Chair Force Engineer

Friday, September 25, 2009

The Other Engine

Whenever I check my gmail account, I see these odd adds from General Electric telling me that I should petition my congressman to continue funding for their F136 turbofan engine. While appeals to the public rarely have the desired effect in defense acquisition, it does raise an interesting point. The F136 program is an unprecedented development in military aviation acquisition: it's the first time that a major weapons system (the F-35 Lightning II) has been authorized with two separate engines before the first flight.

General Electric makes its argument based on their experience producing the F110 replacement engine for the F-14 Tomcat and F-16 Fighting Falcon. Yet the F110 was the product of unique circumstances that are not present in the development of today's F-35 fighter jet and its F135 engine, produced by Pratt & Whitney.

The F110 story actually begins before 1964, when Pratt & Whitney produced the first low-bypass turbofan engine for supersonic fighter aircraft. The TF30 was eventually fitted to the F-111 supersonic medium bomber, F-14 Tomcat (a carrier-based fighter,) and the A-7 subsonic light bomber. All three of these aircraft suffered from engine reliability problems. The F-111's problems were least severe, and the aircraft still flies (in the Royal Australian Air Force) with its original engines. The A-7's performance with the TF30 was so poor that Allison Engines produced a license-built version of the Rolls-Royce Spey to replace it during future production models of the A-7. For the Tomcat, which suffered most from compressor stalls, a replacement engine would have to wait until the late 80's.

The situation was not much better when Pratt & Whitney designed the F100, a fighter turbofan that was well ahead of its time when it first powered the F-15 Eagle in 1972. The engine's reliability problems were less pronounced in a twin-engined aircraft like the F-15, but they became much more critical when the engine was used in the single-engine F-16.

General Electric came to the rescue with a derivative of the F101 turbofan designed for the B-1 supersonic bomber. With some modifications, the F101 was adapted into the F110 fighter engine. The new engine became the powerplant of choice for future F-14 and F-16 production, and was retrofitted to older F-14's. Interestingly, both the F100 and F110 will power South Korea's version of the Strike Eagle (F-15K.)

Having an alternative engine waiting in the wings was a great blessing to both the F-14 and F-16. General Electric reasons that they will be the savior for the F-35 program too. The problem for GE is that the two situations are very different. The TF30 and F100 were designed when supersonic turbofans were still in their infancy. By contrast, the F135 baseline engine for the F-35 draws on mature propulsion technologies developed for the F119 engine in the F-22 Raptor. The chance of F135 becoming a dud like the TF30 or F100 are far slimmer. While the F136 ensures two engine vendors for the F-35, it's a very expensive option for the defense department to retain. It seems like an expensive bailout of GE to keep two major fighter engine manufacturers in business, more than anything else.

Friday, September 18, 2009

Turn on the lite

One curious recommendation of the Augustine Commission is their preference for "Ares V Lite," a proposed heavy-lift rocket which would fill the gap between Ares I and Ares V and be capable of launching the manned Orion spacecraft. When compared to the baseline Ares V, the "Lite" version can only lift 140 tonnes to the reference orbit instead of 160 tonnes. Bear in mind that even the "Lite" rocket has more performance than the legendary Saturn V.

With so little information in the public about Ares V Lite, it begs the question of that makes this new rocket so "Lite" when compared to Ares V? Because the Ares V baseline had switched to a 5.5-segment SRB (with a dummy spacer, to allow for an even longer core,) my guess is that "Ares V Lite" will use 5-segment SRB's similar to the one tested by ATK last week.

It's also possible that the "Lite" core stage is shorter, with the SRB attach points moved further aft on the SRB. This is necessary so the SRB cross-member can pass between the LOX and hydrogen tanks through the intertank structure.

The upper stage probably didn't shrink by much, because the propulsion requirements for escaping low earth orbit are similar (depending on whether Altair and Orion have changed in mass.) The biggest change is whether the upper stage is still expected to accelerate from Mach 12 to orbit, or if the staging velocity has changed. My recommendation to the Ares team is to invest any mass savings from the total system into systems which will reduce the upper stage's on-orbit boil-off, allowing it to loiter for a longer period of time.

