No Shortage of Dreams: ROMBUS: Reusable Orbital Module

As the 1960s decade began, NASA Marshall Space Flight Center (MSFC) had big ambitions. The Huntsville, Alabama-based facility, which left the U.S. Army to join NASA on 1 July 1960, aspired to build ever-more-powerful launch vehicles for the new U.S. space program. The Saturn family of rockets would, it was hoped, lead to the giant Nova booster, which might launch the first lunar landing mission and, eventually, lunar base hardware and spacecraft for crewed voyages to the planets.

Following President John F. Kennedy’s 25 May 1961 call for a U.S. astronaut on the Moon by 1970, however, NASA selected the Saturn V rocket as its Apollo lunar landing mission workhorse. Interest in Nova outside NASA MSFC dwindled. This was bad news for the Huntsville rocketeers; if no Saturn successor was in development as Apollo reached the Moon, they might find themselves celebrating the first lunar landing with layoffs and facility closures.

NASA MSFC’s response was to try to shape the civilian space program’s course beyond Apollo. In May 1962, it hired three companies to carry out the EMPIRE study, which looked at 1970s crewed Mars/Venus flyby missions launched on “post-Saturn” rockets (see “More Information” below). The following month, it awarded contracts to study post-Saturn rockets with capabilities similar to those planned for Nova. The post-Saturn contracts called for designs of launch vehicles that could boost up to 500 U.S. tons/454 metric tons of payload to 200-mile-high/325-kilometer-high low-Earth orbit.

One of the post-Saturn rocket study contractors was Douglas Aircraft Company of Santa Monica, California. Douglas was no stranger to large rockets; the company had won the NASA contract to build the S-IVB rocket stage, the third stage of the Saturn V and the second stage of the Saturn IB, the Saturn V’s smaller cousin.

Philip Bono headed up the 18-member post-Saturn rocket study team at Douglas. In June 1963, when he reported results of the first year of the study at the American Institute of Aeronautics and Astronautics (AIAA) Summer Meeting in Los Angeles, he was Chief Advanced Projects Engineer for Future Space Systems in the company’s Missile & Space Systems Division Advanced Launch Technology Team. 

I have spotlighted Bono’s imaginative concepts on this blog before (see “More Information” below). In 1960, while he worked for Boeing, he described a Mars expedition built around a scaled-up version of the Dyna-Soar space plane the company was developing for the U.S. Air Force. The heavy-lift rocket he designed to launch his Mars ship included plug-nozzle engines, about which more in a moment. In 1962, after he moved to Douglas, he proposed ROOST, a single-stage-to-orbit reusable launch vehicle. 

In addition to a clear view of the ROMBUS plug-nozzle engine (the turbopump exhaust port at the center of the plug and the ring of combustion chamber segments at the plug’s edge are plainly visible), this view of the ROMBUS vehicle contains one inaccuracy and one (solved) mystery. The four landing legs should not be extended while the eight yellow liquid hydrogen tanks and the conical payload shroud remain attached to the centerbody. The large nozzles shown to the left of liquid hydrogen tank 7 and to the right of liquid hydrogen tank 3 are not discussed in written sources used in this post; they are, however, labeled in a blurry illustration (not in this post) as two of four solid-propellant rocket motors for thrust augmentation. Image credit: Douglas Aircraft Company via San Diego Air & Space Museum.

Four views of the ROMBUS launch vehicle. Upper left: bottom of the vehicle. Lower left: top of the vehicle. The center and right views are self explanatory. Image credit: U.S. Patent Office.

The post-Saturn rocket Bono described to the AIAA, which he named ROMBUS, combined the plug-nozzle engine of his Mars expedition rocket with ROOST’s single-stage-to-orbit design and reusability. Of these features, the plug-nozzle engine most defined the ROMBUS design.

The plug-nozzle rocket engine is a form of aerospike rocket engine. As its name suggests, in the typical 1960s aerospike engine a trailing cone or “spike” replaced the bell-shaped nozzle of the conventional rocket engine. In effect, as its proponents liked to say, the aerospike engine turned the bell-nozzle engine inside out.

Rocket plume size is governed in large part by ambient atmospheric pressure, which in turn is governed by altitude. At launch from Earth’s surface, air pressure surrounding the plume is at its greatest. The bell-nozzle plume at launch is typically smaller than the inside of its engine bell, leading to plume instability, which in turn leads to reduced engine thrust. The plume increases in size to fill the bell as the rocket gains altitude and ambient pressure decreases, reducing instability and increasing thrust; in other words, bell-nozzle engine thrust increases toward optimum with increasing altitude. If the width of the plume should expand beyond the lip of the bell, however, instability recurs, again reducing thrust.

The typical aerospike engine is more efficient than its bell-nozzle counterpart because it maintains optimal thrust at all altitudes. Its plume, produced by multiple small thrust chambers mounted in a ring-shaped slot at the outer edge of the spike base, close to where the engine joined the body of the rocket, was shaped on its inner side by the inverted cone or concave curve of the spike surface and on its outer side by ambient air pressure. The plume and thrust level thus remained largely stable regardless of the engine’s altitude, a characteristic Bono called “altitude compensation.” He noted that the plug-nozzle engine’s constant thrust — regardless of altitude — made it well suited for use in a single-stage launch vehicle.

In a conventional bell-nozzle engine, the plume heats the engine bell, so fuel is pumped through tubes in its the walls as coolant to prevent them from melting. In an aerospike engine, the plume would heat the spike and the ring-shaped slot containing the combustion chambers, so fuel — liquid hydrogen in the designs Bono considered — would be pumped through tubes in the spike and slot walls to prevent them from melting.  

The plug-nozzle rocket engine design Bono selected for ROMBUS, based on a design developed at the Rocketdyne Division of North American Aviation as part of an advanced Saturn rocket engine program, replaced the spike with a blunt plug with curved concave sides. The plug was only about a fifth as long as the spike of an aerospike engine with the same engine diameter. A turbopump exhaust port at the center of the plug would produce a plume that would, in effect, stand in for the missing four-fifths of the spike.

In addition to making the plug-nozzle engine more compact, this approach would make it lighter and easier to cool than the typical aerospike engine. Weight saved could be applied to payload.

Further weight savings would accrue through differential-thrust steering. A rocket with a conventional bell-nozzle engine would steer by gimbaling (swiveling) the rocket engine. This would require that the engine be mounted on a complex and heavy armature including flexible plumbing. The plug-nozzle engine, by contrast, could steer its rocket by selectively decreasing thrust in its combustion chambers. For example, reducing thrust from combustion chambers on the plug-nozzle engine’s left side would cause the rocket to turn left. In addition to saving weight, the simpler plug-nozzle steering method was expected to increase reliability.

Bono described a typical ROMBUS mission. It would begin at one of four specially designed launch complexes located north of the Saturn V launch pads at NASA’s Launch Operations Center (LOC) on Cape Canaveral in Florida. (NASA LOC would be renamed Kennedy Space Center six months after Bono presented his paper, following the assassination of President John F. Kennedy.) 

The 14-million-pound/6,350,300-kilogram ROMBUS launch vehicle would ignite the 36 combustion chamber segments in its 80-foot-diameter/24.4-meter-diameter plug-nozzle engine and ramp up to full thrust. Each segment would produce 500,000 pounds/226,800 kilograms of thrust, for a total thrust at liftoff of 18 million pounds/8,164,700 kilograms.

High thrust level would make ROMBUS a very noisy launch vehicle. Extrapolating from noise levels generated during tests of the much smaller Saturn I rocket, the first of which lifted off in October 1961, Bono tentatively estimated that the four ROMBUS launch complexes would need to be built at least 2.15 miles/3.45 kilometers apart to avoid damaging each other through launch noise. Adverse weather conditions at launch — low clouds and winds — could, he cautioned, amplify and reflect sound, potentially causing damage up to 16.8 miles/27 kilometers from a ROMBUS launch complex. 

Engine noise could also damage the reusable ROMBUS rocket itself. Bono expected that repeated exposure to high noise levels during launch would subject its structure to “acoustic fatigue.” He called this “one of the principal structural problems attendant with vehicles of this size and thrust level.” 

Bono designed his ROMBUS launch complex with destructive noise in mind. Each would include twin parallel arch-supported causeways bearing three piers near their centers. Between them the piers would support a launch platform to which the ROMBUS vehicle would remain bolted until its plug-nozzle engine reached full thrust. The causeways would span a 500-foot-diameter/153-meter-diameter, 60-foot-deep/18.3-meter-deep parabolic bowl partly filled with water to form a pool about 250 feet/76.2 meters across and up to 30 feet/9.14 meters deep. 

At ignition, the bowl would scatter sound away from the ROMBUS launch vehicle centerline, reducing acoustic fatigue effects; the water, meanwhile, would be displaced by the plug-nozzle engine plume, forming an “irregular quasi-parabolic shape” that would muffle noise.

Aerial view of a ROMBUS vehicle on its launch complex. A U-shaped mobile platform crawler moves away to the upper right of the vehicle. High noise levels meant that no human could be permitted within 900 feet/275 meters of the vehicle during launch; that is, they would need to stay beyond the ring road and the structures visible at upper center and at left. Image credit: Douglas Aircraft Company via San Diego Air & Space Museum.

Bono did not say how tall the fully assembled ROMBUS vehicle would stand at launch. Drawings such as this, however, which show a facility of known size (the parabolic dish of the launch complex would be 500 feet/153 meters in diameter), indicate that it was a little less than 250 feet/75 meters tall. Image credit: U.S. Patent Office/DSFPortree.

After its engine passed a quick checkout, the ROMBUS vehicle would be released from its launch platform to begin an eastward climb over the Atlantic. Bono estimated that ROMBUS could reach orbit even if six combustion chamber segments failed during ascent. 

During ascent, the plug-nozzle engine would draw liquid oxygen oxidizer from a spherical tank within its “centerbody.” Eight 118-foot-long-by-25-foot-diameter/36-meter-long-by-7.5-meter-diameter cylindrical tanks attached to the sides of the centerbody would provide liquid hydrogen fuel.

Tanks 1 through 4, attached in pairs to opposite sides of the rocket to ensure stability, would supply all fuel during the first 130 seconds of flight; then, nearly empty, they would detach and tumble. At an altitude of about 30,000 feet/9145 meters, they would deploy parachutes, then would descend to a splashdown 34.5 miles/55.5 kilometers downrange of the ROMBUS launch complexes. Gaseous hydrogen pressure within the titanium-sandwich-skinned tanks would help to ensure that they would be sturdy enough to withstand water impact.

A ROMBUS vehicle during the final phase of its climb to orbit. Two liquid hydrogen tanks are shown immediately after separation; they are numbered incorrectly (they should be 5 and 6) and the tank at lower left has apparently experienced premature parachute deployment. The painting shows a small centerbody (with U.S. flag) and a disproportionately long cylindrical extension above it; both have reflective metal skin. Image credit: Douglas Aircraft Company via San Diego Air & Space Museum.

A NASA Landing Ship Dock retrieves the second of a pair of ROMBUS liquid hydrogen tanks. Also visible are the dome-shaped top of the first tank of the pair, the inside of the ship’s well, one open well door, and two cranes. Image credit: Douglas Aircraft Company via San Diego Air & Space Museum.

Bono proposed that NASA acquire a surplus U.S. Navy Landing Ship Dock (LSD) for recovery of the floating tanks. LSDs were designed primarily to deploy landing ships bearing soldiers. Most of the space within an LSD’s hull was taken up by a rectangular well that could be flooded. A pair of doors at the aft end of the ship opened the flooded well to the sea. Two cranes amidships, each capable of hoisting up to 50 U.S. tons/45.35 metric tons, could then be used to pull floating objects — such as ROMBUS liquid hydrogen tanks — into the well. 

After the well doors were closed, the water would be pumped out of the well so that objects taken in could be secured for transport using the cranes. Bono estimated that an LSD could retrieve and return to the LOC one pair of ROMBUS tanks at a time.

As ROMBUS ascent continued, two more opposing liquid hydrogen tanks would supply all fuel to the plug-nozzle engine, expend their contents, detach, tumble, and parachute to a splashdown 357 miles/575 kilometers downrange. The final opposing pair would separate at a low orbital altitude of 57.5 miles/92.5 kilometers and would, 19 minutes after separation from the ROMBUS centerbody, parachute into the Atlantic about 2760 miles/4440 kilometers downrange from the LOC. Bono estimated that tanks 5 through 8 would need additional thermal protection to withstand aerodynamic heating generated during their descent though Earth’s atmosphere.

Bono provided no details of ROMBUS payload deployment or other orbital operations. Instead, he skipped to a novel reentry technique he believed the plug-nozzle engine would make possible. The technique would form a major element of the ROMBUS patent NASA was granted on his behalf in January 1967.

In its 200-mile-high/325-kilometer-high orbit, the centerbody would complete 16 orbits every 24 hours. The last of these orbits each day would carry it over an elliptical ROMBUS landing area centered between Orange City, Florida, and the Atlantic coast, about 30 miles/50 kilometers northwest of the ROMBUS launch complexes. 

Every 24 hours a radio command could thus be sent that would initiate centerbody return to the LOC. Upon receiving the command, the centerbody would face its plug-nozzle engine forward and briefly fire four combustion chamber segments spaced evenly around the combustion chamber slot. This maneuver would slow the centerbody by 500 feet per second/152 meters per second. With its plug-nozzle engine still facing forward, it would then begin its descent to the landing area.

