11 February 2016

What If Galileo Had Fallen to Earth? (1988)

Galileo awaits its chance to fly. Image credit: NASA
The U.S. Congress authorized new-start funding for the Jupiter Orbiter and Probe (JOP) on 19 July 1977, early in the Administration of President Jimmy Carter. When JOP development began officially on 1 October 1977, at the start of Fiscal Year 1978, NASA planned to launch the new robot explorer in January 1982 on STS-23, the 23rd operational flight of the Space Transportation System (STS). At the time, NASA still maintained the hopeful fiction that the STS could begin a series of six Orbital Test Flights in early 1979 and become operational in May 1980. Until 1986, the STS - the centerpiece of which was the Space Shuttle - was intended to replace all other U.S. launch vehicles.

At liftoff, the Shuttle stack comprised twin reusable Solid Rocket Boosters (SRBs), a reusable piloted Orbiter with a 15-by-60-foot payload bay and three Space Shuttle Main Engines (SSMEs), and an expendable External Tank (ET) containing liquid hydrogen and liquid oxygen propellants for the SSMEs. The STS also included upper stages for boosting spacecraft carried in the Orbiter's payload bay to places beyond its maximum orbital altitude. Until the mid-1980s, many in NASA hoped that a reusable Space Tug - perhaps incorporating a propellant-saving aerobrake - would eventually replace the expendable upper stages.

At the start of STS-23 (and, indeed, at the beginning of all STS missions), the three SSMEs mounted on the Orbiter's tail and the twin SRBs bolted to the side of the ET would ignite in sequence to push the Shuttle stack off the launch pad. SRB separation would then take place 128 seconds after liftoff at an altitude of about 155,900 feet and a speed of about 4417 feet per second.

The three SSMEs would operate until 510 seconds after liftoff, by which time the Orbiter and its ET would be moving at about 24,310 feet per second at an altitude of 362,600 feet above the Earth. The SSMEs would then shut down and the ET would separate, tumble, and break up as it fell into dense atmospheric layers over the Indian Ocean.

The Orbiter, meanwhile, would ignite its twin Orbital Maneuvering System (OMS) engines at apogee (the high point in its Earth-centered orbit) to raise its perigee (the low point in its orbit) above 99.99% the Earth's atmosphere. By the time it completed its OMS maneuvers, the STS-23 Shuttle Orbiter would circle the Earth in a 150-nautical-mile-high low-Earth orbit (LEO).

The STS-23 crew would next open the Orbiter payload bay doors and release JOP and its three-stage solid-propellant Interim Upper Stage (IUS). After they maneuvered the Orbiter a safe distance away, the IUS first-stage motor would ignite to begin JOP's two-year direct voyage to Jupiter.

Early days: artist concept of Jupiter Orbiter and Probe. Image credit: NASA
In February 1978, NASA gave JOP the name Galileo. Largely because of its reliance on the STS, Galileo suffered a series of costly delays, redesigns, and Earth-Jupiter trajectory changes. The first of these was, however, not the fault of the STS. As Galileo's design firmed up, it put on weight, and was soon too heavy for the three-stage IUS to launch directly to Jupiter.

In January 1980, NASA decided to split Galileo into two spacecraft. The first, the Jupiter Orbiter, would leave Earth in February 1984. The second, an interplanetary bus carrying Galileo's Jupiter atmosphere probe, would launch the following month. They would each depart LEO on a three-stage IUS and arrive at Jupiter in late 1986 and early 1987, respectively.

In late 1980, under pressure from Congress, NASA opted to launch the Galileo Orbiter and Probe out of LEO together on a liquid hydrogen/liquid oxygen-fueled Centaur G-prime upper stage. Centaur, a mainstay of robotic lunar and planetary programs since the 1960s, was expected to provide 50% more thrust than the three-stage IUS. Modifying it so that it could fly safely in the Shuttle Orbiter's payload bay would, however, delay Galileo's Earth departure until April 1985. The spacecraft would arrive at Jupiter in 1987.

Another delay resulted when David Stockman, director of President Ronald Reagan's Office of Management and Budget, put Galileo on his "hit list" of Federal government projects to be scrapped in Fiscal Year 1982. The planetary science community campaigned successfully to save Galileo, but NASA lost the Centaur G-prime and three-stage IUS.

In January 1982, NASA announced that Galileo would depart Earth orbit in April 1985 on a two-stage IUS with a solid-propellant kick stage. The spacecraft would then circle the Sun and fly past Earth for a gravity-assist that would place it on course for Jupiter. The new plan would add three years to Galileo’s flight time, postponing its arrival at Jupiter until 1990.

