symmetrical. It had five davits on one side, which would grab the port side of the sub, and three on the other side, which would handle the starboard side. This was the direction in which the sub was listing, and the direction engineers feared the missiles might slip out during lift, so that side of the CV also had a set of beams that held a steel mesh net that would, everyone hoped, catch any missiles that slipped out.
Considering that they weren’t allowed to call it the claw, Lockheed’s engineers gave the CV a nickname—Clementine, for the old Western song about the daughter of “a miner, forty-niner.” This name didn’t have to be secret, and because it referred to mining, it actually worked in service of the cover story. Walt Lloyd wanted people in the Bay Area, and in the industry, to know that there was a secret mining machine under construction behind the razor wire in Redwood City, because no one—no one—was to see Howard Hughes’s latest, greatest toy until it was ready to plumb the deeps.
• • •
A second Lockheed base, thirty minutes away, worked on electronic systems. Many of these control systems were subcontracted to Honeywell, in Seattle, but Lockheed handled certain components itself, in particular the critical digital data link (DDL), which would provide commands to, and telemetry from, Clementine on the seafloor and the controls up on the ship. The primary engineer for the DDL design was Ray Feldman, a veteran of the Corona satellite program who had seen the mysterious John P at several key design reviews but was never important enough to actually be acknowledged.
Lockheed’s small Corona office was hidden in a building leased from Hiller Helicopters, in Menlo Park, and it was basically like a smaller version of the Skunk Works. The sign outside said ADVANCED DEVELOPMENT, HILLER HELICOPTERS, and its true nature was known by very few people.
The Hiller shop was classic Parangosky: small, quiet, and nimble. Rather than create a bureaucracy of secret keeping, Mr. P’s security team instructed contractors to use their own methods. In Menlo Park, the engineers were given no official DOD or CIA classification system; they simply went to the local stationery shop and bought a stamp that said SPECIAL HANDLING to mark sensitive materials in a way that wouldn’t be obvious to anyone flipping through their files. There were no secure lines, either. Engineers spoke on open phones, using a dictionary of code given to them by Walt Lloyd’s Corona security—the word “film,” for instance, was never used; it was only “payload”—and Lockheed workers in Sunnyvale weren’t even aware that the facility in Menlo Park existed.
Ray Feldman had spent years working on the Corona project in secret. So when Ocean Systems’ VP Ott Schick called to offer him work on a new mining project, he was excited to move out of the black world and into commercial operations that he could actually talk about. He went for an interview at Redwood City, saw the barge, and found himself right back at Hiller, this time working on the data link for a project even wilder than Corona.
• • •
Clementine needed two redundant electromechanical umbilical cables. Each one was three inches in diameter and had to be capable of sending and receiving data and controls for electrical, acoustic, command, telemetry, and hydraulic systems. It was a fantastically complex system, made up of integrated circuits, because microprocessors didn’t yet exist, and once the cables were designed, they had to be moved out to Redwood City and integrated into both the capture vehicle and the controls, which were being built inside a van that would later be slipped into the Explorer.
After surveying the country for the best company to make the umbilicals, the program office electrical engineer in charge of underwater electronics selected a small outfit in Sugar Land, Texas, that was a subsidiary of the oil services goliath Schlumberger. The engineer put on his fake beard and some glasses and flew to Houston with another Agency plant from Lockheed. They walked into the office of the company’s president looking like two hippie academics and got what one later described as “the Houston howdy.” The man had never been so poorly treated by a contractor in his career and reported back to Curtis Crooke, who called the head of Schlumberger, who called the executive who’d been so rude, to make it clear that those two fellas really needed the cables they’d asked about and he should probably be nice to them. When the two men went back, one recalled, “they did everything but slobber on us.”
The same engineer was charged with finding Clementine’s cameras, and for those he hired the electronics division of Cohu, in San Diego. These were incredible cameras, developed for use on American helicopters in Vietnam, and sensitive down to 10–5 lumens, meaning that they could capture recognizable pictures in even incredibly low light. The military was buying as many of these cameras as possible, so what was available for Azorian were technically “rejects,” but actually they worked just fine.
• • •
The complexity of Clementine’s task was difficult to imagine, even for a company that had consistently pushed the boundaries of aerospace design. The CV had to be remotely controlled from a ship three miles above, by men who needed to have a live, well-lit picture of the target and its surroundings. It had to be strong enough to endure the crushing pressure of that depth—almost eight thousand pounds per square inch—as well as the impact with the seafloor; dexterous enough to navigate forward, backward, and from side to side using a set of eight thrusters mounted to the hull; and delicate enough to grab and hold the object without damaging or dropping it. This was as challenging a task as any Lockheed or CIA engineers had ever faced, even on the A-12.
Later, the CIA would try to capture the enormity by comparison in a review. “Analogies to help grasp the complexity of the task are not completely satisfactory,” an engineer wrote. “But imagine standing on the top of the Empire State Building with a 4-by-8-foot grappling device attached to one end of a one-inch-diameter steel rope. The task is to lower the rope and grapple to the street below, snag a compact-sized car full of gold (for weight simulation, not value) and pull the car back up to the top of the building. And the job has to be done without anyone taking note of it. Mission Impossible? One might think so.”
The Glomar II’s failure to take an accurate survey of the ocean floor in 1970 meant that the engineers had no way to know what kind of muck the sub was resting in. It could be a soupy, porous material or a thicker, gummier mud. Designers would now need to provide enough lifting power so that the CV could deal with any situation. They solved it by giving the CV four massive legs—each one twelve feet in diameter with fifteen-by-sixteen-foot pads, or feet—that would rest on the ocean floor. Once controllers on the ship ha
d positioned the claw around the sub and closed its fingers, cylinders inside each leg would begin to push against the floor. These cylinders were capable of extending up to thirty-five feet using hydraulic pressure provided by 1,240 gallons of seawater pumped down from the surface in the pipe string every minute.
The four legs together were capable of 5.74 million pounds of lifting force, and once the sub was free of the muck, a controller up on the ship would activate a release that pulled pins—again, hydraulically—leaving the legs behind, on the floor, where they would remain in perpetuity, or until they rusted away to fragments and dust.
From that point, it would be up to Clementine.
What Clementine had to do required technical precision of a kind that would be difficult to accomplish on the surface, considering the size of the machines in question, but to do it under the ocean—three miles away from the men who’d be operating the device—would require the most advanced technologies on the planet, including some that would have to be invented for the operation.
The capture vehicle was outfitted with a suite of sensors, including long-range sonar, high-resolution sonar, and altitude/attitude sonar, plus eleven television cameras lit with twenty-two lights to provide the eyes for the men on the controls up on the surface. Really, Clementine needed only the acoustic or the optical system, but because the Azorian team would get only one shot at this mission, every facet of the vehicle’s operation had a redundant system to provide a backup if the primary system failed.
To locate and place itself over the wreck, Clementine carried an array of beacon transponders that were to be dropped around the target. Hydrophones on the CV would listen to these beacons, and onboard software used that data to position the CV over the target with an accuracy of one foot.
All of the audio and video data was transmitted back and forth to the ship by the two fully redundant, 18,500-foot electromechanical cables that would run along the pipe. These were two-way cables, providing up-down telemetry, video data, and acoustic sensor data, so that Clementine’s operators could receive data from the bottom and then give direction in return. The system could handle at least ninety-six hundred bits of data per second and the electrical network that ran it all was also fail-safe—the failure of any one circuit or component would not foil the mission. Every facet of this system was pushing up against the limits of existing technology.
Assembly of Clementine finished near the end of June 1972, and on the final day of the month, the last dry test of the capture vehicle was completed inside the barge. The engineers who built Clementine would also operate her, not only to conserve time, but also because the men on the controls would truly, intimately understand what they were operating.
26
Across the Airport
Up to this point, John Graham had been overseeing the ship’s design from Global Marine’s headquarters downtown. Graham believed in keeping his projects lean, and though he had pulled in engineers to handle the detail work on all of the ship’s major systems, the design team was still small, not even fifteen men. Even that, he’d often grumble, was too many people.
Graham liked to say that managing people was like growing roses, a hobby that had become a passion since he’d quit drinking. He was a gifted manager who made good (often impulsive) hires and then trusted his engineers to do what was asked of them. Every night, he’d take an armful of plans with him when he got into his convertible to drive home to Newport Beach, and the next morning, without fail, they’d all have been checked and marked up by the time he arrived, always within a few minutes of eight A.M. Mostly what Graham asked for was loyalty and hard work, and he often announced himself by barking, “I only want to see assholes and elbows,” as he strode through the engineering sections.
Being inside Global Marine did pose some challenges for Graham in terms of maintaining secrecy, but he had so far succeeded in explaining the rash of closed-door meetings by saying that the “customer” for the mining ship was very concerned about maintaining silence to prevent competition. Still, it wasn’t easy.
Once the board voted to split Global Marine Development off into a separate company, Crooke told Graham that his group would be moving from downtown to the Tishman Building, near LAX, and Graham was happy to hear it. He’d have more freedom to operate without stress, in a larger space where he’d have full autonomy, a short drive from the program office.
• • •
While the program office had both white and black sections, the new headquarters of Global Marine Development Inc., at 5959 West Century Boulevard, was fully in the open. Here, in a fourteen-story midrise identifiable from great distances thanks to the enormous TISHMAN letters just below the roofline, was the new base of John Graham’s engineering group.
Crooke rented a high floor, just about level with the 747s that would lumber past on their descents into LAX, barely a mile west. Graham had filled his team with young engineers like Chuck Cannon, who’d been hired directly out of the University of Michigan, after he read about Global Marine in a magazine and wrote the company a letter.
Cannon thought he was working on the hull of a mining ship, even after the group moved to Tishman, but shortly after the group relocated, Graham’s secretary told Cannon that “something’s come up” and that he should proceed to a phone booth on Sepulveda at a specific time and wait for instructions. When the phone rang, he followed a series of bizarre and precise steps that led him to the program office, where he sat across from Paul Ito—the security officer who’d been planted at Global Marine as the head of HR—and got the entire story. By the end of the conversation, Cannon had finished several cups of coffee and had seen numerous unbelievably detailed photos of a wrecked Soviet submarine. In subsequent days, and always without warning, Ito would appear at the Tishman office at the end of a workday to make sure no one was leaving plans out on desks overnight.
Any engineer cleared into the program got the same cloak-and-dagger entry into the secrets, but not everyone was granted knowledge of the true story. In some cases—such as those on the small administrative staff—there was simply no need to know. In others, there was some security exception, such as in the case of one of the chief electrical engineers, who was unable (or unwilling) to account for a few years of his life in China after World War II, at least in a way that satisfied Mr. P. So that engineer never knew the truth, and he wasn’t alone.
Steve Kemp, one of Chuck Cannon’s closest friends, was hired into GMDI to work on an exciting prototype mining ship and then handed the immense job of calculating weights and centers—which involved looking at every single drawing of the ship, from every angle, and determining the weight of all the steel plates, beams, and pillars, and where that weight was in relation to the ship’s geographic center. The calculations took months, and were done by hand, on desktop calculators.
Kemp was never cleared, perhaps because he was unabashedly anti-Nixon, and he saw every drawing of every piece of the ship except for one. Whenever he requested a drawing or details for the “mining machine,” Graham made an excuse for why that wasn’t possible. Instead, Kemp got weights and measures for the miner, nothing more, and those would have to do.
• • •
Arguably the two most complicated and critical components of the ship itself were the heavy-lift and heave-compensation systems. Heavy lift would deploy and hold the enormous weight of the steel pipe string, the capture vehicle, and the submarine—if the pickup was successful. The heave compensators would keep the string from being stretched and stressed as the ship moved up and down in the water; they were, in essence, gigantic shock absorbers. Graham handed those designs to a pair of mechanical engineers named Abe Person and Vance Bolding.
Both systems depended on another key feature—the gimbaled platform. For the heavy-lift concept to work, the pipe would need to remain as still as possible. Were the string to buck and whip like the ship, according to the motion of the o
cean, it would snap. That meant that the enormous A-frame derrick that held the pipe string up over the moon pool needed to be self-stabilizing, so that it would sit still no matter how much the ship was rolling around. This required the entire platform to float on giant steel bearings. Graham’s engineering deputy, Bill Skipton, gave Silent Jim McNary the job of figuring out how to design and make these bearings, and when McNary sent the specs out to various manufacturers, only one on earth had the ability to produce them—FAG Bearings, of West Germany. Graham found FAG’s solution elegant. He wanted to give them the contract, if “the customer” would allow the job to go to a foreign company.
Graham sent McNary off to share the proposal with Mr. P and one of his engineers. Prior to leaving, McNary was given a ticket by Howard Imamura, the security officer who handled travel for the program on the West Coast, and told to follow specific instructions upon arriving in Washington. He was to rent a car and check into an Arlington hotel, where a “customer representative” would be waiting. When McNary entered the room, all of the blinds had been pulled and the TV was turned up nearly to full volume. A security officer followed him inside and then swept the room for bugs. He found none, but when he discovered a hatpin under a couch cushion, the officer pinched it between his fingers, held it up to his face, then smashed it in dramatic fashion.
The CIA approved the concept and FAG got the contract to manufacture a set of ninety-seven-inch-diameter triple-ring bearings that, at thirty-five thousand pounds each, were the heaviest bearings ever manufactured. Each one had a static load rating of 10 million pounds, and the rollers inside were the size of basketballs. McNary used one to prop open his office door.
• • •