maandag 25 november 2013

Delta IV for exploration

NASA's current focus is exploration, and for exploration of anything, you will need some kind of infrastructure, whether the goal is the moon, asteroids, or Mars. NASA is currently working on two vehicles that are supposed to provide the foundation of all exploration architectures for the coming decades: the Space Launch System and the Orion crew vehicle. These two vehicles are very useful for building a deep-space architecture, but they aren't guaranteed a future, especially with sequestration and other political nonsense lately putting many big NASA projects in danger. And at $1 billion and $1.4 billion annual for Orion and SLS respectively, they are some of the biggest projects around. Therefore, it's always necessary to hold some backup plan around in case Congress goes nuclear on NASA's budget. Here I present an plan that might be part of such a backup plan, that could get humans to places in a more budget constrained environment. 

Delta IV launch vehicle

The Delta IV launch vehicle, in its Heavy configuration, is the most powerful launch vehicle currently available, capable of bringing 28,790 kg to a Low Earth Orbit [1]. The launcher is very reliable, having suffered only a single failure, and is completely American and provides lots of upgrade potential. Delta IV is not exactly cheap at this moment, with a price tag of approximately $370 million[2], but the main reason for this high price tag is the very low flight rate; only once a year for Heavy, with about 3-4 launches per year for the complete Delta IV family. This low flight rate is because of lack of demand, caused by the competition of Atlas V. Increasing this flight rate by two launches and six additional cores every year would reduce the per-unit cost by a decent amount.  

The reason for picking Delta IV for this comparison was the upgrade potential and the fact that all the infrastructure for launching large amounts of cryogenic propellant to orbit are all already included; something that does not exist yet for Falcon Heavy. Delta IV is also being used for an Orion flight test next year, and is therefore likely closer to being adapted for launching Orion than Falcon heavy or Atlas V heavy is. Delta IV Heavy already exists, unlike Atlas V Heavy or Falcon heavy. Also, unlike Falcon, upgrades to Delta can be partially paid for by the Air Force. All in all, upgrading Delta IV for the job could likely be done quicker and cheaper than either Falcon Heavy or Atlas V. However, they remain strong alternatives, and in case of Congress going nuclear, it's up to NASA to decide which one is better.

These alternatives, as well as Atlas Phase 2 and others, can easily be copy-pasted in this paper mission, and I see no reason why it couldn't work. Delta IV was simply assumed here for reasons mentioned earlier.

The exploration gateway

For exploration, some infrastructure to support it will be needed. By far my favorite proposal to do exploration is Boeing's Exploration Gateway plan. This option allows for lower cost access to the lunar surface, and provides a great place to do deep space research and as a staging point to asteroids and Mars. In this post I will focus on the lunar architecture, like I usually do. It can be expanded to Mars later.

The plan I'm presenting here would not dump Orion, but rather, it would dump SLS. As much as I personally support SLS, it's one of the more likely to be canceled and it's more expensive than Orion, and potentially also more replaceable. 

Assembling the station

The Gateway station plan includes a small space station to be assembled behind the moon, at the EM-L2 point. For this plan, the space station would consist of two main smaller components which can be launched in two launches with the Block 1 SLS. The first launch would carry the Science/Power module, the second one the Node and Utility module. They would be docked at the L2 point and a third SLS brings the Orion spacecraft there to man the station and start doing science. The architecture assumes SLS can lift 27 tons to TLI; Delta IV Heavy probably can't even get half of that there in a single launch, so a slight change in architecture will be required.

ACES compared to Centaur and DCSS
In order to reach the 27 TLI goal, a dual launch is obviously required. However, even with a fully fueled DCSS in orbit, the TLI capacity falls a few tons short. In order to reach the goal, an upgrade to Delta IV will be required. The ACES (Advanced Common Evolved Stage) allows Delta IV to lift a whopping 37 tons to LEO [1]. It carried 41 tons of propellant and has an empty mass of ~3.6 tons, and uses four RL-10 engines with a specific impulse of 461.5 [3]. Using a propellant drop tank, it can refuel itself in LEO; such a drop tank could realistically carry about 34 tons of propellant, allowing the ACES to refuel itself to about 83% of maximum capacity. With a propellant load of 34 tons and 1% prop residuals, the ACES stage can send 28.8 metric tons to a TLI trajectory with a delta V of 3200 m/s. While this is not using the most accurate numbers in existence, it nonetheless provides 1.8 metric tons of margin, as well as the not-insignificant launch margin from the payloads (ranging from 1.6 to 4.6 tons of margin). 

