Flight of the Griffin

Last week, Michael Griffin departed his post as NASA administrator to considerable controversy, surrounded by a legacy that will remain tenuous and uncertain for many years to come. Mike Griffin is ESAS, and ESAS is Mike Griffin. The two of them will never be separable, and the success or failure of ESAS will likely mold the verdict that historians deliver on the Griffin administration of the agency.

With that being said, the positive accomplishments of Mike Griffin and his team have often gone neglected. The agency was responsible and courageous in the way it conducted the shuttle program as it returned to flight following the Columbia disaster. Griffin also deserves credit for the way NASA handled the COTS program, or at least for its sponsorship of SpaceX. At this point in time, SpaceX remains America's best hope for closing the spaceflight gap within a reasonable period after the shuttle retires.

At the heart of Mike Griffin's relationship with ESAS and Project Constellation is a fundamental misunderstanding of what the person in the position of "NASA Administrator" is supposed to do. For any management figure, the keys to success are cost, schedule & performance. Specifically, Mike Griffin was tasked with performing the following responsibilities during his 2005-2009 tenure with NASA:
--Keep his agency on-budget in meeting its mission
--Keep his agency on schedule as it meets its goals
--Ensure the agency's performance meets the thresholds & goals that have been set by the national leadership

During his tenure, Mike Griffin's NASA largely abandoned the presidentially-directed 2014 date for getting Orion operational ("operational" being the key term here, although further slips in the Orion schedule are highly likely,) and coalesced around an architecture which requires unplanned budget increases. The result is that the agency's ability to meet its goals, performance, is much less likely.

I would be remiss today if I did not mention Paul Spudis's excellent piece about why Project Apollo was an exception to the rules that traditionally govern national space programs. Apollo's success was due to an unsustainably large development budget. Mike Griffin's challenge, back in 2005, was to figure out how to get back to the moon while operating on a shuttle-sized budget that would only grow to keep pace with inflation. Rather than making cost-control a top priority, he signed on to "Apollo on Steroids" with a shuttle budget. Development would be stretched as long as possible in order to stay within the yearly budgets, which puts the schedule at risk.

If America fails to land a man on the moon by 2020, an argument will be made that if cost was a more serious factor during the ESAS studies, the original program schedule could have been maintained. The counter-argument is that cost-control cannot be used to justify an architecture that does not meet minimum safety requirements. At this point, I believe that NASA can do far more to reduce costs without compromising safety. It begins with a detailed analysis of the baseline mission requirements of how many man-days are required on the lunar surface, and how much volume each crew member requires to make the journey. These requirements are given as Gospel truth in ESAS with little or no justification.

The buck ultimately stopped with him, but Griffin in many ways served as "chief engineer" while forgetting all that is entailed with being an administrator. I would contend that Mike Griffin spent far too much time dictating an architecture to his own employees and then defending it to the world, instead of focusing on letting the engineers develop an architecture which fit within the cost, schedule & performance constraints placed by the nation's leaders. Mike Griffin should have been steering his agency towards meeting its cost, schedule & performance goals. Instead, he took the helm of a stalled Vision for Space Exploration and steered it firmly off-course with his engineers being keel-hauled underneath.

In spite of all the Beltway rumors surrounding the possible choices of the next administrator, I would hope that NASA's next chief comes to the job with experience as a capable manager of successful big-budget programs. His or her success should not be measured by what the changes to the architecture look like, but rather by how much cost and schedule they save the taxpayers.

A Reprieve for SSME?

Rob Coppinger suggests that the days of the Space Shuttle Main Engine may not be over when the shuttle system is retired in 2010. When the shuttle program ends, there will be a number of engines that have not exceeded their lifetimes. It’s conceivable that a commercial rocket could use them on a single-use basis, although it’s unlikely that anybody would want to sink millions of dollars into developing a rocket if the supply of engines is finite.

