On 8 December, a small California-based company called SpaceX did something that had previously been the domain of governments: it launched a space capsule with a pressurised cabin into orbit, then brought it back down again intact. The “Dragon” capsule could have carried a crew. Someday soon it will.

Commercial space flight has been a charged issue in the US ever since President Barack Obama announced in February that he wanted to outsource trips to the International Space Station to private companies. Critics of the plan, including the first moonwalker, Neil Armstrong, claim that only NASA can be trusted to build spacecraft that carry astronauts, as only they know how to do it properly. The success of SpaceX’s Dragon capsule and the Falcon 9 rocket that carried it suggests otherwise.

The feat is all the more impressive when you consider the launch history of NASA’s first capsule, the , which carried the first American into orbit in 1962.

NASA first tried to launch a fully equipped Mercury capsule in July 1960, in part to test its heat shield. Unlike the Dragon test flight, however, it was a complete failure. A structural problem in the attachment of the capsule to its Atlas rocket triggered an explosion that destroyed both capsule and rocket.

Four months later, the next attempt at sending Mercury into space also . An error in the wiring of the Redstone rocket it was sitting on caused the rocket to lift only a few centimetres from the launch pad, before settling back down and shutting down. Despite this, the Mercury capsule dutifully followed its flight plan, jettisoning its escape tower and deploying its parachutes. These dangled from the fully fuelled rocket overnight, until the batteries on the rocket’s self-destruct mechanism ran down. Workers then gingerly moved in to disarm the rocket and drain its fuel.

Finally, on 19 December 1960 – on the third attempt – a Mercury craft successfully flew into space.

Compared to this, the success of the Dragon capsule’s first flight is nothing to sniff at. Maybe somebody other than NASA understands this stuff after all.

The flight is exciting because of its implicit promise that, someday, it won’t just be test pilots and rich space tourists that get to blast into orbit. Maybe the rest of us will be able to go too. In the 50 years since NASA’s first successful capsule test,

it has become clear that government space agencies have no interest in making that happen. But now the game is changing – and about time too.

The recent authorisation bill does direct NASA to proceed (subject to funding) with development of a heavy-lift launcher to support future space exploration. However, nowhere does it require use of the shuttle SRBs. The farthest it goes toward that is to specify that NASA should use existing technology from the shuttle and the now-defunct Ares launchers “to the extent practicable”.

Whether it is necessary, “practicable”, or even desirable to use SRBs in a new man-rated heavy-lift launcher depends very much on who you ask. It seems clear who the – ATK, the Utah-based company that makes the shuttle SRBs and was expecting to make derivatives of them for Ares. Observers with less of a financial stake in the decision might disagree.

Saturn V precedent

Claiming that solid rockets are necessary for a heavy-lift launcher is obvious nonsense. The US’s previous heavy-lift launcher, the , used no solid rockets and lifted a bigger load than the new launcher is required to carry.

What’s “practicable” depends very much on who analyses the problem. Usually the devil is not in the details, but in the assumptions.

One key assumption involves the engine used to power the launcher’s first stage, which is used to push the rocket off the ground. The launcher will probably weigh 2000 tonnes or more at takeoff (the Saturn V weighed about 3000). That requires a lot of thrust, which can be attained with either a few big engines, or a large cluster of smaller ones.

The US has no really big liquid-fuel rocket engine in production today, and historically has been averse to large engine clusters. That leads people to think that SRBs are about the only answer.

Clusters of engines

But there are alternatives. The simplest would be to put the Saturn V’s big first-stage engine, the F-1, back into production. It’s an old design, and not as fuel-efficient as more modern ones, but it worked well and had plenty of thrust. (Or, of course, you could buy big Russian engines. But Congress really doesn’t like that idea.)

The other option is to bite the bullet and use big clusters of engines. The RS-68, currently in production for , could probably do it with seven or eight first-stage engines. (The had eight engines in its first stage, and SpaceX’s has nine.)

This does seem like more work than just using the SRBs. Is there a good reason for it? Yes. Large solid rockets have never been a very good way to build launchers that might have crews on top, especially because of the problems in getting the crew away from a failing launcher.