Cutaway of the ROMBUS launch vehicle centerbody. Features are identified in the black and white illustration immediately below. Image credit: Douglas Aircraft Company via San Diego Air & Space Museum.

The ROMBUS vehicle centerbody with important features labeled. Features are shown more clearly in the color image immediately above. Image credit: Douglas Aircraft Company via San Diego Air & Space Museum.

Bono proposed that the forward-facing ROMBUS plug-nozzle engine could serve double-duty as a reusable active heat shield for protecting the centerbody from reentry heating. He believed that the system that cooled the plug-nozzle engine during ascent — liquid hydrogen flowing through tubes within the plug — could also keep the engine cool during reentry. Liquid hydrogen for the deorbit burn and reentry cooling would be carried in a spherical tank within the plug-nozzle engine. He suggested that, after cooling the plug, the hydrogen could be expelled through the plug-nozzle engine combustion chambers and turbopump nozzle, protecting them from reentry heating and further cooling the outside of the plug. 

In his 1963 paper Bono provided few details of his novel reentry method. In a 1976 book Bono co-authored with spaceflight writer Kenneth Gatland, however, he explained that plug-nozzle engine cooling would activate in “low-flow” mode as the centerbody descended below 400,000 feet/122,000 meters about 11 minutes before planned landing. After three minutes, as reentry heating neared its peak, cooling would switch to “high-flow” mode for four minutes. Then, four minutes before landing, the supply of cooling hydrogen would run out. 

A drogue parachute would deploy from the top of the centerbody to decelerate it below supersonic speed and enhance its stability. At an altitude of 30,000 feet/9145 meters, the drogue would detach and four or five main recovery parachutes would deploy. 

As the centerbody fell below 2500 feet/760 meters, the parachutes would detach and four combustion chambers would ignite. These would further slow descent and and steer the centerbody toward a precise landing spot. Bono suggested that a human on the ground aided by ground-based radar might remotely pilot the centerbody in the final descent phase. 

The centerbody would extend four landing legs. At touchdown, leg compression would trigger engine shutdown. At engine stop, the centerbody would weigh 500,000 pounds (227,000 kilograms) and stand 95 feet/29 meters tall. 

The ROMBUS launch vehicle centerbody ignites four of its combustion chamber segments and discards its main recovery chutes. Parts of each of the two zones (inner and outer) of the elliptical ROMBUS landing area are visible below the centerbody. The barge port is in view on the coast to the left of the centerbody, along with the road connecting it to the inner landing area zone; in the background, to the right of centerbody, Cape Canaveral and some of its spaceflight facilities can be seen. Image credit: Douglas Aircraft Company via San Diego Air & Space Museum.

The centerbody has landed and a mobile platform crawler approaches. After the crawler picks up the centerbody, the four landing legs on the latter will be retracted. Image credit: Douglas Aircraft Company via San Diego Air & Space Museum.

The ROMBUS mission would not be ended, however, for post-flight refurbishment constituted a critical and complex phase of operations. A U-shaped mobile platform on treads would be used to collect the centerbody and carry it overland from the landing site to a port facility on the Florida coast. There it would roll onto a barge to begin a 20-mile/30-kilometer journey south along the Intercoastal Waterway to the LOC. 

At the LOC, the barge would enter a canal leading to a special ROMBUS assembly building located north of the Saturn V Vertical Assembly Building (VAB). At the time Bono presented his paper, the VAB did not yet exist; construction began on 3 August 1963, less than two months after the AIAA Los Angeles meeting. Upon arrival at the ROMBUS assembly building, the mobile platform would disembark from the barge and deliver the centerbody to a refurbishment bay.

Partial cutaway of the ROMBUS vehicle assembly building at the Launch Operations Center, Cape Canaveral, Florida. The assembly building includes three bays. In the one at left, a payload shroud is lowered onto a ROMBUS vehicle; at right, a ROMBUS vehicle has left the building on its way to one of the four ROMBUS launch complexes, the nearest of which is at least 2.15 miles/3.45 kilometers away. At upper right, a barge arrives at the canal carrying a centerbody; at upper left, a Landing Ship Dock delivers a pair of ROMBUS liquid hydrogen tanks. Image credit: Douglas Aircraft Company via San Diego Air & Space Museum.

Bono estimated normal ROMBUS turnaround time — which he defined as the time separating two launches — at 76 days. During that period, the centerbody would be partially disassembled, refurbished, reassembled, fitted with eight liquid hydrogen fuel tanks, loaded with a payload, and fitted with a launch shroud. A mobile platform crawler would then carry the launch vehicle to a ROMBUS launch complex for final preparations and a new flight to orbit.

