Chair Force Engineer

Tuesday, June 24, 2008

DIRECT delivery

One of the most common complaints against DIRECT during the past week has been that the rocket (Jupiter 120) carries too much payload to the International Space Station. In short, the rocket is TOO CAPABLE for the mission, in the reasoning of its detractors. I've looked at this argument and find it to be exceedingly simplistic in the grand calculus of Project Constellation.

Before claiming that Jupiter 120 is too capable for cargo delivery to ISS, let's examine what is really required for ISS resupply. Every year, the station is visited by two Progress freighters, each carrying 2230 kg of cargo (1800 kg dry, 430 kg water.) The station also receives approximately three shuttle flights per year, each carrying 9000 kg of cargo in the 4000 kg Multi-Purpose Logistics Module. All of this is required to support three astronauts per year. After the crew is doubled to six, you can expect to double the resupply requirements. The European ATV will pick up some of the slack, but it can only deliver ~7700 kg at a time.

There's also the down-mass problem, which neither Orion nor COTS have fully addressed. There's a funny anecdote in the astronaut community which claims the purpose of sending the Shuttle to Mir was to get rid of all the trash the Russians had been piling up. The problem with all of the resupply methods currently planned for ISS is that they bring very little, if any, mass back to earth. Much of the downmass requirement for the space station can be addressed by cramming the junk into an expendable payload carrier which will burn up in the atmosphere. Critical items could be stashed in an Orion for return to earth.

While I'm no expert on the ISS consumables situation, it would appear that the more consumables mass you can send to the station, the better. If nothing else, it will make America less dependent on Russian Progress and Soyuz spacecraft.

Jupiter 120 can loft roughly 45 metric tons to the space station, and Orion will only account for half of that capability. So what do we do with the other 22.5 tonnes? A simple solution would be an expendable MPLM with docking adaptors on both ends. It would be stashed in the spacecraft adapter until the Jupiter core reaches orbit. Orion would then separate, dock with the MPLM, and rendezvous with ISS. Orion would then dock the MPLM to the ISS. During Orion re-entry, the MPLM would be discarded before Orion hits the appreciable atmosphere. Of the 45 tonne capacity of Jupiter 120, 22.5 tonnes would be devoted to Orion and 13 tonnes would be occupied by the expendable MPLM. It sounds like a pretty efficient use of Jupiter's capability to me.

In a worst-case scenario, let's assume there's not enough budget or schedule to develop the expendable MPLM. If you look at hardware costs alone, a Jupiter 120 will cost more than an Ares I for an ISS mission. But hardware costs play a very small role in the overall cost equation. The standing army costs for Ares I/V will be bigger than those of Jupiter 120/232. Standing-army costs dwarf the costs of hardware. Development costs are also a major factor, and this is another area where Ares is less efficient than Jupiter.

The last thing worth considering is the short service life of NASA's ISS-resupply system. There will be two ISS missions per year from 2015 thru 2017. Then the US will end participation in ISS. That's a grand total of six flights. I'm incredulous when NASA tells me that the hardware costs associated with six Jupiter flights will be worse than the development and standing-army costs of Ares I.

In looking at the ISS resupply question, we must first ask whether the anointed Ares I meets the real requirements. Even if that condition is met, there is a tradeoff of whether to keep the hardware costs low in exchange for high development costs, a protracted development schedule, and two standing armies for two very different vehicles. In my mind, it just doesn't compute.

EDIT: I overstated the ISS reliance on the shuttle for resupply. Since 2001, the shuttle has delivered an average of one MPLM to ISS per year. Nevertheless, the current schedule of sending an ATV every 17 months does not equal flying an MPLM once per year.

Sunday, June 22, 2008

DIRECTly Seeing the Light

It appears that Mark Whittington is warming to the DIRECT approach for shuttle-derived heavy lift. I take this as a sign that the current ESAS plan, with all of the major revisions that have been made between Fall 2005 and now, is losing support amongst space enthusiasts and amongst technically-inclined observers outside the halls of NASA.

I've always been conflicted between DIRECT and an EELV-based approach to space exploration, from a technical standpoint. I like the free-market approach taken with launching crew on a wide-bodied Atlas, and launching cargo on a cluster of wide-bodied Atlas cores. But DIRECT lives up to its name in terms of being quick to develop and test, and it's markedly efficient at the politically-driven goal of preserving the (inefficient) shuttle infrastructure and jobs.