To recap, in order to assemble and man the station, Delta IV would launch six times:
1. DIVH Launches drop tank, upper stage refuels in orbit
2. DIVH launches Science and Power module
3. (see 1)
4. DIVH launches node and Utility module
5. (see 1, 3)
6. DIVH launches an Orion spacecraft with a crew of four astronauts

The parts would rendezvous and dock at L2; Orion, the Utility and Science module all have their own propulsion system with sufficient delta V to dock.

Lunar surface missions

The Boeing HLO lander
In the Boeing plan, there are two architectures described for landing on the lunar surface. One uses the SLS upper stage as a crasher stage and a methane-oxygen lander. The other one uses a single-stage two man lander with storable propellant and a LTV to get the lander from HLO to LLO and back. The architecture that is most suited for a Delta IV based architecture is the storable propellant lander, since it does not use any parts bigger than SLS block 1 allows for. [4]

A humble change to the architecture is to use Orion's SM instead of ATV as the LTV. ATV production will stop after number 5, and there is not chance for revival. Orion's SM, which is ATV derived, will be able to operate independently from the crew module (required by law) and is European anyway. One of the big reasons to use a modified ATV is to allow for international cooperation to share the costs, but if ESA provides and maintains the Orion SMs for this mission it will do that just as well. Assuming a lander mass of 21.4 tons with 6 tons empty mass, as given in [3], an Orion SM with 316 second Isp and 4115 kg empty mass (Oxygen, nitrogen and water are not needed) can do this job of transferring the ship from HLO to LLO (571 m/s [5]) with a propellant load of 7620 kg, which fits within the 7907 kg maximum prop load Orion can hold [6]. The tanker would still be a single vehicle though; Orion SM with a fuel tank for 15.4 tons of hypergolic fuel would (barely) exceed the 28.8 ton TLI capacity of two Delta IV rockets. It's still a possibility though, but with very little margin, only ~200 kg, which NASA likely wouldn't risk. 

For lunar missions, the Utility module would use its propulsion system to transfer the station from L2 to High Lunar Orbit, where it will remain for the rest of the lunar campaign. After this, the lunar lander is launched fully fueled, followed by the LTV (small enough for a single ACES Delta IV H) and a crewed Orion spacecraft. The crew would dock with the station, transfer to the lander, and would descend to the lunar surface from where they can do a 14 day science mission with 300 kg of scientific instruments. After the mission, they ascend to LLO, where they will dock with the LTV and return to the station. At the station, the crew enters their Orion spacecraft and heads home. For follow-on missions, a tanker vehicle and Orion are sent to the station, taking in total four launches per follow-on mission. 

It would be possible to launch Orion to the station in one go by adding 6x GEM boosters to the Delta IV booster. This would increase payload to about 45 tons [1]. Using Orion and a drop tank at the same time, it could refuel to about 25 tons of propellant. Orion could be short-fueled to about 17 tons for L2 missions, in which case ACES has sufficient delta V to get Orion to the station. A single Delta IV would have about 19 ton TLI capacity, enough for Orion or any other crew or cargo vehicle currently planned. This upgrade would save 1 launch per mission on the manifest, but would require additional upgrades to Delta IV which might end up costing more.

The advantages of a Gateway station

The architecture described by Boeing, and the modifications I made, have a few significant advantages over normal exploration. First, there's technical/economical advantages: the Gateway allows the lunar lander, by far the most expensive part of this architecture, to be fully reusable and used for many times. Unlike letting the thing float in LLO, it can easily be refueled and can be repaired at the station, should maintenance be required. It is also an efficient staging point for missions to asteroids and Mars; while those will likely require bigger launchers in the 80 ton class, the groundwork for such missions could be laid with near-term launchers. 