Bono devoted two pages of his 20-page Los Angeles paper to text and charts detailing ROMBUS costs. He sought to prove that ROMBUS could yield a “four-fold improvement over current vehicles in (a) payload capability and (b) direct operating costs.” (“Current launch vehicles” referred mainly to the Saturn family rockets.) On the whole, however, his cost estimates are not convincing, in large part because he failed to define his cost estimation methodology. 

He wrote of an annual ROMBUS program cost of $1 billion a year, and a total program cost including facilities, operations, and development of between $9 billion and $17 billion. Of this total, development cost ranged from $5.1 billion to $8.6 billion. 

Cost per pound of payload to orbit varied considerably depending on which factors Bono chose to consider. For example, he estimated that a reliability of 0.85 would yield a cost of $12 per pound/$26 per kilogram; upping reliability to 0.95, as might be achieved as NASA gained experience over the course of the ROMBUS program, would reduce cost to $5 per pound/$11 per kilogram.

In his 1976 co-authored book, Bono arrived at a cost of $25 per pound/$55 per kilogram if the ROMBUS vehicle could be reused 20 times; 100 reuses would reduce cost to $10 per pound/$22 per kilogram “or less.” These estimates may reflect methodologies NASA applied in the early 1970s to generate Space Shuttle cost estimates, which would turn out to be seriously flawed.

Bono suggested that ROMBUS might be flown economically with less than its maximum payload. It would then amount to a “reusable ‘trucking’ system” that could replace existing smaller expendable launch vehicles. Flights with reduced payloads would need less liquid hydrogen fuel and liquid oxygen oxidizer; this meant that they would need fewer liquid hydrogen tank pairs. He wrote that, for a mission that would see ROMBUS launch a nuclear-thermal rocket upper stage to an altitude of 106,000 feet (32,300 meters) for a suborbital reactor start, the liquid oxygen tank in the centerbody would need only be filled halfway. 

NASA MSFC pulled the plug on the Douglas post-Saturn launch vehicle study in March 1964. This did not, however, prevent Bono from continuing to propose vehicles that resembled ROMBUS. In his 1976 co-authored book, for example, he described Hyperion, a sled-launched 55-passenger orbital crew transport; Pegasus, a suborbital 172-passenger crew/cargo transport capable of traveling 7456 miles/12,000 kilometers in just 39 minutes; and the Ithacus intercontinental troop transport, which could carry 260 soldiers and their equipment anywhere in the world. All were, like ROMBUS, reusable single-stage vehicles with plug-nozzle engines.

In June 1964, NASA MSFC director Wernher von Braun acknowledged publicly that post-Saturn launch vehicles had no future by calling publicly for future crewed planetary missions to use the Saturn V launch vehicle. By November of that year, President Lyndon Baines Johnson made clear that the NASA space program after Apollo should emphasize Earth-orbital operations and rely on rockets and spacecraft developed for the Apollo lunar landing program.

Sources

“ROMBUS – An Integrated Systems Concept for a Reusable Orbital Module (Booster & Utility Shuttle),” Douglas Engineering Paper No. 1552/AIAA Preprint No. 63-271, P. Bono, Douglas Aircraft Company; paper presented at the First National Summer Meeting of the American Institute of Aeronautics and Astronautics, Los Angeles, California, 18 June 1963.

Design No. 201,773. Recoverable Single Stage Spacecraft Booster, “James E. Webb, Administrator of the National Aeronautics and Space Administration, with respect to an invention of Philip Bono,” US Patent Office, 16 June 1964 (filed), 27 July 1965 (granted),

Patent No. 3,295,790. Recoverable Single Stage Spacecraft Booster, “James E. Webb, Administrator of the National Aeronautics and Space Administration, with respect to an invention of Philip Bono,” US Patent Office, 16 June 1964 (filed), 3 January 1967 (granted).

Frontiers of Space (Revised Edition), P. Bono and K. Gatland, MacMillan Publishing Company, 1976, pp. 63, 66, 69-70, 163-168, 197-200, 218.

Stages to Saturn: A Technological History of the Apollo/Saturn Launch Vehicles, NASA SP-4206, R. Bilstein, NASA, 1980, pp. 37, 50-60.

San Diego Air & Space Museum Image Collection (https://sandiegoairandspace.org/collection/image-collection — accessed 6 May 2025). 

More Information

Dyna-Soar’s Martian Cousin (1960)

Reusable One-stage Orbital Space Truck (ROOST) (1962)

EMPIRE Building: Ford Aeronutronic’s 1962 Plan for Piloted Mars/Venus Flybys

X-15: Lessons for Reusable Winged Spaceflight (1966)