• Configuration
• Ares I-X
• Design Reference Mission

• Configuration
• Design Reference Mission

• Lunar Orbit Rendezvous

• Safety
• Practical and Affordable
• Common Hardware
• Heavy Lift Capability

• Ares I
• Ares I-X

• Ares IV
• DIRECT Launch
• FAQ

• News Articles
• Image Gallery

• Space Foundation
• Space Coalition
• The Planetary Society
• NASA

For more information regarding Exploration Launch Vehicles, contact: safesimplesoon@atk.com

©2007 Alliant Techsystems Inc All Rights Reserved

Myth Busters

DIRECT Launch

One of the biggest clubs that has been used recently to pummel the Ares I program is the "DIRECT" launcher concept, which called for an inline launcher with two standard RSRMs, an ET with five RS-68's underneath and an upper stage. The idea was to go with a single common launcher, rather than the Ares I and Ares V launchers. The story was that it would be easier, faster and above all cheaper.

Small problem: it won't work as advertised.

Dr. Doug Stanley, who headed the ESAS study, was asked by Ross Tierney (the main DIRECT spokesman) to examine the DIRECT proposal. Here are some of his observations, which also provide some interesting and useful data on the Ares I:

As I noted above, there are a number of major problems with the "Direct" concept and claimed advantages, both technical and cost-related, but I will begin with the largest technical one, the RS-68 REGEN engine assumptions. The Direct vehicle concept was analyzed by NASA and reviewed by me using the same models and assumptions used by ESAS and the current Ares V program to ensure apples-to-apples consistency. When this was done the vehicle came up way short of the claims for payload capability. It was over 16 tonnes short to LEO and 19 tonnes short to TLI — per launch. This means that the approach is not even close to being able to launch the lunar mission in two launches.

One of the main reasons for this is the use of RS-68 Regen vacuum specific impulse assumptions of 435 sec rather than less than 420 sec as verified with NASA and Pratt & Whitney/Rocketdyne. The thrust/weight was also found to be higher than possible. Detailed performance analysis was run at the maximum power setting of the RS-68 and using a regen nozzle to gain additional performance. This was run by the RS-68 contractor and verified by NASA's own internal analysis. After consultations with the VP in charge at Pratt & Whitney and the Ares V Program Manager, they agreed to let me publicly release these performance numbers so they would not be from "anonymous sources". They are willing to stand behind these numbers (posted below)! Neither they nor I have been able to find any possible source of confusion that could have led to anyone quoting the higher numbers, which Ross said he got second hand from a source that is anonymous to me.

The second less major reason for the performance shortfall is in the assumptions about the amount of propellant that the core can actually hold. Because of geometric and structural considerations, a core vehicle of this diameter cannot hold more than 1.6Mlbs of propellant due to tank clearance issues with the structure required to react SRB loads and thrust structure packaging considerations. A very detailed analysis of this was done during ESAS when we did the LV 24/25 concepts. For a given diameter core, going to 5-segment solids has the added advantage of allowing you to make the hydrogen tank longer to hold more propellant.

There are a number of more minor discrepancies, but Ross and I agreed that the RS-68 issue alone is enough to make the concept infeasible, so I won't bother discussing them here.

I don't want to be pulled into a discussion of costs, because the numbers are SBU, but the claims of significant cost savings are just not correct. I have every cost number we generated at the most detailed level. I don't recognize at all the claimed numbers for CLV or CaLV DDT&E costs. The DDT&E costs for the ESAS CLV is a factor of over 3 less than $16.8B. Even if you add up all of the DDT&E plus the three test flights and all KSC ops costs and other "keep alive" costs, they don't add up to a number like that.

The cost of developing a heavy lift vehicle big enough to do the ESAS lunar mission in two launches (which the one in the proposal is not capable of doing) will be a few billion more than the baseline ESAS CLV. You need a vehicle at least as big at the LV 24/25 vehicle from the ESAS report. BTW, despite the claims in the proposal, we did look extensively at this issue. The 24/25 vehicle was not quite large enough (but very close) to launch the final ESAS lunar mission even with the suborbital burning of the EDS. There was also a bit of a mismatch between the needed launch mass for each vehicle (EDS on one and LSAM/CEV on the other — as there also is with the Direct concept). Even doing the "split" LOR mission with two smaller EDS was marginal and this was unattractive for other reasons... Hence, we needed a larger vehicle (with 5 segment solids) to do the two-launch solution with sufficient performance margin.