Another advantage is the international aspect of the plan. The costs are split over several participating countries. For example, the US provides the LV, the Node and Utility Module, the Orion crew module, and commercial resupply. Russia provides the Science and Power module and does most of the lunar lander. They also provide crew access and HLV capacity later on via Angara 7, PTK NP, and Sodruzhestvo (Zenit super Heavy). Europe contributes to the lander, the LTV and provides a LV for commercial resupply. The advantages of such an approach are obvious. America can provide their part within the current exploration budget, while the high cost of a lander of ~$7 billion [7] is split over two space agencies rather than one. 

All in all, an international Gateway station program allows for a flexible, low cost exploration program supporting a wide range of missions, and Delta IV with the ACES upper stage upgrade is very well suited for laying the groundwork for an exploration program and supporting the exploration of cislunar space. 


[5] Done using the Vis-Viva equation: 232.6+(1972-1633.9)=571 m/s Delta V

donderdag 7 november 2013

Low Cost Lunar Missions; To the moon with Ariane 6

Lunar exploration has always been a huge interest of mine. The moon is our closest neighbor and could teach us tremendous amounts about the history of our solar system and our planet. It could also function as a place to gather resources to explore further into the solar system. In short, there's plenty of reason to go there, but how? Many earlier plans to got there have yielded nothing but powerpoints and pretty animations. Getting there in a low-cost manner would be critical.

(For info on the plan discussed, scroll to the bottom.)

Constellation: How not to go there

The Constellation program was initiated by NASA as part of Bush's Vision for Space Exploration policy in 2005. It had the goal to return Americans to the moon by 2020 and give America independent manned access to space by 2014 after the Shuttle's retirement in 2010. However, the way they wished to accomplish this was doomed to fail from the beginning. 
The first fatal flaw of the program was the way two different LVs were used for achieving one goal. Using two different vehicles delayed Heavy Lift capability and costs a lot of additional money for developing the extra vehicle. Even if the smaller vehicle is a lot lower in cost than the big one, which Ares 1 certainly wasn't, you still wind up having to pay more. Using two or more launches of the same launch vehicle, like is the plan with SLS, significantly reduces the costs and allows the mission to take place much earlier, and doesn't require a vehicle as big as Ares V, which by itself was way too big and expensive.

The second fatal flaw in the program is that it's very ambitious. Landing four people on the moon for at least 7 days, anywhere on the surface, is a very big requirement, which ends you up with a spacecraft and lander which are both extremely huge and very expensive to develop. It also caused Ares V to be huge; up to 188 metric tons to LEO according to some sources. Even the biggest version of SLS won't go over 130 tons, and that is still years into the future. Ares V would have used a new, bigger core, new engines, new boosters and would have nothing in common with the space shuttle. It was a completely clean-sheet design and it would have cost over 20 billion dollars to develop. The whole constellation program would've cost over $40 billion dollars just for the first lunar flights. NASA doesn't have the funds for that.

Lastly, another big flaw in the program was Ares 1 itself. The vehicle was designed from the start as "it must use a space shuttle solid rocket booster as the first stage" and that is were the problems started. Even when using the powerful RS-25 engine for the upper stage, it was underpowered and provided almost no margin for Orion. The slightest grow of Orion or performance reduction of Ares would have made the LV useless.

Early Lunar Access

ELA lander docking with EDS. Credit: NASA, Wikimedia
By far one of the better ways to do "budget moon flights" is the Early Lunar Access architecture designed in the early 1990's. It was based on the basic "Faster Cheaper Better" spirit at NASA at the time. It requires only two launches of Medium Lift Vehicles; one space shuttle and one Titan IV or Ariane 5. The total mass in LEO is only 52 tons. It would use only near-term hardware and could be ready by the year 2000, and would bring two people to the surface of the moon. While less capable than constellation, the use of smaller expendable LVs allows more missions to be done for a lower cost, more than making up for the smaller capability. The idea is great, but slightly outdated. Many modernized versions of this architecture are possible, using newer launchers like Falcon Heavy, Delta IV, Ariane 5 or SLS. This time, however, I'll focus on an architecture using Ariane 6, Europe's new, low cost launcher.

Ariane 6

Ariane 6. Credit: ESA

Ariane 6 is the successor to Ariane 5. It is expected to become operational after 2021, and cost approximately €70 million a piece, which is about $93 million dollars. It can get 6.5 tons to GTO, but its LEO performance has not been released yet. The vehicle is designed for GTO and that's where its market is at, so LEO performance isn't as important. However, it is possible to estimate the performance of the vehicle.

