#tedioushighlights

In general, a space tug is just a spacecraft with no payload of its own that can be used to boost essentially any generic payload such as a satellite, interplanetary probe, or telescope into a different orbit. Essentially we already have and use expendable space tugs in use in the form of upper rocket stages like the Centaur. But a devoted, reusable, in-space refuelable space tug is something we don’t have quite yet.

Proposals for reusable space tugs aren’t anything new, and a flexible, reusable, modular space tug was a large part of the initial concept for the STS project, which was eventually trimmed back to only include the Space Shuttle. For that matter, ULA’s proposed successor to the Centaur upper stage would itself be a reusable space tug.

The STS space tug was meant to be the one-stop solution to all space activity beyond Low Earth Orbit (LEO).

The shuttle would be responsible for essentially all deliveries from Earth to LEO, and from there different configurations of space tugs would lift crew or satellites to geosynchronous orbit, lunar orbit, or even clear out of earth orbit and on their way to other planets.

Within LEO the tug was meant to be a workhorse, assembling space stations and adjusting their orbits, performing repairs and maintenance on satellites by delivering robotic manipulators, and effectively increasing the launch capacity of the shuttle be being able to meet it in an orbit too low to be stable for long and carry its cargo up to where it was needed.

For the biggest of missions, multiple tugs would be stacked on eachother in order to make multi-stage rockets.

So what’s the catch? Because you know there is one.

Well, at heart the catch is that for a space tug using chemical fuel, the extra weight of the hardware for refueling and the extra fuel that needs to be used to return the tug to its fueling point are all but guaranteed to outweigh the mass savings of not having to send up a new set of engines and pumps. Expendable upper stages like the Centaur are already roughly 90% fuel after all. So while you can save the cost of building a new rocket stage each mission, you don’t really save any launches.

That doesn’t really start to change until you look at more advanced spacecraft engines. The solar panels or nuclear reactors necessary to run things like NERVA or an ion thruster weigh enough relative to the propellant they use that it should be possible to get significant reductions in the number of launches needed to do the same missions.  

(from this 2015 NASA slideshow)

(from NASA)

At heart, I think these two images sum up the justification for the SLS. Commercial launchers like ULA and SpaceX have rockets that can handle payloads of around 10 to 20 mt to LEO and are set to start taking over crewed missions to the ISS or really anywhere in LEO in the next few years. Going beyond LEO takes more speed, which takes more fuel, which means a bigger rocket that can lift more mass. A 70 mt rocket like the first version of the SLS should allow long duration crewed missions beyond LEO but still within the Earth-Moon system. A 130 mt rocket like the projected upgraded version of the SLS should allow very long duration crewed missions to Mars orbit, or bigger and better missions on the surface of the Moon or the moons of Mars. 

And for non-crewed missions, a bigger rocket allows quicker trips or larger more capable robots to be sent to the outer solar system. The big sale for robotic missions seems to be that the SLS can do a mission to Europa in roughly 1/3 the time of a mission launched with an existing by taking an inefficient direct flight rather than relying on several gravity assists.

(from the same slideshow)

The other advantage for the SLS is that the plan is to have the option of using larger payload fairings, which will allow larger diameter payloads to be sent to space than current rockets are designed to handle. This is key for a few things like certain telescopes and potentially space station habitats.

So why don’t I like it?

I went into that a bit last week, but at heart my complaint is that building a bigger rocket every time we have a bigger mission gets us stuck in a trap. The only advantage to going to the moon on the SLS instead of on the Atlas V is that with the Atlas V you’d need to use several launches and assemble your spacecraft in LEO while with the SLS you can go in one launch. Less launches sounds nice, but the SLS is supposed to cost close to 10 billion dollars to get its first launch into space while ULA is selling its launches for an average of $225 million. I think if we had more than 40 launches to practice with, we could probably figure out how to lash three small payloads together into one big payload.

Sooner or later we’re going to have to bite the bullet and figure out how to put together big spacecraft, telescopes, and satellites in orbit instead of launching them in one big prefabricated piece. Once we do that, then we can use whatever rockets we have to send missions out however far we want to instead of spending $10 billion for a new moon rocket, another $15 billion for a mars rocket, and $20 billion for a Ceres rocket and only getting a dozen launches out of each new rocket before we retire it. Once we’re already putting together more than one 70 mt spacecraft in orbit every year, then we can start talking about whether or not it’s a good idea to build a rocket to send that same craft up in one push.

(excerpts from document 6, here)

That’s the core of what the publically revealed plan was. It was a sold as a simple, disposable, trial version of a space-station meant to be launched with Gemini hardware.

The formerly secret documents show things a little differently.

(From the NRO. Document 20, here)

Interestingly enough to me at least, there’s still quite a bit of redacted information in the released documents.

(from that same document 20)

But so far, aside from the extra tidbit about still-redacted mission information, I’m just rehashing last week’s post. What I really want to look today at is this:

(From another of the NRO’s goodies: document 35, here)

I realize the text is a bit small, but this is the occupied portions of one of the proposed MOL configurations. Essentially a Gemini capsule would be mounted to a cylinder containing one or, in this case, two pressurized compartments containing the living quarters and control systems for the laboratory. The entire thing was 10 feet in diameter, to match the Titan III interstage, and about 40 feet long. With less than half of that length being living space, the two astronauts would have had to spend a month cooped up in a space the size of a 20’ office container.

The configuration above shows an inflatable external tunnel to allow astronauts to travel from the capsule to the living quarters, but there were a number of options the looked at to solve that problem:

(from the same document)

Personally, I kind of like the hinge option. Not for any good reason, I just like the idea. 

One of the good things about a program that drags on without ever quite getting full funding is that it starts to gather lots of alternative designs and revisions. I think my favorite part of the design proposals was the idea that the MOL could be modular and expandable.

(and again, from the same document)

The program was cancelled of course, but things did get as far as an unmanned test launch in 1966.

(image from NASA, here)

I said last week that the cancellation of the MOL was a turning point for US manned space activity. I think I’ll wrap up by explaining what I meant.

As I see it, the end of the MOL marked a split between the world of manned space activity and unmanned space activity in the US. Up to that point, and through the Gemini program, astronauts and satellites rode essentially the same rockets to get to space. The MOL would have continued that trend, with a series of space stations or a series of missions to the same station using Gemini capsules and Titan rockets. Instead, from the mid 1960s to the retirement of the space shuttle in 2011, the US launched the bulk of its satellites and interplanetary probes and other unmanned missions on relatively simple expendable rockets and its people on much more expensive rockets like the Saturn V and the Space Shuttle. And in my mind the basic, although probably unintended, effect of that split was to make it harder to justify manned missions by making them substantially more expensive than they needed to be.

The other effect of cancelling the MOL was that it essentially put on hold any progress the US might have been able to make on the in-space assembly of space stations or spacecraft and kept us from developing any real ability for in-space maintenance and repair of equipment until we began work on the ISS.

The one major complaint I have about the Slate video is that it presents the MOL as top secret. That’s only half true. The MOL itself was essentially never secret, and was in fact publically announced in 1963 as an orbital laboratory along with the cancellation of DynaSoar. What was an open secret was that the intended purpose of the MOL was to serve the role that would later go to spy satellites, collecting imagery and signals intelligence. The only real secrets were the specific technical details of the equipment.

(DynaSoar, also from our main pdf)

By the time MOL was proposed, unmanned satellites were already being used to photograph targets within the soviet union and to intercept signals. The trouble was that those satellites weren’t terribly reliable, the control systems were almost hopelessly inadequate at reacting to things like cloud cover or emerging situations, and there wasn’t much that could be done to correct any malfunctions.

A manned station, it was thought, would solve those problems while also helping the Air Force develop the experience it felt it needed operating in space in case orbit became a front in the next war.

Astronauts would be able to scan earth using simple spotting optics and only use the ludicrously powerful camera and its incredibly valuable film if they knew they could get a useful image. If something went wrong, or if command wanted to look at something else, the astronauts could react flexibly according to whatever priorities they’d been given. And if the equipment failed, there’d be astronauts with wrenches on hand to put it back in order.

(not the most luxurious of office spaces, from the NRO’s MOL picture page)

It was a good justification, and one way or another the US military was going to get its eyes in the sky, but there were problems and those problems ended up killing the program.

First, always first, was money. MOL had to compete with the Vietnam war for funding. The war won.

Second, while the Air Force wanted the MOL as much to simply give their own people extended experience with manned space flight, the reconnaissance mission was the entire purpose on paper, and the mission shared with the public overlapped with NASA’s scientific mission. On the reconnaissance side, MOL had to compete with rapidly improving technology that eventually erased the major justification for a manned surveillance station. In the world of public opinion, MOL had to compete with the Apollo program.

By the late 60s, the arguments in favor of MOL were almost entirely based on either the capability for continuous surveillance with a manned station in a synchronous orbit or on the power of the camera. Unfortunately for the MOL, its unmanned competition was already projected to offer nearly the same image resolution.

There are two basic reasons that it matters that both suborbital rockets and the first stages of orbital rockets can land softly now.

First is the dream, reusability. A rocket costs about the same to make as a passenger jet. A big part of the reason it costs so much more to send something into space than from New York to L.A. is that we don’t throw airliners away after one flight. Until last month we’ve essentially never gotten the part of a spacecraft that did the lifting back down to earth in one piece.

