VII

… probably to confuse the enemy …

            At a length of 78 feet, from tip of reentry vehicle to end of thrust chambers, the cold, brightly lit metal shape hanging beneath its recumbent launch boom in the chill of the narrow launch bay doubtless could create some uneasy emotions if one were to think too deeply about its actual function.  While the missileers were completely dedicated to their mission, they were also young men who understood what a war requiring the unleashing of this device might mean to the future of the entire planet, and themselves.  They understood that a war requiring the unleashing of this device would change the future into something — unfortunate.

            After adjusting to the idea of being in the presence of a mindless machine with such god-like potential for wrath, the second thing any visitor to the bay would notice was the cleanliness of the surroundings — to military specifications.  As throughout the missile complex, the unpainted concrete floors in the bay had been sealed and waxed.  No oil stands on the floor.  No loose litter or tools.  Nothing out of place or unsecured.

            Jack Roberts recalls, “It was quiet in the launch bay — none of the echoes you might ordinarily expect in a concrete building.  I guess all the equipment in there didn’t give sound a chance to bounce around.  And in the winter it tended to be cold.  I don’t recall any heating in the bay.”

The machine itself was clean.  The bullet-like cleanliness of its airframe was designed to slip through the lower, denser atmosphere with little resistance.  Most components fixed to the missile outside the line of its bullet shape were housed in equipment pods covered by aerodynamic nacelles.

1st Lieutenant Eldon Wilford in the launch bay of Fairchild's Launch Complex #3, located east of Rockford.

        The stabilizing fins of classical rocket design were gone, unneeded since all steering corrections were made by directing the thrust from the three main and two steering engines.  Machine senses, far quicker than human senses, would keep the rocket standing squarely on the column of its own thrust as it pushed skyward. 

            Standing is a word describing our conditioned view of a rocket at rest or in flight.  But a more proper way to think of a rocket, at least the complex ones designed as ICBMs or spacecraft, is as a boat.  Because many of them do stand upright at times, and then spend part of their time in an environment where up and down are little more than mathematical coordinates, it can become confusing.  Though bow and stern easily convert to top and bottom when the rocket is upright, the confusion begins when having to think in terms of right (starboard) and left (port), or dorsal (upper side or back) and ventral (lower side or belly).  And then more confusion is add by nautical terms indicating motions, terms such as yaw, pitch, and roll.  Perhaps the vessel most closely related to the rocket is the submarine, since both operate in three dimensional oceans.

            Regardless of such comparisons, the machine lying in the launch bay was most certainly a sleeping shark-like predator.  Perhaps at least part of the quiet chill the missileers felt it its presence was related to that fact.

            At the nose of the rocket proper was a 48 inch diameter, four foot deep, conical reinforced adapter ring that provided attachment points for the reentry vehicle.  Recessed inside this ring was the domed forward bulkhead of the 18,600 gallon liquid oxygen tank.

            From the leading edge of this adapter ring, and over the next fifteen and a half feet, the missile’s diameter flared at an angle of 10 degrees to the full ten-foot diameter of its two tanks.  Though the length of the oxygen tank was just over forty feet, the aft fuel tank’s domed bulkhead protruded into the interior of the forward oxidizer tank almost four feet — reflecting the fact that the lower tank, full or empty, was always maintained under higher pressure than the upper tank.

            The hardened adapter ring used to mount the warhead was also the point of suspension from which the recumbent launch boom grappled the forward part of the missile.  This boom attachment was also the point at which mechanical stretch could be applied to the airframe if depressurization, by intent or accident, occurred.

            Documents indicate that the thickness of the liquid oxygen tank’s stainless steel skin varied from 0.012 of an inch to 0.038 of an inch — roughly the thickness of one to four standard playing cards.  This is why, without internal pressure or mechanical stretch, the airframe would collapse under its own weight.

            The interior of the LOX tank, as well as the propellant tank behind it, contained a series of ultrasonic sensors running down the full length of each tank.  Each sensor, as the fuel level fell below it, sent a signal to the propellant utilization system.  If the pre-programmed usage ratio between the two tanks became unsynchronized, valves were manipulated to increase flow from one tank and decrease flow from the other tank until the LOX/RP-1 usage returned to the correct balance.  If one or more of those sensors needed attention, the Missile Maintenance Technicians had to enter the tank.

            “Entry into either the liquid oxygen or the propellant tank was a pretty rare event,” Jack reported.  “I was only inside the tanks three or four times.  Because of the extra difficulty of entering the rear tank — the central engine with all its attachments had to be removed — that was always done at the base maintenance shop.  But we could enter the front tank with the missile still hanging beneath the launch boom at the bunker.”