Using the LV performance calculator, which should be added is just an estimation tool, I've been able to model the performance of 6.5 tons to GTO using the following numbers:

Stage 1, 2 and boosters: 135 tons of propellant, 11.8 tons empty mass, 4500 kN of thrust, 280 second specific impulse. 

Stage 3: 28 tons of propellant, 4.94 tons empty mass, 180 kN of thrust, 464 second specific impulse. 

All these values are based on Vega's P80 lower stage (which forms the basis for A6's P135 main stages) and Ariane 5's upper stage information (propellant mass fraction was improved, because A6 doesn't have the same volume restrictions as A5). For LEO, a mission to a 200x200 km orbit with an inclination of 6ยบ, it has a payload capacity of 17.3 tons. This seems a little optimistic to me, so I took out a chunk of payload by including a 12.5% performance margin for error, reducing payload to 15.1 tons to LEO. This is a lot more realistic for this vehicle, and it's the LEO payload I went with for the rest of this article. 15 tons is a lot less than the 23 tons Ariane 5 ME can get into orbit, but it does so at a much lower price; only $93 million instead of the estimated $210 million for Ariane 5. That's a 32% decrease in cost per kg (and pretty close to the claimed 30% cost decrease for A6). For this reason, Ariane 6 is the vehicle I went with. That, and it's European, and the European space program is a great interest of mine.

The mission 

Launch 1: The first Ariane 6 launches the Lunar Landing and Return Vehicle. This vehicle consists of a capsule (3.7 tons fully loaded), a propulsion module carrying an engine (Aestus 2, storable propellant, 55.4 kN and 340s Isp) and landing legs and a propellant load of approximately 9.3 tons.  Total mass of the system in LEO is 15 metric tons. The hypergolic Aestus 2 engine was chosen because it allows fuel to be stored in space for much longer. The Aestus 2 is by far the most efficient upper stage engine available for this purpose right now. It is launched without a crew: to prevent another Ares 1 fiasco, it's probably better not to launch humans on a solid powered rocket without any significant margin, especially not because it doesn't have a Launch Escape System yet. This part of the ship has 3226 m/s of ∆V.

Launches 2,3 and 4: Carry small propulsion modules. They each have a mass of 15 tons and a structural index of 10%, meaning that 10% of their mass is non-propellant. Thee of them give the stack, including the LLRV ∆V, a total of 8697 m/s, which is just enough for the total round trip from LEO back to Earth. 

Soyuz can provide crew transport to the LLRV
Launch 5: Should Ariane 6 not be capable of launching humans, which it probably won't be, a 5th launch would be necessary. This would be a Soyuz, launched from Kourou Space Center in French Guyana. It would bring a crew of 2 to the loitering spacecraft in LEO. The Soyuz facilities at GSC are made so that they can easily be adapted to human launches, so this shouldn't be a big problem. Landing locations after the flight are a different story, but they aren't supposed to reenter in Soyuz anyway.

Major mission events include a 3100 m/s burn to a trans-lunar trajectory, a 2700 m/s burn to land, surface exploration by the two astronauts, an 1872 m/s burn to enter LLO and a 1025 m/s burn to return to Earth. In order to save ∆V, the spacecraft sent on a trajectory straight to the moon and does not enter LLO before landing. This saves about ~300 m/s compared to entering LLO first (only 200 m/s was assumed here). However, in order to allow some more global access to the moon, the craft does enter LLO on the return trip

The launch costs of this mission are not insignificant. 4x Ariane 6 is a total of $372 million dollars, and a Soyuz adds another 40-60 million. In the worst case it's $432 million total. However, this is still a lot lower than the cost for two SLS launches (which I previously estimated at $1.4 billion a piece, for $2.8 billion total). It also helps kick up the flight rate of Ariane 6. In order to get a low price out of the vehicle, a low price is required. It is supposed to launch 7-15 times a year, and reach the 70 million euro goal at 7 launches a year. Costs can only go down by flying the vehicle more often, and four extra launches will certainly help with this. 

Sources and other interesting information: