Forty years ago, in April 1981, the Shuttles began their flights into space, and ten years ago they stopped. Usually they write about them: they completed the tasks, they just turned out to be too expensive. We will show the opposite: the original goal of these reusable ships was cheapness, so from the very beginning the project was a disaster. The trouble is this: the general idea behind the Shuttles is similar to that behind the fundamentally new means of space flight - Starship. Will Elon Musk's brainchild repeat the unenviable fate of space shuttles?
On April 12, 1981, the space shuttle Columbia made its first flight - and we would have published this text the day before yesterday, if not for fears that it would be lost against the background of the 60th anniversary of another, larger event. In July 2011, the Atlantis shuttle landed, completing the history of the entire program.
Since then, there has been continuous debate about it. What was it? Success, according to NASA officials? (Until they resign, of course, after which they begin to vigorously criticize). Or, as critics of the program claim, it became a catastrophe that delayed the development of American and world astronautics for decades? Let's try to figure it out.
To Mars, to the Moon, or even lower?
To understand where to go, you first need to understand why. Why did NASA want to create a Shuttle capable of flying only to near-earth orbit, instead of the already tested bundle of the Saturn-5 rocket and the Apollo spacecraft? Moreover, it could fly to the moon?
It all began in 1969, when Richard Nixon instructed the vice president to develop a further strategy for the conduct of the United States in space. The first required $ 57-72 billion a year (hereinafter - in 2021 dollars) and allowed the States to prepare a flight to Mars. The second was supposed to continue to study the moon - and cost no more than 57 billion. The third was limited to stations in low-earth orbit and flights to it - using space shuttles. It was only worth 36 billion.
Anyone can notice that these sums are much more than NASA spends on space annually and today. Naturally, President Nixon did not like this program. He believed that space was too expensive - and wanted to save money on it. Therefore, the authors of the programs were asked to choose the third option and greatly reduce it. NASA believed that presidents come and go, but space remains. If the Shuttles were abandoned in 1969, space flights would remain relatively expensive. And then the new president would not approve of the LEO project, the agency thought.
But if you make reusable shuttles, it will drastically reduce the cost of space flights, that is, the creation of such a station will become cheaper. Sooner or later, NASA reasoned, Nixon's successors would decide to build a station.
It all seemed so logical. Real life has turned plans into dust.
However, we are wrong. The Shuttle program didn't seem logical from the start. In 1970, RAND Corporation ordered an analysis of this entire project (the link leads to an American military site, that is, its holders block Russian citizens, but even a free VPN is enough). The document examined whether it would be advisable to develop a Shuttle to reduce the cost of creating an orbital station. The analysis showed that there was no point.
The point is that the bulk of the cost of a typical space program is absorbed not by mass production, but by the cost of developing the necessary products. Take the Saturns and Apollo. If the entire lunar program cost $ 160 billion (again, in modern dollars), then R&D on it cost 80% of this amount.
Let's go back to the Shuttles. According to NASA, one new launch of each shuttle cost about 0.5 billion, and taking into account R&D - one and a half billion. It is easy to see that the development of the materiel took two thirds of each flight cost. In the RAND analysis, the shuttle's R&D expenditures were estimated according to the promises of its developers, so they were half the real ones. In total, the development of shuttles and related equipment took at least $ 130 billion, while RAND was expecting only $ 57 billion. This is the same as the cost of the entire lunar program - both R&D and the flights to the Moon themselves. All the same, the RAND developers made the right conclusion: the existing disposable media ("Saturns") were cheaper.
How to make black white (you can skip this piece for a former civil servant)
It would seem that after the RAND report, “no one is going anywhere”. What is the point of creating the Shuttle, a reusable carrier that reduces the cost of building space stations, if it is cheaper to replace it with existing disposable carriers, thereby saving on R&D? Moreover, R&D is the most expensive part of any space program.
However, it is not what is considered important, but who and how. Did RAND assess the situation objectively? Nothing, the new head of NASA, James Fletcher, was an inveterate connoisseur of how to promote his point of view in large corporations. He commissioned a summary report on the shuttle perspective from the newly formed company Mathematica. We weren't behind the scenes when he was explaining to her how to write the report, so we won't speculate.
Let us confine ourselves to the facts. Mathematica made a presentation in which it considered the only conceivable possibility of economically viable implementation of the Shuttle project. To do this, the authors of the report took the figures for the most frequent shuttle flights - up to a level of more than a dozen per year. Then, due to the theory of the lower cost of one flight, it was possible to recoup R&D costs. Alas, the Mathematica talk did not explain where the Shuttles would get targets for such frequent flights.
But the politicians liked him: they were promised "cheaper" - and this, alas, was all that seriously interested the officials of the 70s in the space program. As we will show below, in the end, I had to pay more.
The birth of the shuttle: where are we going - to space or to money?
NASA originally planned a full-fledged reusable shuttle ship. Liquid fuel for all engines, making it easy to reuse all shuttle system components.
However, being born under the slogan of "cost reduction", the shuttle from the very beginning took the path of avoiding all difficult issues. The accelerators were replaced with solid-fuel ones (which is bad for the safety of the crew). The external fuel tank with liquid fuel at first wanted to be fished out of the ocean for reusable use, but then they simply forgot about it. After all, this would raise the price of R&D.
NASA developers knew from the very beginning that normal ejection is available only if there are no more than four ejection seats. But making a ship for less than seven people was considered wrong: then it would not have turned out to be as versatile as it was originally seen. Making catapults for seven was difficult - and they just … weren't done.
In fact, the very question of creating an emergency rescue system for crews on the Shuttles was raised. But the bottom line was simple: it would greatly weigh down its design. And raised the cost. It's easy to guess that NASA chose to fly this way - without an emergency rescue system.
In a literal sense: Vostok-1, from which manned astronautics began, had ejection, but no shuttles that flew decades after it. What if an accident happens? So everyone will die, what can you say.And in general, do not escalate, what accidents, everything is reliable with us!
In the 1980s, while investigating the first Shuttle disaster, physicist Richard Feynman wrote:
“What is the reason for the fantastic belief of management in the reliability of the Shuttles? The first reason is to try to convince the government of NASA's excellence in order to successfully receive government funds. The second reason is a sincere belief in the veracity of [low accident rate expectations], showing an almost unbelievable lack of communication between management and engineers [NASA]."
In practice, the Shuttles have had quite a few weaknesses. Their aluminum body was covered with 24,000 ceramic tiles - very difficult to install and often falling off. In addition, some of the surfaces were designed exclusively for aerodynamic loads. As a result, their strength turned out to be insufficient.
In light of all this, it is easy to understand why Feynman states: the engineers estimated the probability of the Shuttle disaster at 1% (the actual one was ~ 1.5%). That is, every hundredth flight, from the point of view of engineers, was programmed to be catastrophic. Considering that the resource of each shuttle was planned to be exactly 100 flights, according to its engineers, each of them had a more than decent probability of dying.
For a system initially targeting dozens of flights a year, a 1% disaster probability is a level of safety that should rather be called the “level of guaranteed danger”.
But the administrators of the project, notes all the same Feynman, believed that the probability of a catastrophe is only 0.001%.
How could they be so unaware of what the engineers think? We have already mentioned such a scheme more than once in various areas of human activity. When we divided all spheres of life into highly specialized sectors, we immediately laid the possibility of a typical scenario: no one knows what is happening nearby.
A specialist is so specialized that he can no longer explain his knowledge to a layman. The administrator tries to hear from a specialist, but he himself has such a poor understanding of engineering issues that he is not able to clearly understand what the design engineer thinks about the issue.
And it should be borne in mind: Feynman's report after the first disaster was never able to reach the brain of NASA administrators. They still did not realize how high the accident rate of the system created in their organization was.
"The wing is strong and our Shuttles are fast": how the inability to grasp the inadequacy of the Shuttle design led to their second accident
It is highly probable that this is why the crew of the space shuttle Columbia died in February 2003. As you know, at launch, a piece of polyurethane foam insulation from the external tank of the Shuttle broke away from the incoming air and hit the carbon fiber edge of the Columbia's wing. The edge was very thin and had a hole in it. After a couple of weeks in orbit, the Shuttle returned to Earth, but upon entering the atmosphere, hot gases entered the wing and destroyed it from the inside. The shuttle lost stability and began to fall. There was no escape system, so all the astronauts were killed.
But this was not at all a necessary development of events. The fact is that the next launch of the Atlantis shuttle was scheduled for March 1 of the same year. In theory, with fewer crew members, he could take seven astronauts from Columbia.
The trouble is, for that, NASA administrators had to want it. "Columbia" was in space in mid-January 2003, the reserves on board made it possible to survive in space for a month and a half, but subject to austerity from the very beginning of the mission.
At first glance, nothing stood in the way. It was known about the damage to the wing by a piece of thermal insulation foam from the moment of launch - it remained unclear only where it was located. Moreover, a number of NASA engineers asked the US Department of Defense to take a photo with a spy satellite in order to understand the extent of the damage. Others suggested that one of the Shuttle crew members go outside and inspect the damage. Here, however, there were few chances: there were no spacesuits for extravehicular activity on the Shuttle that time.After any of these actions, it was possible to understand exactly where the hole is in the wing. From this the need for a rescue mission naturally followed.
The Columbia Disaster Inquiry Commission concluded:
"Such a rescue operation was a challenge, but it was quite surmountable."
None of this was done. Cause? First, before the flight, despite Feynman's report after the first Shuttle disaster, no one seriously considered the possibility of a new disaster. Therefore, no completed rescue operations were planned. NASA administrators rejected requests from engineers to the Ministry of Defense.
Second, when computer simulations (using the Crater program) showed that there might be a hole in the wing, the simulators simply said it was exaggerating the threat. They were sure: well, a piece of polyurethane foam cannot kill the Shuttle, well, the program is wrong.
We really don't know if everyone at NASA thought so. For example, the person who was then in charge of flight control at the agency explained to his subordinate his approach to damaging Columbia:
"You know, even if they knew about it, it would be better not to know … Wouldn't it be better for the crew to have a happy flight and die unexpectedly at its end, than to stay in orbit, knowing that nothing can be done?"
The commission's investigation established what could have been done. But for this one had to want it: think over plans, work. Nobody did this in advance - and in January 2003 itself, none of the NASA administrators could react in time and understand that people can be saved by Atlantis.
Operating price: we didn’t manage to fly to the money either
Contrary to initial beliefs, the complexity - and cost - of preparing the Shuttles for the next flight turned out to be much higher than planned. Problems poured literally everywhere: tiles fell off more often than they thought, checking and replacing them turned out to be extremely time-consuming.
The idea of a dozen or more flights a year, as expected, did not come true. Therefore, according to the plans, the Shuttles were supposed to carry cargo into space for 560 dollars per kilogram, and in real life they carried them for 18 thousand (excluding R&D), or even 60 thousand (including R&D). The error in this key parameter was about 30-100 times.
For 30 years of flights "Shuttles" flew 135 times, 4, 75 times a year. The average cost of a flight, taking into account R&D, turned out to be one and a half billion dollars, without them - 0.5 billion. The average cost of the program per year has exceeded seven billion in modern money. Total - more than 210 billion, one and a half lunar programs.
In the 21st century, former NASA chief Michael Griffin tried to pose a completely reasonable question: what if everything went according to the RAND scenario, without the development of Shuttles, but only with the use of Saturns and Apollo?
His conclusions: due to the complete completion of R&D on these rocket and spacecraft, NASA would have borne only operating costs for them. The cost of a single flight to the Moon, excluding R&D - which has already been completed and therefore does not require new costs - was in the region of a couple of billion dollars. Flights of "Saturns" to a lower orbit, to orbital stations such as "Skylab" - are much cheaper. The typical budget for a NASA manned program is no less than ten billion modern dollars a year. This is the cost of two flights to the moon and four flights to a near-earth orbital station. Thus, 30 years of space shuttle flights cost the United States at least 60 flights to the moon.
By the way, the Saturn rocket was more suitable for flights to orbital stations than Shuttles. She had a version without the first stage, Saturn-IB. He put 21 tons into low-earth orbit, close to the Shuttles - only his one launch cost only $ 0.35 billion, much cheaper than $ 0.5 billion for the Shuttles (in both cases, prices are given without R&D).
By launching Saturn-5 into space instead, it was possible to put into orbit not only a payload, but also one empty stage with each launch - from one such NASA made the Skylab station in the 1970s.
The step was so large that the station left it larger than Mir and like half of the ISS. Of course, empty steps could be docked. That is, if NASA wished, it could create an orbital station every year - no less than the modern ISS.