In July 1982, Congress overruled the Reagan White House when it mandated that NASA launch Galileo from LEO on a Centaur G-prime. The move would postpone its launch to 20 May 1986; however, because the Centaur could boost Galileo directly to Jupiter, it would reach its goal in 1988, not 1990. NASA designated the STS mission meant to launch Galileo STS-61G.

Artist concept of Galileo on a Centaur G-prime stage. Image credit: NASA
There matters rested until 28 January 1986, when, 73 seconds into mission STS-51L, the Orbiter Challenger was destroyed. A joint between two of the cylindrical segments making up the Shuttle stack's right SRB leaked hot gases that rapidly eroded O-ring seals. A torch-like plume formed and impinged on the ET and the lower strut linking the ET to the SRB. The plume breached and weakened the ET's liquid hydrogen tank, causing the strut to separate. Still firing - the SRBs were not designed to be turned off once ignited - the right SRB pivoted on its upper attachment and crushed the ET's liquid oxygen tank. Hydrogen and oxygen mixed and ignited in a giant fireball.

Despite appearances, Challenger did not explode. Instead, the Orbiter began a tumble while moving at about twice the speed of sound in a relatively dense part of Earth's atmosphere. This subjected it to severe aerodynamic loads, causing it to break into several large pieces. The pieces, which included the crew compartment and the tail section with its three SSMEs, emerged from the fireball more or less intact. The mission's main payload, the TDRS-B data relay satellite, remained attached to its two-stage IUS as Challenger's payload bay disintegrated around it.

The pieces arced upward for a time, reaching a maximum altitude of about 50,000 feet, then fell, tumbling, to crash into the Atlantic Ocean within view of the Shuttle launch pads at Kennedy Space Center, Florida. The crew compartment impacted 165 seconds after Challenger broke apart and sank in water about 100 feet deep.

NASA grounded the STS for 32 months. During that period, it put in place new flight rules, abandoned potentially hazardous systems and missions, and, where possible, modified STS systems to help improve crew safety. On 19 June 1986, NASA canceled the Shuttle-launched Centaur G-prime for reasons of safety. On 26 November 1986, it announced that a two-stage IUS would launch Galileo out of LEO. The Jupiter spacecraft would then perform gravity-assist flybys of Venus and Earth. On 15 March 1988, NASA scheduled Galileo's launch for October 1989, with arrival at Jupiter to follow in December 1995.

One month after NASA unveiled Galileo's newest flight plan, Angus McRonald, an engineer at the Jet Propulsion Laboratory (JPL) in Pasadena, California, completed a brief report on the possible effects on Galileo and its IUS of a Shuttle accident during the 382-second period between SRB separation and SSME cutoff.

McRonald was not specific about the nature of the "fault" that would produce such an accident, though he assumed that the Shuttle Orbiter would become separated from the ET and would tumble out of control. He based his analysis on data provided by NASA Johnson Space Center in Houston, Texas, where the Space Shuttle Program was managed.

The Space Shuttle was by far the largest spacecraft to launch with astronauts on board. It was immensely capable - but with capacity came complexity, making it vulnerable. Image credit: NASA
McRonald also examined the effects of aerodynamic heating on Galileo's twin electricity-generating Radioisotope Thermoelectric Generators (RTGs). The RTGs would each carry 18 General Purpose Heat Source (GPHS) modules containing four iridium-clad plutonium dioxide pellets each. The GPHS modules were encased in graphite and housed in protective aeroshells, making them unlikely to melt following an accident during Shuttle ascent. In all, Galileo would carry 34.4 pounds of plutonium.

McRonald assumed that both the Shuttle Orbiter and the Galileo/IUS combination would break up when subjected to atmospheric drag deceleration equal to 3.5 times the pull of gravity at Earth's surface. Based on this, he determined that the Orbiter and its Galileo/IUS payload would always break up if a fault leading to "loss of control" occurred after SRB separation.

The Shuttle Orbiter would not break up immediately after loss of control occurred, however. At SRB separation altitude, atmospheric density would be low enough that the spacecraft would be subjected to only about 1% of the drag that tore apart Challenger. McRonald determined that the Shuttle Orbiter would ascend unpowered and tumbling, attain a maximum altitude, and fall back into the atmosphere, where drag would rip it apart.

He calculated that, for a fault that occurred 128 seconds after liftoff - that is, at the time the SRBs separated - the Shuttle Orbiter would break up as it fell back to 101,000 feet of altitude. The Galileo/IUS combination would fall free of the disintegrating Orbiter and break up at 90,000 feet, then the RTGs would fall to Earth without melting. Impact would take place in the Atlantic about 150 miles off the Florida coast.