If we go back to my very first highlight, buried in the documents I read while I was researching the Sea Dragon was the interesting breakdown of the estimated costs of the different parts of the rocket. The first stage was about 29 million dollars, the second stage was about 8 million dollars, and the payload - which is normally the only part we recover- was about 150 thousand dollars. I use the Sea Dragon as an example because I have those documents handy, I doubt those ratios are directly applicable to any real rockets. But I suspect that the first stage of a rocket generally represents anywhere from half to three quarters of the cost of the entire vehicle. It’s the biggest single part of the rocket both in size and in cost and because of the lower velocity it reaches compared to the rest of the rocket it’s the low hanging fruit for recovery after the capsule itself.

So for a 61 million dollar per launch rocket like the Falcon 9, if we assume the rocket itself is half the launch cost, getting back the first stage means not throwing away maybe 15 to 20 million dollars of hardware (probably 15, going by the difference in price between the falcon 9 and the falcon heavy with its two extra boosters).

That’s the dream.

The other reason it matters that SpaceX and Blue Origin got their soft landings is pretty simple. Until now no one’s really been able to inspect the first stage of a rocket (or the only stage of a single stage rocket) after it’s been launched.

When SpaceX had their failure in June, the only way they really had to determine what the problem was was to look at the instrument data they received back from the craft and to check their supplied parts against their models of how the spacecraft would have experienced mechanical stresses in a launch. What they didn’t have, what essentially no one has ever had, was actual observed information on stress and wear from inspections performed on rockets that had actually flown.

Both the SpaceX and the Blue Origin vehicles are essentially gold mines of information for improving the designs of similar vehicles in the future.

And that’s the rest of why it matters. There’s a chance that there could be large cost savings from reusable rockets or rocket stages coming out of this, but even if that falls through there’s now the opportunity to learn information for improving rocket design that could not have been gathered in any other way.

So the difference between the two craft?

Blue Origin’s rocket is a suborbital rocket. It’s a single-stage rocket that hits a speed of about 1 kilometer per second, goes more or less straight up all the way into space, releases a capsule which will in theory be filled with experiments, tourists, or astronauts in training, experiences freefall for several minutes, and then falls more or less straight down to where it started.

SpaceX’s first stage is the main booster for an orbital rocket. It travels mostly perpendicular to the surface of the earth, hits a speed of about 2 kilometers per second, reaches space, releases a smaller rocket that then accelerates up to at least 8 kilometers per second to reach orbit, and then the main booster flips over, thrusts in the opposite direction and travels back to where it started.

(what the leaving and landing of the same rocket actually looked like in a long exposure shot, image from SpaceX)

If the SpaceX deal sounds a lot more complicated, that’s because it is.

But the Blue Origin flight was certainly impressive in its own right.

What these companies learn from their experience could lead to major advances in launch vehicle design and manufacturing in the years to come. And those advances could lead to significant cost savings for everything everything associated with space.

Massively oversimplifying and misrepresenting things, in the late 50s and early 60s reaching the moon was seen as the end goal of manned space flight for the foreseeable future. There were two basic paths that were being looked at. In the path we took with Apollo, the idea was to focus everything on that end goal alone. A really big rocket would need to be built to send a really small spacecraft to the moon and an even smaller spacecraft from the moon back to earth. The other path was slower, and focused on building in-space infrastructure and developing the capacity for in-space vehicle assembly and fueling so that smaller rockets could be used to send up the pieces and the fuel for spacecraft of essentially any size desired which could then be sent on missions not just to the moon, but to anywhere in the solar system.

In that second path the space station that was called for wasn’t a simple tube with life support and a bit of space for scientific experiments, it was a dry dock facility.

You can almost see the frustration of the people who pushed for that second path in the record of the designs that were proposed and then rejected one by one.

Von Braun wanted a large, rotating space station that weighed nearly 100 tons and was almost entirely assembled in space. The in-space assembly was seen as too difficult, and the design started to shift to inflatable or collapsible structures. Even though I know nothing about the individuals or their personalities I can almost imagine profanity-laden discussions within Von Braun’s team about how the difficulty of in-space assembly was half the point of doing it. If they couldn’t assemble the space station, how were they going to use the station as a base to assemble spacecraft?

Langley thought a space station should have its own self-contained purpose. They wanted a station as a place for science.

Ultimately, with Apollo’s time crunch making it impossible to build an orbital assembly facility and still get boots on the lunar ground before Jan 1, 1970, Langley’s rationale for a space station was all that remained. And that inevitably killed the ideal of a large rotating station.

There’s really only unique condition in low earth orbit for scientific purposes. Microgravity. For that, you don’t just not need a rotating station, you don’t want it. And the size of the earlier proposed stations always had just a little to do with the assumptions engineers were making about how a human would hold up to living in a rotating environment.