            “First the weapons team removed the reentry vehicle from the rocket.  We put a work platform at the nose, and covered the overhead, sides, and floor of our work area with 6 mil. plastic, forming a contamination control tent of sorts.  A plastic sheet divider was dropped to isolate the adaptor ring from the rest of the tent.”

            “To gain access we had to remove the tank’s boil-off valve.  This valve was a large apparatus situated in the center of the domed front of the tank.  Its function was to vent oxygen vapor from the missile.  When upright and fueled, liquid oxygen, growing warmer and evaporating out of its liquid state, would buildup the vapor pressure inside the tank.  If this wasn’t vented, the pressure could build high enough to rupture the tank.  The white plume seen drifting away from the top of the missile during fueling is gaseous oxygen escaping through this valve and chilling atmospheric moisture into visible condensation.”

            When liquid oxygen, under normal barometric pressure, warms up to 297 degrees below zero Fahrenheit, it begins to boil.  As it turns into a gas, again under normal barometric pressure, it expands to 860 times its original liquid volume.  Though the LOX in the tank wasn’t boiling, the increase in evaporative vapor pressure due to the oxygen being pumped into the tank’s much warmer environment was doubtless substantial.

            “The boil-off valve was closed just before launch.  At launch., high pressure nitrogen was vented into both tanks to help force the contents toward the engine pumps.”

            “To remove this valve, we would first put the missile into stretch, then release the tank’s internal pressure.”

            “The valve was about two and a half feet in diameter.  There were over a hundred screws holding it in place.  We didn’t have electric or air-operated screwdrivers.  Every screw had to be broken loose and threaded out by hand.  And when replacing, each hand driven, and individually torqued.”

            “The valve didn’t weigh much more than 25 pounds, but it was so awkward it took two of us to handle it.  After removal, we wrapped it in plastic to prevent contamination.  And then we could slide through the opening into the tank.”

            “Working inside the oxygen tank was a real pain.  We had to wear moon-suits and breathing apparatus — partly to protect us from the chemicals we were using, and partly to protect the tank from us.  And then there were air and safety lines to contend with.”

            “In mechanical stretch, the tanks deformed, developing very large wrinkles.  About as close as I can come to describing the sensation of walking around the interior is to say that it was like stepping on a big, curved steel waterbed.  If we needed to get higher up the walls, we used scaffold boards, with the ends heavily padded and wrapped in plastic, to span from side to side of the tank.”

            “Much of this contamination prevention was due to the specific worry that hydrocarbons might be introduced into the interior of the oxygen tank.  Hydrocarbons — petroleum products — and liquid oxygen do not react well together.”

            “When the maintenance was done, we wiped down the interior of the tank with trichloroethylene, and tested a randomly selected area with a small piece of filter paper.  We sealed this test patch in a plastic bag and passed it through the boil-off valve opening to a lab technician.  He looked for contamination on the patch through a microscope.  If it showed more than so many particles per square inch, regardless of what those particles might be, we had to wipe the whole thing down again.”

            “Working in the RP-1 tank — the rocket propellant tank — was not such a clean procedure.  There was always and inch or two of residual fuel sloshing around.  We just wore rubber boots and waded.  Getting into the tank was the headache.”

            The propellant tank, almost 25 feet long, held 11,455 gallons of fuel.  The thickness of its skin varied between 0.027 and 0.038 of an inch.  The intermediate bulkhead, the forward end of the fuel tank, bulged about four feet into the aft end of the liquid oxygen tank.  This intermediate bulkhead was insulated to keep the 300 plus degree below zero LOX from being heating by the fuel — and the super-refined, kerosene based fuel from being chilled into jell by the LOX.

              The power from the two booster engines was transferred to the body of the rocket through a hardened thrust ring girdling the bottom of the fuel tank.  Thrust from the central sustainer engine was transferred through a thrust ring on the truncated bottom of the fuel tank’s downward bulging lower bulkhead.  This also formed the access portal into the fuel tank — and was the reason the central engine had to be removed to gain access to the tank.

            Standing at the rear of the missile, looking at the horizontal line of thrust chambers, the second most innovative design feature of the Atlas missile, after inflatable propellant tanks, was not readily apparent.

            Classic rocket design called for the use of multi-stage vehicles — each stage a complete rocket consisting of fuel tanks and engines.  The Atlas missile’s stage and a half design, at stage separation, only jettisoned the unneeded engines and engine support equipment.  The entire aft end of the rocket, minus the sustainer engine, was designed to slide away.  Both outside booster engines would shut down, but the central sustainer engine would continue burning.  Latches holding the booster section on would slam open, locking pins on hoses and connections would snap open, and the thrust from the sustainer would push the missile away from the loose booster.  The booster section, with a gaping hole in its center where the sustainer engine had once rested, would drop away.