Parachute danger

For one thing, shutting solid rockets down in an emergency is difficult. Contrary to popular belief, it’s not impossible, but it’s a somewhat violent process and that creates its own problems. (Solid boosters were chosen for the shuttle on the assumption that they could be shut down on command, but the decision wasn’t reconsidered when that proved to create impossible structural loads.)

For another, while solid rockets can be fairly reliable if you spend enough money on them, when they do fail, their failures are often catastrophic. (Ailing liquid-fuel engines, by contrast, usually just shut down when they fail – this happened several times on the Saturns.)

Moreover, a disintegrating solid booster creates its own special hazard to an escaping crew capsule: a huge shower of chunks of flaming fuel, the heat from which can .

Finally, all of this is on top of the problems that solid fuels create even when everything works perfectly: vibration, pre-launch handling hazards, and atmospheric pollution.

There is plenty of reason to avoid using solid boosters in a new heavy-lift launcher, and there are feasible ways to do it. Steering clear of them would present a political problem: the loss of jobs in Utah. It’s worth it.

Henry Spencer is a computer programmer, spacecraft engineer and amateur space historian

Today is the 10th anniversary of the start of permanent occupation of the International Space Station (ISS). For those whose memories have grown dim, here are some highlights from the station’s first 10 years.

29 July 1970

NASA abandons all hope of further production of its mammoth Saturn V rockets, which launched Apollo astronauts to the moon. As a result, it cancels the big space station it had planned to build as a successor to . With Skylab itself nearly finished, space station work enters a long hiatus.

25 Jan 1984

In his state of the union address to Congress, President Ronald Reagan announces: “Tonight, I am directing NASA to develop a permanently manned space station and to do it within a decade.” With contributions from Europe, Japan, and Canada, this becomes Space Station Freedom. Unfortunately, it is repeatedly redesigned, getting further and further behind schedule and over budget. By 1993 it is clearly headed for cancellation.

Sept 1993

With the Cold War over, a US-Russian agreement combines Space Station Freedom and Russia’s planned Mir 2 station to produce ISS.

20 Nov 1998

Zarya, the first ISS module, is launched from Baikonur, Kazakhstan.

6 Dec 1998

Unity, the first US module, is docked to Zarya by space shuttle Endeavour. This flight also marks the first crew visit to the station and includes the first station spacewalks. An 18-month hiatus follows, due to funding problems with Russian modules and technical problems with US modules.

21 May 2000

Space shuttle Atlantis visits, preparing the station for resumption of assembly.

26 July 2000

Zvezda – a.k.a. the Service Module, originally built as the core module of Mir 2 – joins the station.

2 Nov 2000

The “” crew arrives by Soyuz. Permanent occupation of the station begins.

10 Feb 2001

The Destiny lab module arrives. ISS internal volume now exceeds that of Mir.

23 March 2001

The Russians reluctantly de-orbit Mir, removing ISS’s only competitor.

23 April 2001

Canadarm 2, the station’s manipulator arm, arrives on the shuttle Endeavour. This is the first non-US/Russian station component, added relatively early because it’s needed for later assembly work.

30 April 2001

Dennis Tito becomes the first space tourist to visit ISS, over NASA’s strenuous protests.

15 July 2001

The module, a pressurised module that serves as an entryway and exit for spacewalks, arrives. ISS is now heavier than Mir was.

21 July 2001

First spacewalk done from the station itself, using the Quest airlock.

4 Feb 2002

The first significant station emergency occurs when control over the station’s orientation, or attitude, is lost for several hours, endangering both power generation and temperature control. Rapid response from the crew and ground controllers resolves the problem before any damage is done.

1 Feb 2003

Space shuttle Columbia is lost during atmospheric re-entry from a non-ISS flight. The shuttle fleet is grounded, making the station entirely dependent on Russian Soyuz and Progress spaceships for crew and cargo deliveries. The resident crew is cut to two, with activities largely limited to maintenance.

6 April 2003

An old rocket component makes a close approach, causing the station crew to briefly retreat to their Soyuz lifeboat in case of collision.

28 July 2005

Discovery makes the first shuttle visit since the loss of Columbia. (The next doesn’t happen until a year later, due to continuing problems with foam shedding from the shuttle’s external fuel tank during launch.)