In a recent post, I discussed the weight issues associated with Ares V (probably to be renamed Ares VI if the extra RS-68 engine is slipped in.) The rocket is growing to address performance shortfalls, but it has become too heavy for the existing crawlers, too heavy for the existing launch pad, and too heavy for the hard stand on which the mobile launcher sits. I suggested that NASA should have initially determined weight and size limits on their rocket, based on the existing infrastructure, and limited the weight and size of Ares V to fit within those requirements. If that rocket were insufficient to meet the lift requirements for Project Constellation, use two heavy-lifters instead of one heavy-lifter and one crew launcher.

In that case, the resulting heavy-lift rocket would probably look a lot like the Jupiter-232. But in the current political climate, it will probably not happen for a variety of reasons. For one, Mike Griffin's NASA didn't invent it. Will NASA be able to swallow its pride and accept an outsider proposal? Probably not, at least not under the current leadership. Secondly, DIRECT is a modern update of the Martin "New Launch System" proposal, done by a small group of industry outsiders and assisted by NASA employees working off-the-clock. If there is to be any honest, technical discussion about the merits of DIRECT versus ESAS, the NASA "traitors" who assisted the DIRECT team will need immunity. Finally, DIRECT doesn't rely on five-segment SRB's, which will deny billions of R&D dollars to ATK.

At the same time, it's clear that "A Change is Gonna Come," to quote Sam Cooke. The current architecture is not viable politically, fiscally, or technically. Mike Griffin, by his own admission, is done when the Bush Administration leaves office. The next president, regardless of party, will be under tremendous political pressure to save shuttle jobs. It would not be a stretch for the incoming NASA administrator to order a 60- or 90-day study of existing launch plans. At the end of the study, the new administrator would announce that NASA has "refined" its launch vehicle concept, whether it be Ares II/III (renamed Jupiter 120/232) or Shuttle-C for cargo plus EELV's for crew launch.

DIRECT is admittedly not the perfect solution to the problems of shuttle-derived heavy lift. Modifications will still be required at the launchpad and on the shuttle's fixed service structure, although none of them will be as drastic as what's currently in store for Ares V. The upper stage of Jupiter-232 is a question mark. Conceptually, it's a bigger version of the Centaur from Atlas V, with two J-2X engines in place of the RL-10A-4's. It remains to be seen how well the Centaur "balloon tank" concept will scale up (although the Jupiter upper stage will not be a purely pressure-stabilized design.) The problem once on-orbit is that the two J-2X's of the Jupiter upper stage have too much thrust to push the Orion-Altair stack without breaking the docking mechanism. There are many ways around this problem, such as throttling down the engines, or having Orion and Altair delay their docking until both have arrived in lunar orbit.

It's impossible to say what's going to become of ESAS, between the reported technical problems and the ever-shifting whims of Washington. But it's safe to say that if NASA can swallow its pride, the DIRECT guys have offered them an easy way out.

Labels: , , , ,

Friday, June 20, 2008


The people who knew me in high school and college would probably not recognize me today. This statement is most true when it comes to space.

There was a time when I was enamored with all things space. It was a part of the action cartoons I grew up with as a child. It gave me something to be awed by, and something to strive for, as I progressed through my education. I became a hardcore devotee of Robert Zubrin and his Mars obsession. I wholeheartedly embraced the Von Braun-ian vision of bold, nationalistic space programs that brought glory to the state.

Then a funny thing happened along the way: I started to work in the MilSpace community. And I quickly learned to hate everything about it. Not only did I learn to hate the military-industrial complex, but I quit being fascinated with space. It just seemed to me that there are too many priorities that need to be dealt with before space can be addressed. I viewed this priority shift in terms of pulling the federal government back from space, and from pulling myself back from space so I can focus on things in life which I now view to be more important.

I still have tremendous faith in the private sector. I wholeheartedly support the desires and pursuits of individuals who seek to explore space using their own money. But I no longer see manned spaceflight as an appropriate realm for NASA or for any federal government agency. I don't believe that taxpayer dollars should be used to enable some cocky fighter-jock's joyride beyond the atmosphere.