For an intermediate case - for example, if a fault leading to loss of control occurred 260 seconds after launch at 323,800 feet of altitude and a speed of 7957 feet per second - then the Shuttle Orbiter would break up when it fell back to 123,000 feet. Galileo and its IUS would break up at 116,000 feet, and the RTG cases would melt and release the GPHS modules between 84,000 and 62,000 feet. Impact would occur in the Atlantic about 400 miles from Florida.

A fault that took place within 100 seconds of planned SSME cutoff - for example, one that caused loss of control 420 seconds after launch at 353,700 feet of altitude and at a speed of 20,100 feet per second - would result in an impact far downrange because the Shuttle Orbiter would be accelerating almost parallel to Earth's surface when it occurred. McRonald calculated that Orbiter breakup would take place at 165,000 feet and the Galileo/IUS combination would break up at 155,000 feet.

McRonald found (somewhat surprisingly) that, in such a case, Galileo's RTG cases might already have melted and released their GPHS modules by the time the Jupiter spacecraft and its IUS disintegrated. He estimated that the RTGs would melt between 160,000 and 151,000 feet of altitude. Impact would occur about 1500 miles from Kennedy Space Center in the Atlantic west of Africa.

Impact points for accidents between 460 seconds and SSME cutoff at 510 seconds would be difficult to predict, McRonald noted. He estimated, however, that loss of control 510 seconds after liftoff would lead to wreckage falling in Africa, about 4600 miles downrange.

McRonald summed up his findings by writing that Galileo's RTG cases would always reach Earth's surface intact if an accident leading to loss of control occurred between 128 and 155 seconds after liftoff. If the accident occurred between 155 and 210 seconds after launch, then Galileo's RTG cases "probably" would not melt. If it occurred 210 seconds after launch or later, then the RTG cases would always melt and release the GPHS modules.

STS flights resumed in September 1988 with the launch of the Orbiter Discovery on mission STS-26. A little more than a year later (18 October 1989), the Shuttle Orbiter Atlantis roared into space at the start of STS-34. A few hours after liftoff, the Galileo/two-stage IUS combination was raised out of Atlantis's payload bay on an IUS tilt table and released. The IUS first stage ignited a short time later to propel Galileo toward Venus.

Free at last: Galileo and its two-stage IUS shortly after release from the Space Shuttle Orbiter Atlantis, October 1989. Image credit: NASA
Galileo passed Venus on 10 February 1990, adding nearly 13,000 miles per hour to its speed. It then flew past Earth on 8 December 1990, gaining enough speed to enter the Main Belt of asteroids between Mars and Jupiter, where it encountered the asteroid Gaspra on 29 October 1991.

Galileo's second Earth flyby on 8 December 1992 placed it on course for Jupiter. The spacecraft flew past the Main Belt asteroid Ida on 28 August 1993 and had a front-row seat for the Comet Shoemaker-Levy 9 Jupiter impacts in July 1994.

Flight controllers commanded Galileo to release its Jupiter atmosphere probe on 13 July 1995. The spacecraft relayed data from the probe as it plunged into Jupiter’s atmosphere on 7 December 1995. Galileo fired its main engine the next day to slow down so that Jupiter's gravity could capture it into orbit.

Artist concept of Galileo in communication with its Jupiter atmosphere probe. Blue dots linking the low-gain antenna and Jupiter represent radio signals. After the spacecraft's large main antenna jammed partly open (left), the low-gain antenna, much less powerful, became Galileo's link with Earth. Image credit: NASA
Galileo spent the next eight years touring the Jupiter system. It performed gravity-assist flybys of the four largest Jovian moons to change its Jupiter-centered orbit. Despite difficulties with its umbrella-like main antenna and its tape recorder, it returned invaluable data on Jupiter, its enormous magnetosphere, and its varied and fascinating family of moons over the course of 35 orbits about the giant planet.

As Galileo neared the end of its propellant supply, NASA decided to dispose of it to prevent it from accidentally crashing on and possibly contaminating Europa, the ice-crusted, tidally warmed ocean moon judged by many to be of high biological potential. On 21 September 2003, the venerable spacecraft dove into Jupiter's turbulent, banded atmosphere and disintegrated.