My biggest question to the Ares V team is about the number and type of core engines. The baseline Ares V had six RS-68B engines. But if heating from the SRB's is an issue for the core engines, this is all subject to change. A regen-cooled RS-68R, or expendible engines based on the Space Shuttle Main Engine, might be better suited to surviving the thermal environment. They would also improve the overall performance of the rocket, at the expense of reduced thrust. Switching to regen-cooled engines results in a smaller core stage (either shorter or narrower, depending on the whims of the team at NASA-Marshall) carrying less fuel. Depending on how much lighter the core stage gets, the reduced thrust of expendible SSME's might not be an issue.

Ares V Lite might be moot if NASA doesn't see the $3B/yr budget increase requested by Augustine 2.0. Nevertheless, it would be interesting to see what Marshall cooked up, and whether it could live up to its performance estimates.

Thursday, September 10, 2009

The Shape of Things to Come

A lot of Constellation critics (particularly those who have singled out the Orion spacecraft) have questioned why we're going back to the Apollo shape for the capsule which will return astronauts to earth when the mission ends. Why not go back to the Soyuz shape? For that matter, can't we examine the shape of the Mercury/Gemini capsules, or come up with something new?

While some might attribute the Orion shape to romancing the past, there's one very good explanation for the reason why Orion looks like Apollo. It's because NASA and the industry have an extensive database of thermal and aerodynamic data on the Apollo shape, based off the capsule's performance during the Apollo program. Most importantly, this includes data from re-entries at lunar flyby and lunar return velocities.

Could the Soyuz shape make for a good lunar spacecraft? The answer is yes, if history serves as a guide. Soyuz-derived capsules flew around the moon during the Zond program, which was supposed to put a Soviet Cosmonaut around the moon before its cancellation. At one point, the Soviets looked into an upscaled, reusable version of Soyuz. Named Zarya, the new capsule would have fit within the 15 tonne estimate for a multi-man capsule equipped for low earth orbit missions. But one possible drawback to the Soyuz "headlamp" shape for future American capsules is whether American agencies and engineering firms could access the full database of thermal and aerodynamic data that the Russians have accumulated over the years. Without full unfettered access, there will be a significant hurdle to reusing the Soyuz shape.

The last viable option is the shape pioneered on Mercury and used again for Gemini. A favorite of the armchair engineers is the proposed Big Gemini, which seems to meet most of the Orion requirements and weighs in under 16 tonnes. While American firms have access to the data generated on these missions and during the development of these capsules, there's no real-world data for how these capsules behave when returning from lunar trajectories at the velocities these trajectories dictate. Perhaps McDonnell did this analysis during the 60's, but nothing beats real flight-test data. Additionally, I'd be interested to see how Big G compared to Apollo in terms of crew volume available to each astronaut. While Apollo was cramped, it was still quite an improvement over Gemini where each astronaut was prettymuch confined to a seat for up to two weeks.

Interestingly, it would appear that the SpaceX Dragon stays true to the basic geometry of the Gemini spacecraft. I don't have exact figures to see if the cone angle on Dragon is the same as on Mercury/Gemini, but they both capture the "tall cone" profile. Dragon also corrects a big problem that I saw with Big G: the awkward docking system. Since Dragon is designed from the ground up and doesn't incorporate any Gemini legacy systems, it is free to contain a docking tunnel in the "nose" of the spacecraft. Big G would have required an additional set of aft windows for docking, and introduces the risk that hot plasma could enter through the heat shield hatch during re-entry.

One last shape worth considering was developed during the CORONA program. Again, this would need to be subjected to rigorous analysis before it was ready for a lunar flight program.

It's not so easy to qualify a new shape for a re-entry capsule, especially if it's going to be returning from the moon. Re-inventing the Apollo capsule might not be the optimal solution, but it's fast one that gives the space program a proven result.