I believe that space will inevitably be explored once a commercial rationale exists do do so, or if there is an essential national defense need that can only be filled through space assets. Frankly, the justifications I've seen behind some MilSpace programs truly stretch the bounds of what could be considered a legitimate "national defense" need. I speak from the perspective that every national defense dollar wasted on extraneous space assets is a dollar that could be spent on ensuring that our armed forces have the best armaments, armor, and vehicles for the current conflicts in which we're engaged.

When I started this blog, I was still a committed believer in NASA, the Air Force, MilSpace, and the Von Braun-ian vision of nationalist glory. If you pay close attention to my old posts, you might be able to pinpoint the period in which I became the bitter, jaded individual I am today. It wasn't simply the realization that ESAS was an untenable plan for getting back to the Moon on the current budget, and it wasn't solely the increasing dissatisfaction I've had with being a pawn of MilSpace. But I think the confluence of both facts can probably explain it.

So now I've traveled full-circle, from Space Cheerleader to Space Critic. Feel free to call me a flip-flopper. You have my permission to dismiss me as bitter and my opinions as irrelevant.

Tuesday, June 17, 2008

Staging Strategies

In comparing Ares V to its spiritual predecessor, Saturn V, the difference between the two main staging methods becomes apparent. Saturn V was a three-stage vehicle, with each staging event occurring in serial. Ares V will be a "2.5 stage" vehicle, with one of the two staging events occurring in parallel.

Each approach to staging has its ups and downs. In brief, the staging comparison can be summed up thusly:

Serial staging:
++Most effective shedding of unused tankage mass
++More situations for a survivable abort
--Taller vehicle
--No way to verify that upper-stage engines work prior to staging-event

Parallel Staging:
++Shorter vehicle
++Confidence that engines work at liftoff; no ignition event during staging
--Less efficient shedding of unused tankage mass
--Problems in one parallel stage tend to quickly snowball into problems affecting the entire vehicle

I wanted to expound on these differences in a little bit of detail. Consider the difference in staging events between Saturn V and Ares V. On Saturn, the stages would burn out and drop off in succession. All of the tankage and structure associated with Stage 1 was shed prior to Stage 2 ignition. On Ares V, there are really three stages. The first stage consists of the SRB's and all of the core propellant that is burned during SRB-powered flight. The second stage is the remaining core propellant and structure. The third stage is the serially-staged Earth Departure Stage.

The inherent inefficiency of Ares V is that when the SRB's burn out, the core is unable to shed the remaining "first stage" mass: the portion of structural mass which would have contained the expended core propellants. When the SRB's drop off, Ares V is carrying a lot of dead weight along the way.

The same argument can be made ad nauseum about any rocket: it will be more efficient to have infinite drops of unused structural mass. But practical considerations, including complexity and reliability, limit most launchers to two or three stages. A simple optimization routine demonstrates that there is little more to be gained for every stage beyond three.

The other problem I really see with parallel staging is the ability for problems in one booster to grow into vehicle-threatening problems. The Challenger disaster is case in point: a fairly small amount of flame escaping through the aft O-Ring led to structural failure of the SRB attach strut, followed by structural failure of the external tank, followed by destruction of the orbiter by aerodynamic stresses.

The parallel staging argument actually works in favor of Ares I as a safer alternative to Delta IV Heavy for crew launch. If any of the three core boosters loses an engine, the vehicle is lost. There's no redundancy, since there's no propellant cross-feed system. (Besides, the loss of either outboard engine would probably make the vehicle uncontrollable.)

On the other hand, serially-staged vehicles have a potential safety advantage: if one of the lower stages leaves you a bit short on velocity, you merely ignite an upper stage engine to abort. Saturn V had several abort modes which relied on the third-stage engine or the Spacecraft Propulsion System. I'd assume that Ares I would also be capable of an abort using the Orion main engine.

With all this being said, it should be asked if any of these lessons can be applied to Ares V. The answer is "yes," but only if NASA is willing to sacrifice schedule and admit that there were mistakes in the previous course of action. As it stands, Ares V is taller, heavier, and possessesmore thrust than Saturn V. Yet the performance isn't appreciably better (depending on who you listen to for the Saturn V's performance.) It might be possible for Ares V to get by on a smaller core and carry additional propellant in a pair of drop tanks that would be drained during the SRB burn, and discarded during SRB staging. But this would add complexity to a very complex vehicle.