We could back off in requirements and make it fit on the 24/25 system with 4-segment solids. This was still a possible option at the end of ESAS, since we selected a 4-segment CLV. One reason we selected a 4-segment CLV was to preserve the maximum flexibility in future CaLV decisions. Although the final report presents the 1.5 launch solution as the preferred option, we preserved the option of eventually going to the two-launch option (with 5 or 4 segment SRBs) which also looked quite attractive. Our intent was that further study would confirm the best answer. When ESMD later changed to a 5-segment solid on the CLV (which I don't necessarily agree with), the four segment option was taken off the table...although the 5-segment booster CaLV two-launch option remains...but NASA prefers the 1.5 launch option because it 1) has a better LOC and LOM, 2) can land significantly more cargo on the lunar surface in a single launch, 3) provides a larger vehicle for Mars missions, and 4) allows you the flexibility of launching humans to LEO or cis-lunar space without having to carry cargo also with a more costly vehicle.

As the Direct proposal points out correctly (and as we did in the ESAS report), the two-launch option has somewhat lower annual operating costs and life cycle costs. If Ares 1 is developed and then decommissioned, the 2-launch option has a LCC of a few billion less that the 1.5 launch option.

This gets to the main unique issue with what was being proposed by Direct...why not forego the development of a CLV altogether and save some money.? Although this is a moot point now, we did look at this, and the life cycle cost savings (purposely vague...) was less than $5 billion (not $17B). The problem is that, despite any of your claims to the contrary, heavy-lift vehicle capable of a two-launch lunar solution cost a few (purposely vague...) $Billion more through first human flight due to the higher DDT&E cost relative to the simple ESA CLV and the higher costs of the 3 flight tests. Despite assertions to the contrary the Direct core vehicle or the ESAS 24/25 core vehicles are almost completely new hardware developments with little STS ET heritage other than materials and diameter because of the different load paths. The development time was also found to be over 2 years longer (using detailed apples to apples schedules). The schedule driver now is actually the available budget...not technical considerations. NASA wanted to close the gap as quickly as possible and wanted a system to go to the ISS with high safety (the ESAS CLV is significantly higher LOC than a HLLV).

I will also take this opportunity to address on the record some of the alleged "issues" with the ARES 1 vehicle from "anonymous sources" that have been discussed in this forum and certain NASA-Related-Personal-Axe-to-Grind-Single-Source-is-Good-Enough-Blog sites.

An entire section of the Direct proposal is devoted to alleged "Flaws With the Ares Launch Vehicle Family". Very little that was written in the section concerning "problems" with the current Ares Program is correct. The premise that the current (or original ESAS) Ares 1 approach is "broken" and needs to be "fixed" by something like what is being proposing is simply not correct! I will attempt to address some of them in this section for the record in one place. All of the data in these responses come directly from the knowledgable NASA people in the responsible engineering or program office...

DIRECT Assertion: "The original design, of 4-segment SRB with Space Shuttle Main Engine Upper Stage, would probably have lived up to expectations — if the SSME could have been air-started. It can not. NASA is left with a compromise which attempts to fulfill the same requirements, but which fails to."

NASA Response: This is not true. NASA was confident in its plan to air-start the SSME and no showstoppers were identified at the time NASA elected to change the Ares I baseline. NASA switched to the 5 segment/J-2X approach to achieve greater commonality with the Ares V, reducing the number of developments required — resulting in significant development and recurring costs savings (billions). This included moving from 2 SRB's (4 and 5 segment) to one (5 segment), 2 upperstage engine developments (alt start SSME and J-2X) to one (J-2X), and moving to a low cost, commercially developed core stage engine flying on the Delta IV today (RS-68) vs. an engine unique to NASA needs (SSME derivative).