Another (remote) possibility is that the SSME’s may find their way back onto the Ares V, as per the original ESAS studies of Summer 2005. The thought of recovering the SSME’s in a fashion similar to that proposed for the Atlas V engines had occurred to me. It’s harder to pull off a recovery of this nature on a cluster of six SSME’s, but it’s certainly worthy of study. It would improve Ares V’s performance to orbit (SSME has higher specific impulse than the RS-68,) but the thrust levels would drop off. The change to a six-engine cluster (versus the five engines on the original Ares V) might help offset that. Conversely, it might allow a switch back to the shuttle-derived 8.4-meter tankage instead of the all-new 10 meter tankage of the current design, and avoid the expense of changing the spacing between the SRB cut-outs on the Mobile Launch Platforms.

On a related note, NASA is looking at other uses of Ares V, especially for science missions. I must admit that the prospect of a massive space telescope, used to spot earth-like worlds around distant stars in our galaxy, is too exciting to ignore. At the same time, it's not realistic to think that Ares science missions will appreciably increase the Ares launch rate and amortize the big booster's massive standing army costs. Big-budget science missions on the scale of Hubble Space Telescope only launch once every few years. The obvious way to double the Ares launch rate is to dump both Ares I and Ares V in favor of two launches of an intermediate-capability rocket. You know, something that looks kinda like Jupiter-232. Or you can further amortize the existing standing army costs for the existing Delta or Atlas rockets by using six EELV's or so per lunar mission. The point is that the inherent fiscal inefficiencies of the Ares system will not be fixed through the addition of an occasional science mission to the manifest. Only a marked reduction in fixed costs or a steep increase in flight rate will make the system more efficient.

Boundary Conditions

The staff of LaunchSpace wants everybody who cares about the fate of Project Constellation to keep quiet and let the big boys at NASA get on with business of going back to the moon. It's a nice sentiment, if you buy into the belief that NASA is nigh-infallible and that their plan will get us back to the moon with a reasonable amount of time and money.

At the same time, the most visible of the alternatives to Project Constellation is getting exposure in Popular Mechanics and even got a meeting with the Obama transition team. The promoters of DIRECT may see the change in administrations as their best (and perhaps final) chance to salvage Project Constellation from the budget and schedule nightmares that lie ahead.

We have a situation where multiple groups all believe they have the best solution to the challenge of going back to the moon. For what it's worth, twelve industry contractors submitted Crew Exploration & Refinement studies in Fall 2004, and their results were drastically different from either DIRECT or the NASA baseline.

How could fourteen different studies result in fourteen different conclusions about the best way to go back to the moon? The answer can simply be summed up as a difference in assumptions and evaluation criteria for what is truly best. How big of a crew are you sending to the moon? How many days will the mission last? What sites on the lunar surface will be accessible? How much volume does each crew member need in the capsule?

The criteria and assumptions that went into the ESAS study of Summer 2005 have been widely and justifiably criticised. The expendible SSME and air-start SSME which formed the basis of the early Ares designs were unrealistic. It was assumed that the existing Atlas and Delta rockets could not tailor their trajectories to enable aborts during all flight phases. On-orbit assembly and orbital refueling were viewed as highly undesirable, at best. It's enough to make one wonder what other flawed assumptions and judgement criteria went into the study.

While Ross Tierney relied on post-Challenger design studies for his original “DIRECT 1.0” concept, it should be noted that the Jupiter design was independently validated by Stephen Metschan using his FrameWork CT optimization software. FrameWork CT evaluated numerous configurations in a fashion similar to ESAS, but with different ground rules and assumptions. Neither approach is perfect, but I’m inclined to believe that ESAS is heavily flawed based on its low-balling of the cost for the five-segment SRB and other invalid assumptions.

The subjectivity of evaluation criteria is clear from the simple question whether Ares I is safer than Jupiter 120. The answer is conceivably yes, because it has less engines. But Jupiter 120 mitigates this by lighting both RS-68’s on the ground prior to igniting the SRB’s (which have an incredibly reliable ignition system.) Jupiter 120 gives you the confidence that all engines have been tested at liftoff. By contrast, Ares I relies on a staging event in which you hope and pray your second-stage engine ignites. Additionally, Jupiter 120 has more performance, allowing for greater redundancy to be added to the Orion capsule. This is the area where evaluators have to look at the probability of a staging failure on Ares I versus a loss of engines early in the Jupiter 120's ascent, or the probability that extra redundancy in certain capsule systems will allow the crew to survive in an emergency scenario. The analysis of these probabilities, and the weights given to predicted reliability numbers, present plenty of room for the experts to debate until they are blue in the face.