26 Nov 2006

A rendezvous antenna aboard a Progress M-58 cargo ship becomes tangled with the end of the station’s Zvezda module during docking. It is eventually freed by spacewalkers with cutting tools on 22 February.

11 Feb 2008

The Columbus module – the main European ISS component – arrives. Designed as a lab for biology, physiology, fluid physics and other experiments, it is roomy enough for three crew members to work in at once.

3 April 2008

The European Space Agency’s “Jules Verne” ATV supply ship arrives. It is the first non-US, non-Russian vehicle to visit.

4 June 2008

The first part of Japan’s bus-sized Kibo lab arrives on shuttle Discovery.

29 May 2009

At long last, the permanent crew is enlarged to the intended six (requring two Soyuz lifeboats docked to the station at all times), and substantial science work can start.

17 July 2009

With the shuttle Endeavour visiting, there are 13 people aboard, a new record for ISS and for any single spacecraft (and a tie for the all-time record for the number of people in space simultaneously).

17 Sept 2009

HTV-1, the first Japanese supply ship, arrives.

10 Feb 2010

Five vehicles are docked, a “full house”: two Soyuz capsules, two Progress freighters, and the shuttle Endeavour.

31 July 2010

A pump failure in the cooling system for the power equipment temporarily limits available power and badly curtails operations. Fixed, after several spacewalks and some unexpected difficulties, on 17 Aug.

When the Obama Administration announced a new plan for NASA last February, it set the stage for a major battle with congress. The first phase of that battle concluded yesterday, and the result is, well, okay. It’s not perfect, but it could have been much worse.

On Wednesday, the House of Representatives endorsed the , temporarily abandoning their . Most everyone interested in promoting cheaper and better spaceflight heaved a sigh of relief.

Some parts of the administration’s policy ended up being uncontroversial. The real uproar was about getting NASA out of the space transportation business, and helping commercial suppliers to get into it.

Predictably, the loudest objections to this came from congressional districts that get a lot of NASA space-transportation money. Their congressmen in particular have been denouncing the new policy at the top of their lungs.

Heavy-lift launcher

Clearly, the administration wasn’t going to get everything it wanted, especially in an election year. Some concessions would have to be made to reduce the loss of jobs in NASA’s existing empire.

So the senate NASA authorisation bill diverts about half the money that was supposed to go to supporting commercial spaceflight and long-term technology R&D. Instead, it puts it towards starting work on a government heavy-lift launcher, preferably derived from shuttle designs, and recommends that various other in-house NASA efforts continue.

This isn’t great, but it’s livable. Even the remaining half of the original request would be a big step forward, and NASA was given considerable flexibility about details.

More for less

The House version, by contrast, looked like a disaster. Essentially all commercial support and R&D money would go instead to continuing large parts of the Constellation programme to return to the moon, which would be told to do the same jobs on the same schedule for less money, somehow. (As the Augustine committee found last year, coming anywhere close to meeting those goals was going to need more money, not less!)

Would-be commercial launch providers would get no support, while government rockets would act as subsidised competition. Commercial rockets would also get a bunch of new requirements, which would not be imposed on their government competition. It’s hard to imagine a more effective way of discouraging investment in commercial spaceflight.

Purse strings

This was a recipe for making US crewed spaceflight depend on high-priced rides on Russian Soyuz spaceships into the indefinite future, while wasting the money that could wean that dependence.

Fortunately, the House version stalled when it became clear that there was considerable opposition to it, and an attempt at a revised version didn’t fare any better. Faced with the prospect of leaving NASA adrift for months without an agreed-upon direction – and potential voters angry – as everything stops for the mid-term election, the House caved in and accepted the Senate version.

This isn’t the final word. The original backers of the House version haven’t given up, and even the Senate bill isn’t perfect. The authorisation bill’s official plan now needs to be turned into specific funding allocations in an appropriation bill, and that too needs both House and Senate agreement. But an important first step has been taken, and in more or less the right direction.

Henry Spencer is a computer programmer, spacecraft engineer and amateur space historian

See also: Space 2020: what NASA will do next

NASA’s Constellation programme, which was going to fly manned capsules to the International Space Station in (maybe) 2015, to the moon in (maybe) 2020, and to Mars someday, is dead. Some people are mourning it. I’m not.

Is manned space exploration important? Yes – not least because it simply works much better than sending robots. When you look past the rhetoric and superstitions and compare the results, today’s robots (and tomorrow’s too) are pitifully limited, painfully slow, and not really all that cheap. (Case in point – NASA recently gave up trying to free the Mars rover Spirit from a sand pit it had been stuck in for nine months. But when the crew’s lunar rover got bogged down in loose soil, the astronauts got off, picked it up, moved it, got back on, and drove away – all in maybe two minutes. Robots do fine when everything goes as planned, but that’s rarely true on complex, poorly-known planetary surfaces.)

Exploring with robots looks cheaper only because we set our expectations so much lower. Bolder goals need humans on the scene. Nevertheless, I’m not shedding tears for Constellation. Why not? Because it wasn’t going to get us there.

First, it probably wasn’t going to work. Even so early in its life, the programme was already deep into a death spiral of “solving” every problem by reducing expectations of what the system would do. Actually reaching the moon would probably have required a major redesign, which wasn’t going to be funded.

Second, even if all went as planned, there was a money problem. As the noted, Constellation was already underfunded, and couldn’t ever get beyond Earth orbit without a big budget increase. Which didn’t seem too likely.

Finally, and most important, even if Constellation was funded and worked … so what? The programme was far too tightly focused on repeating Apollo, which was pointless: we already did Apollo! Early ideas of quickly establishing a permanent lunar base had already been forgotten. Constellation was going to deliver exactly what Apollo did: expensive, brief, infrequent visits to the moon. That was all it was good for.

Sure, there were hopes that Constellation’s systems could later be adapted to support more ambitious goals. But Apollo had those hopes too. It didn’t work in 1970, and it wasn’t going to work in 2020.

The demise of Constellation is not the death of a dream. It’s just the end of an illusion.

The proposed replacements are a mixed bag, with reason for optimism but some reason for concern.

More R&D is good. Space flight technology has stagnated badly in recent decades; for example, we badly need a robust, fully-reusable, low-maintenance heat shield system for atmospheric re-entry, and promising concepts from 50 years ago are still awaiting flight tests. My one concern is that when money gets tight, it’s easy to cut R&D funding that isn’t tied to a specific project – look at what’s happened to NASA’s aviation research.

Switching to commercial space transportation, for both cargo and crews, is long overdue. For over 20 years, NASA has been legally required to use commercial space transportation whenever possible – and has used every possible excuse not to. It’s high time to end this.

Opponents’ main argument is that NASA will have trouble assuring astronaut safety if it uses commercial launch services. Hogwash. To be blunt, NASA has no financial incentive to build safe spaceships – the shuttle, on average, has killed its entire crew about every 50 to 60 flights, and yet it has kept going. (Indeed, after the Challenger disaster, NASA’s shuttle budget increased; compare that to what happened after Boeing’s first two launches of its rocket failed – no one bought more launches and the rocket programme folded.) Spaceship builders who have a direct financial interest in safety should do better, not worse.

At the very least, safety assessments should be done by an independent authority, with no vested interest in the answer. When NASA was launching its Orion capsule on Atlas or Delta rockets five years ago, its safety standards () were very strict indeed – and the commercial rockets ended up losing out. But when NASA started applying those standards to its own rocket when development of the Ares rocket and Orion crew capsule got well underway a few years later, suddenly the standards were in need of revision, and the revised ones () were much less demanding. It’s time to give commercial space flight a fair trial.

The one aspect of the announcement that is worrisome is the vagueness of the long-term plans. “We’ll do neat stuff someday” is a recipe for going back into the holding pattern NASA has been stuck in for many years, driving endlessly around the parking lot without ever getting out onto the road. Even if the transportation is going to come from commercial providers and the details will depend on them, they need some rough idea of what NASA wants to do and when. A clearer explanation of NASA’s new exploration goals is urgently needed.

Henry Spencer is a computer programmer, spacecraft engineer and amateur space historian