DIRECT Assertion: "The 'Stick' Crew LV's biggest selling point was its high safety figures. However, the difficulties the design is suffering from today are continually whittling those away, with each 'fix' causing ever larger penalties to the performance.

NASA Response: This is not true. NASA currently projects a loss of crew of 1 in 2,150 — a robust vehicle when compared to STS and with any other alternatives evaluated and consistent with ESAS projections.

DIRECT Assertion: "The new 5-segment SRB's and J-2X engines are both completely unproven."

NASA Response: This is not true. A 5 segment ground test motor was fired in October, 2003. The J-2X is a derivative of the Saturn J-2 and J-2S engines, elements of which (turbopumps) were recently utilized on the X-33.

DIRECT Assertion: "Together their performance is so desperately low that other parts of the vehicle are having to be designed down to dangerously minimal weight, in order just to get the system to fly at all. Performance of just 22mT -30x100nm 28.5deg is at best, mediocre, at worst, anemic. This poor performance is causing detrimental domino effect throughout every phase of the development of the new vehicle."

NASA Response: This is not true. The original ESAS baseline CLV delivered ~27mT (without performance margin) to LEO/28.5°. This was with a much lighter launch abort system and before wind tunnel data was available. Using much more detailed models, the current Ares—I is projected to deliver ~26mT (without performance margin) to LEO/28.5° — equivalent to ESAS. Orion is being designed to weigh no more than 22mT (in ESAS, this was 23mT, but was a 5.5m diameter capsule). Ares I will be the largest heavy lift capability in the U.S. until Ares V is developed.

DIRECT Assertion: "A normal rocket is naturally stabilized throughout its flight by having the Center of Gravity (CofG) ahead of the Center of Pressure (CofP). Like a thrown dart, the rocket will naturally fly nose-first. But the Ares I's CofG is behind the CofP — which causes the rocket to want to flip around in mid-air. Only with very precisely applied Thrust Vector Control, can the rocket be kept on track without applying very high stress loads to the structure. The first stage has a very slow Thrust Vectoring system, simply because it is a Solid Rocket Booster. This is causing concern during the first minute after launch, before speed builds and aerodynamics affect the ascent. It is the job of the SRB's Thrust Vectoring system to keep the very tall and ungainly rocket stable and pointing in the right direction as it lifts from the Pad. It is a problem often equated to balancing a pencil, on end, using your finger. The nozzle at the bottom of the SRB is proving to be a very slow 'finger' performing the balancing act. If the rocket becomes unbalanced, perhaps due to crosswinds, the nozzle may be too slow, and be forced to apply very high bending moment forces on the structure in order to try to re-stabilize."

NASA Response: This is not true and shows a lack of understanding of large rocket design. Typically, large, orbital capable rockets have a C.G. aft of the C.P., hence you utilize a TVC system. NASA has conducted over 1,500 wind tunnel tests of the Ares I configuration, and conducted analyses on the flight control system design. While Ares is a long and slender vehicle, it is within the control dynamics experience base of previous programs, most notably the Saturn V. 6DOF simulation results indicate a ~2x margin on first stage thrust vector control (angle and rate) and an ~8x margin on the vehicle structural response to control frequency ratio.

DIRECT Assertion: "The two issues above can cause forces which, quite literally, try to bend the vehicle in half. The SRB is a very strong structure. The pressurized Upper Stage tanking is also a very strong structure. But the Interstage between them is a hollow cylinder, 18ft (5.5m) wide, and 40ft (12m) long, with walls only 1.25" (3cm) thick — and complicated further by a conical structure changing diameter from 13ft (3.9m) to 18ft (5.5m). The Interstage will be the "weak point" if the vehicle suffers instability issues during flight. It is the structure which would fail first if the rocket goes off-course and takes too much time to be forced back on course. The Ares I test vehicles' Interstages are being specifically over-built to combat this problem in a bid to dissuade disparaging comment from the space community, who is already well aware of this concern. But the final flight versions of Ares I must be built down to the lowest possible weight limits in order to keep performance high enough — which means this will be the weakest structural point in the final design. The SRB first stage is currently 18,000lb overweight because the seals around all of the segments need additional, unplanned, strengthening. This is because the in-line design, with the stage and payload located above the booster instead of beside it, are experiencing different loads during flight from the SRB's intended design — so require additional strengthening at these joints to compensate."