Galileo: Uncontrolled STS Orbiter Reentry, JPL D-4896, Angus D. McRonald, Jet Propulsion Laboratory, 15 April 1988

Mission to Jupiter: A History of the Galileo Project, NASA SP-2007-4231, Michael Meltzer, NASA History Division, 2007

More Information

A 1974 Plan for a Slow Flyby of Comet Encke

What If an Apollo Saturn Rocket Exploded on the Launch Pad? (1965)

What Shuttle Should Have Been: NASA's October 1977 Space Shuttle Flight Manifest

Where to Launch and Land the Space Shuttle? (1971-1972)


  1. If Galileo's capabilities were re-made with today's technology how much less mass/more capable would it be?

    When I think about my 1980s era computer vs what I now have, and relate this to what the Pioneer/Voyager, Galileo/Cassini technologies have been, I can't but imagine the geometric increase in capability. I think we saw some of that with New Horizons. Looking forward to Juno.

  2. Ben:

    That capability jump began when SDI tech started dribbling over into the space program. That was the real origin of the Discovery Program, I'm convinced of it. New Millennium, the tech development program, should never have become burdened with so much science. As soon as things went wrong - which is to be expected in a tech development program - confused people called it a failure and ended it. Of course, the fact that it might have developed tech that would have enabled more exploration was enough for some anti-exploration people to turn against it.

    I mention those programs because they had such short turn-around times - from concept to mission in a couple of years, usually. That has, naturally, stretched out some, and the cost has gone up. I don't think it's truly possible to keep low-cost programs low-cost indefinitely. The temptation is too great to pile on requirements.

    Galileo was actually a Pioneer spacecraft - or at least a cousin of Pioneer. So, one might say it was a Pioneer that went wacko when it came to cost and complexity. Of course, as folks ladled on the requirements, that was bound to happen. Galileo was said by some to be the last of the really big missions. Of course, Cassini now holds that dubious honor. It is hard to fault Cassini, though, because it has been such a rewarding mission. Plus I kind of like the thought of something the size of a school bus that we built and launched flying around the Saturn system.

    Anyway - I don't know the best way to answer your question. New tech tends to shape what we try to do. Dawn is a good example - no way it could have been done without long-lived electric propulsion. We could have done multiple flybys of Main Belt asteroids, and such a mission was indeed proposed repeatedly beginning as early as the 1970s. But ion drive meant we could skip over that and orbit two of the big, interesting Main Belt bodies. I hope the proposed Pallas orbiter goes ahead. A big mostly iron asteroid is bound to be interesting.

    New Horizons to me is less capable than it needs to be. It shouldn't take so long to return data. The area explored at high resolution was actually quite small. If something goes wrong with the spacecraft - always a possibility - then the data will be lost. Call me ungrateful, but I wish they'd traded one or two instruments for better transmission capability. The Pluto flyby concept started with a big dish and a small payload - the dish stayed the same or got smaller (depending on which design we're talking about) and the payload grew.

    In an ops vs science situation, ops can be neglected. Kind of like trimming Curiosity's mass by making its wheels less sturdy. It's a rover, dammit - the things that allow it to rove should be the last thing one cuts.


  3. Hi David -
    Good grief, yes! on New Horizons data transmission. I didn't build it, and have no idea the trade-offs that were argued, but what a tremendous leap (Pluto!) to be force through such a small pipe (data transmission rate).

    Time to be alive comment:
    You and I close in age. In half a century the knowledge of astronomy has exploded. It's daunting and thrilling. In the last ten years, data analysis/analytics capabilities have just begun to expand. The probes which will come ought to be massively more capable, with better data transmissions and AI.

    The tick list for the next 50 looks pretty exciting. Good thing the local theocracy doesn't burn scientists at the stake anymore (I guess they just cut the NASA budget), there'd be alot of heretics on fire these days. :)

    1. I sometimes say we should adopt a new theme - "Tau Ceti by 2100!" I think that, unless our civilization comes unglued, we might see star probes by then. We certainly could if we used that goal to push technology. Of course, that's only my optimistic opinion.


  4. "The move would postpone its launch to 20 May 1986; however, because the Centaur could boost Galileo directly to Jupiter, it would reach its goal in 1988, not 1990

    "and had a front-row seat for the Comet Shoemaker-Levy 9 Jupiter impacts in July 1994. "

    So had Challenger not exploded, Galileo would have only been 4 years into it's mission when SL9 crashed into Jupiter, and Galileo would have been directly in the line of fire.

    1. Challenger didn't explode - it broke apart because of aerodynamic stresses. The fireball was, as my post states, caused when propellants that escapes from the External Tank mixed and ignited. I don't mention this in the post, but the Orbiter main engines continued to operate normally even as the Orbiter came apart - they shut down only after they were deprived of propellants.

      Space is big and Galileo could have been put into a distant orbit around Jupiter to avoid the SL9 fragments if anyone thought that was desirable. Rather than worrying about comet fragments hitting the spacecraft, I expect scientists would have been completely thrilled to be even closer than they were when the fragments hit. The likelihood that harm might have befallen Galileo in July 1994 was very small. The likelihood that it Centaur G-prime stage would not have worked as planned was probably greater.