Every design will also possess unintended consequences that must be worked out once it has been chosen. Prediction of these consequences is a reflection of how valid and effective the original design trade was. For Ares I, launch drift and thrust oscillation are the challenges that we’ve bought. Jupiter would certainly have its own challenges, although it’s hard to see them being worse than launch drift or thrust oscillation.

I would like for the next NASA Administrator to call time-out and order a re-evaluation of crew and cargo launch strategies that takes development costs into account with infrastructure and operational costs for the expected duration of Project Constellation (from now until at least 2025.) The agency should look at permutations of all realistic crew launch & cargo launch designs. Examine Ares I, Jupiter 120, Atlas V Heavy, Delta IV Heavy, and Wide-Body Atlas for crew launch. Take a gander at Ares V, Jupiter 232, and a side-mount Shuttle Derived Vehicle similar to Shuttle-C. Take a realistic look at the assumptions which are driving the Orion capsule weight (especially the amount of volume available to each crew member) and the number of man-days the Altair lander is expected to support on the lunar surface.

The Atlas of LC-39

Sir Sydney Camm famously observed that “All modern aircraft have four dimensions: span, length, height and politics. TSR-2 simply got the first three right.” He was referring to Britain’s stillborn attempt to design a medium jet bomber during the 1960’s, but the same could be said about using Atlas or Delta for space launch during the post-shuttle era. If we accept that NASA must retain government jobs, Atlas & Delta will not be acceptable solutions to the problem of manned space launch after the shuttle retires.

An option I have thought about (and I certainly was not the first) is whether the shuttle’s launch complex could be converted to launch Atlas or Delta rockets while retaining much of the shuttle workforce. Such an option is not economical, from the perspective that it’s not leveraging the infrastructure & personnel investments that United Launch Alliance has already made. But the idea of a NASA-run space launch program is inherently un-economical already.

From a technical perspective, it would not be too difficult to convert the shuttle launch facilities over for Atlas or Delta operations. It was already accomplished once, when the mothballed SLC-6 at Vandenberg was rebuilt from a Titan IIIM pad to a Shuttle pad and to a Delta IV pad. Prior to the conversion for Delta IV, Athena rockets were launched from a specially-designed stand placed over the flame trench for the shuttle SRB. It’s not hard to imagine a single-core Atlas or Delta being supported over an SRB flame-trench at Kennedy Space Center, using a modified version of the current pad’s crew elevator and access tunnel supported from the umbilical tower.

The Atlas V launch pad is designed with the umbilical tower already mounted to the transporter. One would conceivably need to be added to the Mobile Launch Platform if Atlas is flown from LC-39. Delta IV is transported horizontally, so the MLP would need some mechanism to erect the rocket at the launch site. Neither issue is non-trivial. But the alternative of modifying the MLP’s for Ares I also involves significant work.

I can see the Defense Department supporting the idea of establishing separate EELV pads at LC-39. It is always nice to have a backup launch pad in the event that an exploding rocket damages one of the existing launch pads on the east coast (LC-37B or LC-41.) Assuming that Ares V survives the anticipated cuts to Project Constellation, one of the pads at LC-39 will likely be reserved for the new super-rocket. The other launch complex could conceivably be designed for both Atlas and Delta operations, but the Mobile Launch Pads would be drastically different. I would anticipate that only one of the EELV’s will make the cut. My preferred configuration would be a wide-body Atlas with two RD-180 engines on the first stage, but a three-core Heavy EELV would be cheaper from a development perspective.

Recent rumors have indicated that the Obama Administration plans to euthanize Ares I in favor of Atlas, Delta or both. If the rumblings are true, the space community should be warned that the substitution will not be a simple slam-dunk. NASA will have to re-examine its workplace retention issues and its relations with United Launch Alliance. Ultimately NASA will find a way to retain as many existing jobs as possible through re-use of LC-39, regardless of how expensive and unnecessary it may be.