And so that gets us to Skylab. It wasn’t a base for constructing a moon or a mars mission, it was just a place for astronauts to do science.

And for that, it was successful.

And when you get down to it, considering that indoor air quality concerns and other life support system issues made it nearly uninhabitable by the end of its life, it was probably a necessary first step in long-term manned space habitation.

Certainly the Soviet Union developed quite a bit more essential engineering and scientific knowledge than NASA did by flying their own series of small stations like Skylab. Without the knowledge they developed, the ISS would likely have been entirely unworkable.

But Skylab is in some ways a nice study of one of the failure modes of committee-designed projects. One way to fail is like the Space Shuttle or the Bradley Fighting Vehicle, throw every idea into the design until the end result isn’t actually good at doing anything. The other way to fail is strip the idea down to the one objective everyone can actually agree on, and in the process miss the point of what everyone really wanted all along.

We never really built on Skylab. It was mostly a dead-end for the United States. And I think part of the reason for that is that no one involved really felt invested in the final product. It was just some small disappointing thing. The sad little shadow of grand dreams.

So what are those standards? If you really want to know, you should look here (or really here might be better as standards do update as time passes). If you’re not actually planning to build one in the immediate future, then it’s probably enough to know that the basic CubeSat is a cube with 10 centimeter faces and a total mass of less than 1.33 kilograms. Larger CubeSats can be made by essentially stacking up to three cubes together to make a single bread-loaf sized sattellite.

There’s plenty of other guidance in the standards beyond size, all meant to ensure that the CubeSat is compatible with the standardized launcher that can either be attached to someone else’s rocket as a secondary payload or is already mounted on the ISS if you want to send your cube up on one of the routine ISS resupply missions.

Here’s what an older model of the launcher looked like:

(from that first document)

And a newer version with some sats:

(from the second document)

Yes, that really is basically just a box with a spring inside. I honestly have no idea where people get the idea that stuff in space is complicated.

So I’m calling the CubeSat a powerful tool, which in the context of how I opened this post means that I think the CubeSat opens up many possibilities. But how much can a satellite somewhere between the size of a coffee mug and a bread loaf really do?

Well, for just a few examples it can test a solar sail, or electrodynamic propulsion, or provide communications to earth for a mars lander.

What makes CubeSats so powerful is that with our increasing ability to operate in-space hardware using modern off the shelf electronics we can now fit meaningful scientific payloads into the leftover space in more routine launches of things like ISS resupplies and communications satellite launches. And the standardization of the CubeSat means that so long as the individual satellites fit within the standard, they can avoid much of the expensive testing and design work necessary for a customized design.

So good news all around right?

What I won’t be talking about is anything to do with their Kickstarter campaign, but if you’d like you can browse the comment thread here to read the lamentations of people who are deeply offended that their ‘space selfies’ are late.

So first off, what is Arkyd? Well, in a sense it’s the loaf of bread sized thing in the picture at the top of the post. It’s also the name of what’s supposed to be a series of missions that Planetary Resources is planning to use to identify near earth asteroids for mining.

Like I said last time, we don’t know anywhere near enough about the size and composition of near earth asteroids to actually pull off an asteroid retrieval mission without first doing some prospecting. That’s where Arkyd comes in.

The idea is to build small, light, inexpensive spacecraft to do the surveying. Several of these craft then head out to flyby or intercept near earth asteroids and gather useful information like how big the asteroids are, and broadly speaking what they might be made of.

But it isn’t like NASA or the rest of the world’s space agencies are just being intentionally wasteful in their spacecraft design. If the Dawn mission that recently visited Ceres weighed over a ton and cost 450 million dollars, how light and cheap can Arkyd possibly get and still do useful work?

Probably pretty light, and in the context of space missions probably pretty cheap.

First there’s the obvious savings, Arkyd is going to be designed to survey near earth asteroids instead of several main belt asteroids. Dawn had to be able to pull off 10 km/sec of acceleration even after it detached from its rocket. Arkyd can probably get away with 2.

Next is electronics. Space is hell for electronics. I won’t go into exactly why, but ionizing radiation screws everything up. You can get around the radiation problem to an extent using hardened electronics, but they aren’t off the shelf components, and they aren’t on par with the latest technology. Arkyd is part of a push to essentially use off the shelf components anyway, and deal with the problems caused by radiation by a combination of redundancy and frequent partial or complete resets.

That change in the way the electronics are handled lets every instrument be smaller, lighter, less energy intensive, and more powerful. And that lets next week’s highlight, the CubeSat, do useful science for a cost of $150,000. For something like the prospecting versions of Arkyd, which needs to be able to maneuver and send information back to Earth from well beyond earth orbit the cost floor until launch costs come down is probably in the single digit millions of dollars.