  5. https://www.sciencenews.org/article/support-grows-return-ice-giants-uranus-and-neptune

    This is timely.

    I wonder how many small mini-probes could be loaded to a "mother-ship" as either landers/crashers (think 1960s Ranger probes) or atmospheric penetrators (ala Galileo atmospheric probe into Jupiter)

    I know Huygens was complex, heavy, and delicate, but could you load say, 20 3-pound probes that would be the equivalent of a go-pro with a camera, some sensors and a transmitter to the mother-ship? The mother-ship then functioning as Cassini has for the last several years...

    1. Interesting piece, though I disagree with its characterization of Uranus. The Seventh planet is a fascinating place *because* it's "weird." We don't know enough about other planetary systems to know whether it is non-representative or not. Also, it has mid-sized moons which in the Saturn system have turned out to be quite interesting (Miranda is about the same size as Enceladus). Neptune has Triton, but I think the eighth planet should not jump the queue - it should wait its turn. Uranus is next in line!


  6. https://en.wikipedia.org/wiki/CubeSat


    1. I've come late to the Cubesat party. I thought they were a gimmick for a while, but I've developed a better understanding and appreciation for them.


    2. Same here. The name sounds too cute. Still, if we can have micro drones, the scale should translate. I think in terms of we dropped one 350Kg probe (Huygens) on Titan, how much value would we get if similarly we could drop 100 3Kg probes? Assuming 50% or even 25% capability per mini-probe, the science Return-On-Investment per Kg jumps geometrically.

      Full value of Cassini-Huygens listed at $3.26B and mass at 2500Kg (Wiki) comes to about $1.3M/Kg. Now theoretically, we can get 100 probes, albeit presumably reduced capabilities, for the mass-price of one.

      How much more science value, using the currency metric of study papers submitted, would that have provided to have 40 mini probes to Titan, 5 for Enceladus, Iapetus, etc and a few thrown into Saturn's atmosphere to boot?

      The potential is enticing. Hopefully this kind of thinking makes it into Uranian probe planning and for other missions currently in concept.

  7. Sorry for muddying up your blog with my excessive responses. The topic pretty much captured my attention. (hard to tell, eh?)

    Go to Google Scholar, type : cubesat planetary exploration
    Gobs of papers out there already.

    Two of many...



    Anyway, the gist is that every little fiefdom, university, and interest is now putting a satellite into orbit as a cubesat piggyback on scheduled missions. So Soyuz, DragonV2, and Orion will have to dodge cubesats from Andorra and Liechtenstein soon.

    But further, the planetary implications are under study by many multiples of scientists.

    1. No worries about "muddying up" the blog. You'll note that I included reference to cubesats toward the end of my "On the Moons of Mighty Jupiter" post. Interestingly, Robert Staehle and Stacy Weinstein, who were instrumental in making New Horizons happen, are heavily involved in cubesat engineering. They see cubesats as extensions of the original philosophy behind their proposed small Pluto mission.


  8. Back in the early 1990s there was a PBS NOVA show on Galileo called "The long rocky road to Jupiter." Generally, it's the same as this article. Problems with the shuttles, cost over runs, Centaur issues, old technology (70s and early 80s computer chips; at least two generations out-of-date by the time of launch). Your article left out the little item when the high gain antenna failed and NASA went up to the early 1990s era congress and asked to put a communications satellite to Jupiter. That was a hoot because the engineers had to explain why the high gain antenna failed to work (see, we didn't count on the sun getting the probe so hot flying inside of the orbit of Venus and never thought to look at the design after the new mission profile was done). All in all Galileo was a mild failure. It was a very improved Voyager. Had a Titan Centaur had been available for launch it would have turned out much, much better.

  9. I saw the Nova program, and it didn't include McRonald's analysis of Galileo's likely fate - and that of its nuclear power system - in the event of a catastrophic Shuttle Orbiter loss of control after SRB separation. That's what this post is about. Everything else is context. I mention the high-gain antenna's failure to unfurl, but the relay sat proposal will have to wait for another post - one in which it is more relevant. Galileo was fitted out with sun shields and insulation before it was launched past Venus. The precise cause of the jammed high-gain has never been ascertained, though there are several theories. Galileo was actually a Pioneer that grew. Ask the scientists who use Galileo data whether it was a failure, mild or otherwise; you'll find little sympathy for your assessment.



I like hearing from my readers. No rules except the obvious ones - please keep it civil and on topic.