NASA Response: This is not true. While the Shuttle RSRM was not originally designed to have a second stage ride atop it on the way to orbit, this is a very robust stage which carries the entire load of the Shuttle External Tank and offset load of the Shuttle Orbiter. In addition, it was sized to carry the offset load of 3 Space Shuttle Main Engines firing at ignition ("twang load") which Ares will not have due to its single engine, in-line first stage configuration. The NASA team is using proven, validated engineering tools and loads models and conservative margin factors at this stage of the design. Analyses performed to-date indicates that the existing Shuttle RSRM cases, joints and aft skirt have sufficient design capability to support the Ares in-line configuration and are not "overweight" as characterized above. In addition, the upperstage and interstage are being designed for the loads expected on the ground and in-flight. The upcoming Ares I-X flight test in 2009 will give NASA important data early in the development cycle.

DIRECT Assertion: "The roll-control system was not predicted to be as considerable an issue as it is proving to be. It requires an extra system which was unplanned originally, which impacts the weight of the vehicle, and increases the number of systems which can cause an expensive Loss of Mission or, worst of all, a Loss of Crew contingency."

NASA Response: This is not true. Characterizing and controlling roll torque has been a high priority since Ares' inception. NASA has utilized what it believes are worst case roll torque predictions and then designed the control system to handle 1.7 times that torque using RCS thrusters. Our goal now is to further refine the roll torque predictions through ground test firings of the motors with calibrated sensors, analyzing similar launch systems (Athena, for example) and the Ares I-X flight test in 2009. We believe we have utilized very conservative predictions and then used a conservative design approach.

DIRECT Assertion: "The original "Stick" launcher utilized the Upper Stage to reach an initial elliptical orbit of 60x160nm, then that Upper Stage to then perform the Circularization burn to achieve the stable 160x160nm orbit. The Orion is now required to perform a 1000ft/s high-Delta-V burn to reach an initial orbit of just -30x100nm — that means the low-point is 30 nautical miles under the Earth's surface."

NASA Response: This is not true. In ESAS and until last Spring, the Ares—I injected the Orion into a 30x160nmi transfer orbit and the Orion then circularized itself, to avoid the complexity of deorbiting the large upperstage. Working with Constellation and CEV project teams, the program elected to change to a -30x100nmi orbit to move the ocean impact of the CLV upperstage to the Indian Ocean from the South Pacific to stay away from populated islands. This also allowed the impact point for both ISS and lunar missions to be in the same general vicinity. Appropriate performance was transferred from Ares to Orion so that the spacecraft was not penalized. Performing multiple OMS types burns is commonplace on STS today and does not increase risk. Also, Orion does not have to do a burn to reach -30x100nmi — Ares places it in that orbit. Orion carries 1,000 ft/s to perform all orbital maneuvers, including transferring from -30X100 to 160 circ, and on to 220 for ISS, rendezvous, prox ops and docking, and deorbit.

DIRECT Assertion: "Together, this reduces the original 'Stick' concepts Loss of Crew (LOC) figures below the stated 1 in 1918. The Ares I's fundamental design requires that the Upper Stage engine be ignited at altitude, only after the SRB First Stage has burned-out. There is no guarantee that any engine will start correctly, or safely, let alone at altitude. If there is a problem, the mission would become an abort, requiring the use of the escape system. NASA has yet to publish new, independent, 'apples-to-apples' comparison safety figures between the original CLV and the current evolution. The figures will obviously be lower today. Loss of Crew (LOC) safety figures of between 1 in 1500 to 1600 are rumored for the current risk factor as this paper was compiled — so the gap to DIRECT's 1 in 1355 LOC risk is now very narrow indeed.

NASA Response: This is not true. The current Ares probabilistic risk assessment, which is much more comprehensive than the PRA used in ESAS is predicting a loss of loss of crew of 1 in 2,150 (mean). This is ~1.6x the DIRECT claim of 1 in 1,355 LOC (which is not supported with any analysis — the best ESAS vehicle in this "direct" class had an LOC of 1 in 1,170).