If Mike Griffin and company want to out-do Von Braun and his compatriots, they'd be wise to follow him more closely. Dump the SRB's entirely, and split the Ares V core into two serial stages. Stage 1 would use high-thrust LOX-Kerosene engines, while Stage 2 would be extremely similar in size, thrust, and Isp to the Saturn V second stage (complete with five J-2X engines.)

Building a new first stage similar to Saturn V's first stage will be challenging, but will probably be easier than when the Saturn V first stage was developed. The RD-171 engines currently available have slightly more thrust than the F-1, and significantly higher Isp. Another bonus is that the first stage has a fairly low staging velocity (around Mach 3,) so splashdown recovery might be feasible.

For all-out performance, I do have a preference for serial staging. And in looking upon the Saturn V, the genius of the Von Braun team becomes apparent with each passing day that today's NASA spends re-baselining the Ares V.

Tuesday, June 10, 2008

Liquid Courage

When NASA adopted the Ares launch vehicles in its lunar mission strategy, it wedded itself to a new five-segment solid rocket being developed by ATK. That fateful decision from three years ago has created particularly thorny challenges for NASA. Much effort is being spent on mitigating thrust oscillations on Ares I, while Ares V may necessitate building another new SRB with 5.5 segments.

It would have been worth asking whether a liquid-fuel booster would be a better choice than the new solid rocket. The most deceitful sales pitch behind ESAS is that the SRB was "shuttle derived." The truth is that any change in the dimensions of a solid rocket is non-trivial. As for liquid-fuel boosters, changes in length are comparatively easy (while changes in diameter involve somewhat more cost, schedule and risk, but are not unheard of. After all, the Saturn I upper stage was lengthened, widened, and re-engined to form the upper stage on the Saturn IB and Saturn V.)

What if a widened Atlas core, using two RD-180's, replaced the five-segment SRB in NASA's plans? For starters, crew launch could be accomplished with commercially-purchased Atlases using a widened Centaur-derived upper stage and minimal modifications at the Cape's LC-41. The new boosters could then be applied to Ares V, and stretched or shortened as mission needs would dictate.

Liquid fuels will always possess certain advantages over their solid brethren. The LOX-kerosene combination has better specific impulse, lower structural mass requirements, and lower density than solid propellants like the ones on the shuttle. The result is a more voluminous booster for the same mission, but one that weighs less.

The weight consideration is an important one. The behemoth Ares V weighs so much that new "super-transporters" will be built to replace the 40+ year old crawlers which transport the shuttle to the pad. The super-transporters will not come cheap. My personal preference would be placing a gross weight restriction on Ares V that would fall within the limits of the current crawlers. I would also use two Ares V's per mission, to get around the performance limitations on a smaller Ares V.

No design trade is without its drawbacks, and the Achilles Heel of liquid boosters is the limited thrust when compared to solids. It takes roughly 5.5 RD-180 engines to match the thrust of the Shuttle's two SRB's (and the new Ares SRB will have even more thrust than the Shuttle SRB.) It can reasonably be expected that three or four evolved Atlases would be necessary to replace the two SRB's in Ares V. More boosters leads to higher failure rates (even if the RD-180 pairs are replaced with RD-171's, which are functionally equivalent but possess fewer parts than the RD-180 pair.) There's a good reason why the Soviets needed four liquid boosters on their Energia-Buran shuttle system, with each booster powered by a single RD-170 (functionally equivalent to the RD-171 or two RD-180's.) The only upside is that the lower liftoff mass will mean that less thrust is needed. Still, it's not enough of an offset to get away with only two boosters for the payload masses that NASA wants.

The solid-liquid trade study is one that couldn't have been adequately analyzed during the 60 days of the ESAS study, and will likely end up as an interesting footnote in the Ares story. The question is whether the Ares story will fall into the genre of historical nonfiction, or fantasy and tragedy. If the latter is true, perhaps liquids were the answer after all. But the decision to not cap the weight of Ares V (even at the expense of payload) is one that taxpayers shouldn't forget if the massive rocket, and its shiny new infrastructure, ever get off the drawing board.

POST SCRIPT: It just so happens that the Delta IV core, at just over 5 meters in diameter, is almost exactly as wide as the SRB's aft skirt. The wide-bodied Atlas will be slightly bigger at 5.4 meters. The SRB cutouts on the mobile launch pad would seem like an ideal fit for EELV-derived boosters. But they will likely need to be moved farther out from the center of the pad, as there's only seven meters separation between the two.