Let's summarize. The shuttles were created with one main goal: to fly into space is cheaper than before - due to its reusability. Numerous attempts to justify their existence by “the possibility of satellites being taken from orbit” and other functions do not find any confirmation in real sources from the time of the development of shuttles. Neither the RAND report, nor the Mathematica report, nor in NASA documents from the era of the creation of the Shuttles, does their ability to remove satellites from orbit as one of the main functions. This is understandable: at the time of creation, there was no significant market for satellite repair, so no one would base the development of a huge project on purely hypothetical ideas about such a market.
Stories like “but the Shuttles could have removed the satellite from orbit” chronologically appear very late (mostly already in the 21st century) and almost always in the context of justifying the numerous problems of the program. During the period when the Shuttles were selected as the future space system, everyone was not talking about this, but only about the future low cost of their space flights.
In real life, the Shuttles turned out to be poorly designed: the external fuel tank was not reusable, maintenance of ceramic tiles and overly complex engines required a lot of money and time.
Very little attention was paid to the security of the system. No one even calculated the possibility of events like the death of the Columbia crew. The engineers guessed that every hundredth Shuttle flight would be a disaster (in practice, one in 68), but the administrators and astronauts were not aware of this.
Simply abandoning the Shuttles and flying on the Saturns, the Americans would have time to study the entire surface of the Moon, including the circumpolar regions with their water ice, and, if desired, even build a permanent base on Selene. Along the way, the States could build a mega-ISS once a year, if they had such a desire.
We have no choice but to agree with the conclusions of Michael Griffin, who headed NASA in 2005-2009: "Instead of creating Shuttles, it was worth gradually modernizing the Saturns and Apollo".
A warning for Starship?
All disposable space systems are equally happy, but all reusable space systems are unhappy in their own way. Despite this general rule, Shuttles and Starship have a couple of things in common.
First, neither one nor the other has an emergency rescue system. The Shuttles justified this simply: with an emergency rescue system, like the Apollo or Soyuz, the payload weight would be reduced to indecency. Or it had to be done much more - and more expensively.
Starship is in a tougher situation. Its capacity - with an internal hermetic volume of more than 800 cubic meters - is hundreds of people. These are not the seven brave ones on the Shuttle: the emergency rescue system for them will be cyclopean. If created, it would render the spaceship meaningless and unsuitable for a flight to Mars. And Musk conceived Starship precisely for this purpose: the Moon, we recall, is not of interest to him, because it cannot be terraformed with the available technical means.
The second similarity between Starship and Shuttle is thermal insulation tiles. The steel giant's need for them is much less than that of the Shuttle. SpaceX's materials are more heat-resistant (no CFRP or anything else). This is an important point: the stainless steel of the new ship does not lose strength until at least 800 OC, and CFRP starts to do this already at 150 OC. But still, there are tiles on a small fraction of the ship's surface. Even if they are made of glass-like material, not ceramic - Elon Musk has yet to show that he can cope with the problem of tearing these tiles off during flight. If the tiles are dropped regularly, it will become quite difficult to fly into space at the required frequency of the Starship.
Of course, this does not mean at all that the new American reusable space system will have the same catastrophic fate as the old one. Musk is right when he says that there is no emergency rescue system on airliners, nevertheless, they are much safer than a car. He plans to bring the security of Starship exactly to the level of airliners, and if this works out, he does not need a rescue system.
The situation with tiles and overheating when entering the atmosphere should also be simpler: fortunately, the overall resistance of stainless steel to thermal loads is higher than that of the Shuttle materials. This, by the way, is a big plus for the entire Starship concept, and one cannot but rejoice that Musk, despite the resistance of the engineers, insisted on this.
The main reason for optimism about Starship, however, is not technical, but organizational. The NASA administrators who promoted the Shuttle did not design the rockets themselves. They were "specialist management specialists." Subordinate engineers proposed options, and administrators selected and adjusted them. So they adjusted the original "Shuttles" to "semi-reusable", having received a "cheap reusable ship" with a cost of operation much higher than "expensive one-time missiles".
In contrast, Musk has grown out of himself as a good space technology designer over the past 18 years. It is he who is behind the introduction of supercooled fuel on its carriers, the choice of landing on the tail for the Falcon and the main ideas for the Starship. This means that the key decisions for a new space reusable system are made by an administrator who is at the same time an engineer (albeit without a corresponding diploma). Hopefully it does better than its NASA predecessors.