Thursday, September 03, 2009

Orion Revisited

Let's assume for a minute that the post-Augustine national space policy emphasizes the commercial development of a human spacecraft to replace the space shuttle. Obviously SpaceX has the inside track with Dragon, which might be capable of manned flight by 2012. But where would this leave Orion in the grand scheme of things? It might continue at a reduced funding level as a backup to the commercial capsules, but some strategic decisions by Lockheed Martin (and perhaps Bigelow Aerospace) might put Orion back on track to thrive in the commercial space market.

Unlike it's intended Ares I booster, the Orion spacecraft is a generally sound idea. It's rooted in mature technologies and design concepts that would assure safe human return from the moon. (A biconic design might be better for returns from Mars, but there's a lot of controversy about which shape is best for the Mars return vehicle, and how far off that objective really is.)

Orion's problems up to this point have stemmed from NASA's shifting requirements. Until Ares I flies (if that happens at all,) the booster's performance (and Orion's mass budget) are uncertain. The original Orion specifications called for a maximum of six passengers, the ability to operate autonomously in lunar orbit, enough consumables to support four humans during a lunar round trip, and landings on terra firma. The new Orion specifications are for a four-passenger capsule that will likely maintain a pilot while in lunar orbit, and will land in the ocean at the mission's conclusion. Even the capsule's diameter changed, shrinking from 5.5 to 5.0 meters early in the design cycle.

Orion's mass budget always seemed optimistic in my view. The Apollo capsule, designed for a maximum crew of five and sized to a 3.9 meter diameter, weighed in around 30 tonnes in its lunar variant (although the earth-orbit missions loaded less propellant and consumables, weighing in under 15 tonnes.) The mass target for Orion was under 23 tonnes, even though it was sized for 5 meters diameter. Mass savings can be attributed to less propellant (Orion doesn't perform a lunar orbit insertion burn) and using solar panels instead of fuel cells.

Maybe these mass savings offset the added mass of the larger capsule when comparing Orion to Apollo. It's hard to say from my "Monday Morning Quarterback" chair. But it does seem safe, based off the Apollo experience, that a 5-passenger capsule with lunar-capable heat shield would weigh in just under 15 tonnes when configured for missions to the space station. Previous NASA estimates for comparable capsules, conducted during Mars design reference mission studies, came up with a mass under 10 tonnes. This lines up pretty well with SpaceX's estimates for the mass of a fully-loaded Dragon.

With all that being said, LockMart's work on Orion up to this point is far from a waste. If nothing else, LockMart should press on with its own money to get a commercial version of Orion flying (if the commercial capsule option comes to fruition.) Bigelow Aerospace has already proposed "Orion Lite" to launch on an EELV by 2013, but it's unclear if this idea has any official support from Lockheed Martin. And it's clear that for Orion to be commercially-competitive, it will need to fly in a "lite" version without all of the lunar frills.

A commercial Orion would need to keep its mass low, to fit on existing commercial launchers. Even if the capsule was scaled back to Apollo's 3.9 meter diameter, this may still be a tough order. After all, the "Zero Base Exercise" from the 2007 time frame cut Orion back to dangerously low redundancies in critical systems. Redundancy was only restored if extra performance squeezed out of Ares I freed up the mass budget to put it back in.

At the same time, a commercial variant of Orion would be free from NASA's fickle whims and requirements creep. There's no reason not to touch down on dry land. Soyuz has been doing it for over 40 years, thanks to six braking rockets in the capsule's heat shield. Orion could also move to two rows of seating and add more paying passengers (something currently banned under NASA's human-rating requirements, although it's featured in commercial capsule designs like Dragon.)

In short, commercializing Orion would give Lockheed Martin a head-start in the race for commercial orbital spaceflight and make for an interesting race between Orion and Dragon. But Orion would need to go on a massive diet, and the commercial Orion would be a very different beast compared to today's Orion designs. Ultimately it's in America's best interest to have at least two commercial capsules in operation, and it would be exciting to see Orion continue on in this fashion.