            The coordinate system for knowing where things were located around the circumference of the missile could best be visualized at the rear.  Two lines were drawn — one horizontally through the centers of all three thrust chambers — the other vertically through the center of the sustainer’s thrust chamber.  This divided the missile, down its long axis, into four quarters or quadrants.

            “The quadrants were numbered one through four clockwise, starting at the lower right,” Airman Roberts said.  “That’s why the booster engine on the left, centered on the line between quadrants two and three, was called the B-2 engine — booster-2 engine.  The equipment pod located along that same line just forward of the booster section — the pod containing the computer, inertial guidance system, inverter, batteries, and such — was called the B-2 pod.  I don’t know why they used this particular grid system — probably to confuse the enemy.”

            Most of the equipment inside the thrust section of the rocket was intended to serve one purpose — to turn the liquid oxygen and RP-1 pumped from the forward tanks into enough thrust to lift and accelerate the machine — its fuel and its warhead — into a ballistic trajectory toward Russia.  This was not a simple task.

            The booster engines each produce 165,000 pounds of thrust.  The central sustainer produces 57,000 pounds.  The two small vernier steering rockets, both located forward of the booster section — one on the dorsal (upper) side of the recumbent missile, along the line dividing quadrants three and four, and the other on the ventral (lower) side, on the line between the first and second quadrants — each produced one thousand pounds of thrust.  All together, this gave the missile 389,000 pounds of lift.

            To get that lift, each booster engine consumed something over 75 gallons of LOX/RP-1 mixture a second.  That rate was maintained until booster shutdown and jettison 140 seconds into the flight.  The sustainer, during it’s 270 seconds of burn, volatized about 25 gallons a second.  And the verniers, operable for a time period comparable to the sustainer, each burned about half a gallon of mixture a second.  These calculation, by BMAT Bob Lemley, assume that only a small safety margin of LOX/RP-1 would have been left in the tanks at shutdown.

            Each gallon of liquid oxygen weighed 9.52 pounds.  Each gallon of RP-1 — slightly lighter than kerosene — weighed 6.625 pounds.  The fuel and oxidizer at takeoff weighed slightly over a quarter million pounds.  So the combined weight of missile, warhead, and fuel came in at just under 275,000 pounds — or just over 137 tons.

            The classic problem with two chamber rocket engines — an upper combustion chamber and lower thrust chamber — is heat.  The scorching blast of flame rushing out of the combustion chamber and down the throat of the rocket’s thrust chamber tends to erode away any metal softened by that heat.  In the thrust chamber this erosion can be slowed by cooling the metal in the chamber’s walls through a process called ‘regenerative cooling’.

            As Bob Lemley explained, “The thrust chambers were made of square, hollow tubes welded together.  Every other tube was either a ‘down tube’ or a ‘up tube”.  The RP-1, as it came from the turbopump, was sent down the ‘down tubes’ into a collector ring around the outside-bottom of the rocket tube.  The fuel returned from the collector ring though the ‘up tubes’.  This liquid, moving at a very high speed, carried the heat away from the thrust chamber’s walls.  It was then directed into the showerhead injector nozzle at the top of the engine’s combustion chamber.”

            Positioned at dead-center/top of the combustion chamber, the injector plate consisted of concentric rings of injection holes.  Each ring alternated between RP-1 and liquid oxygen.  Besides cooling the surface of the plate, the fuel and oxidizer created a laminar-flow boundary layer as they entered the chamber that further protected the injector plate from burn-through.

            The outside row of holes in the showerhead propellant injector sprayed streams of PR-1 onto the walls of the combustion chamber — a process called ‘spray cooling’.

            Jim Geoghegan added, “The combustion chamber was large enough to disperse the combustion pressure over a large area.  That’s why the turbopumps could push LOX/RP-1 into the chamber at a pressure less than the total thrust being developed by the engine.”

            When used as a rocket’s regenerative cooling agent, the heavy form of kerosene burned in jet planes proved inadequate.  The scorching heat inside the thrust chamber’s cooling tubing caramelize elements out of the jet fuel.  These varnishes tended to collect as sludge in the regenerative tubing and in the showerhead injector plate, clogging the passages, and causing hot spots, burn through, and catastrophic engine failure.  RP-1 — a highly refined fractional form of kerosene — provided a low varnish solution.

            Attempting to spray RP-1 and liquid oxygen together into the combustion chamber before combustion had been initiated caused another set of problems.  As Colonel John Voss pointed out, “I remember a training film presented by engineers from General Dynamics Aerospace.  They took minute amounts of rocket propellant and liquid oxygen, mixed them, and remotely dropped a weight on top.  It sounded like a shotgun going off.  Liquid oxygen and petroleum products are not compatible.  When mixed, the least shock will detonate the concoction.”

            Spray induced turbulence inside the combustion chamber would certainly constitute a shock.  The volume of fuel and oxidizer injected in just a fraction of a second, if shock detonated, would doubtless be enough to blow the chamber apart.  To prevent this, a healthy flame front had to be built at the very beginning of chamber injection.  Creating this front required the sequenced injection of a hypergolic agent.

            A hypergol is a chemical that remains non-reactive to one element of a propellant, yet spontaneously ignites upon contact with another.  In the Atlas, the hypergolic charge, or ‘slug’, was added as a ‘burst cartridge’ — a tube filled with the chemical, then capped at both ends with a diaphragm designed to ‘burst’ under pressure.  This cartridge was installed in a by-pass from the main fuel line.  A portion of the initial RP-1 flow was diverted into this line.  Pushing through the first ‘burst’ diaphragm, the fuel flow forced the slug of hypergol through the second diaphragm, then through the showerhead injector and into the combustion chamber.  Meeting the incoming spray of liquid oxygen, the hypergolic immediately flashed into flame.  The fuel following the slug maintained the combustion until the main flow of RP-1, having first circulated though the thrust chamber’s regenerative cooling loop, entered the showerhead.  As it sprayed into the combustion chamber, it found a sustained flame waiting.

            Three high-speed turbopumps, one plumbed into the workings for each main engine, drew down the approximately 175 gallons of liquid needed by the missile each second all engines were under power.  The two steering rockets — vernier engines — were plumbed into the turbopump supplying the central sustainer engine.  That way, when the two booster engines dropped away at staging, the steering rockets, along with the sustainer engine, would continue to function.

            Turbine style impellers gave the turbopump its name.  In the missile, each turbopump has three impellers working off a common driveshaft axle.  The impellers on each end were fluid pumps — one for liquid oxygen, the other for the kerosene based fuel.  The center impeller in each turbopump used a high-speed stream of combustion gas, ducted over its blades, to spin the pump’s common axle and drive the impellers on either side.

            To produce this high-speed jet of combustion gas, a gas generator — essentially a small rocket engine in itself — was installed upstream of each turbopump.  Since the same LOX and RP-1 used to drive the big engines was burned in these gas-generators, the problem became how to supply fuel to the gas-generators to produce the exhaust to spin the turbines to pump the fuel into the gas-generators.  The answer was a solid propellant gas generator ignited by an electrically detonated squib.

            Ducting, essentially a closed causeway for exhaust from the gas-generator, ran aft from the generator, through the turbopump’s central impeller, then out an exhaust pipe at the base of the rocket.  A solid propellant gas-generator canister was fixed to the upper portion of this causeway.  Upon ignition, the solid fuel produced the initial blast of pressure needed to spin the turbine to speed.  As the solid propellant gas-generator ignited, pyrotechnic igniters inside the combustion chamber of the liquid fuel gas-generator also ignited.  These slow burning igniters insured that when the LOX and RP-1 being forced forward from the turbopumps reached the liquid fuel gas-generators, there would already be an adequate flame inside the combustion chambers to ignite the incoming mixture before turbulence-induced shock, and a resultant explosion, could occur.

            By time the solid fuel inside the propellant canisters was exhausted, the liquid fuel generators were fully operational.

            Clearly visible at the rear of the rocket, stubby gas-generator exhaust pipes angled out-board form the two booster engines.  A perforated muffler wrapped around the base of the center engine’s thrust nozzle allowed exhaust from the sustainer’s gas-generator to escape.

            During certain procedures, such as Dual Propellant Loading Exercises or Operational Readiness Inspections — times when LOX and RP-1 were pumped into the rocket’s tanks — the solid propellant gas-generators and various pyrotechnic squibs and chemical hypergolics would be removed to prevent even the remotest chance of this engine ignition ladder sequence being accidental started, and one or more of the rocket engines igniting.

            With the rocket upright, the blast from the engines would be directed into the flame pit directly beneath the base of the launch boom/gantry.  From there, the blast would be turned and channeled south through the horizontal flame tunnel.  At the end of this tunnel, the blast was angled upward into the open.  When the complex was sealed, the tunnel’s outside opening was capped by a sliding hatch designed to keep explosions, possibly nuclear, occurring above the complex from entering the launch bay through the flame tunnel.

            During the four year operational life of the Atlas E series ICBM, no flame tunnel ever sprouted fire, and no launch bay endured the ferocious blow-back from an actual launch.  But for two unrelenting weeks deep in the Autumn of 1962, as the nation teetered on the nervous edge of DEFCON 2, the launch crews of the 567^th Strategic Missile Squadron waited in near certainty for exactly that.