Wednesday, December 26, 2012

SLS for Return to the Moon by the 50th Anniversary of Apollo 11, page 2: Orion + SEV design.

Copyright 2012 Robert Clark

 It is generally acknowledged that the SLS is based on the DIRECT teams "Jupiter" launcher. Then their respective launchers closely mirror each other in their payload capabilities for versions with similar components. The Block 0 SLS was initially planned to have a 70 mT payload capability, as mirrored by the corresponding DIRECT launcher:

 In reports on the Block 0 SLS, NASA discussed the option of it using 4 or 5 segment SRB's as if it were no big deal. But I was surprised when I looked at the 5 segment version on the DIRECT teams site, that the payload jumped to ca. 95 metric tons:

  Ed Kyle who operates the site also estimates this first SLS version will have a payload to LEO of 95 mT. A jump in payload of 25,000 kg is a big deal. It's the difference in payload for instance between the 105 metric ton Block 1A version, and the 130 metric ton Block 2 version of the SLS. It would also mean the Block 0 given 5-segment SRB's would be close to the "magic" 100 metric ton payload number. And with just the interim upper stage, it would certainly exceed that.

 Judging by this Chris Bergin article, we would expect the 5 segment SRB's to be ready by the 2017 first flight of the SLS:

ATK and NASA ground test their SLS-bound five segment motor.

September 8th, 2011 by Chris Bergin
    As far as ATK’s role in SLS, documentation (L2) shows the Utah-
based company have proposed a Firm Fixed Price (FFP) contract for 10
boosters, available between 2012-2015, whilst noting available assets
that can support up to 11 SLS missions prior to asset depletion in

The current plan now is to go directly to a Block 1 launcher, scheduled for a 2017 flight date. This will use 5-segment SRB's instead of the regular 4-segment ones planned for the Block 0. But the DIRECT teams 5-segment version of their Jupiter rocket has nearly a 95 mT capability. Moreover, NASA wants to give the Block 1 an additional SSME core engine and stretch the tank. Then it will have even greater payload than the 95 mT of the corresponding DIRECT teams launcher.

So NASA is still using the 70 mT payload number of the Block 0 in discussing this initial flight of the SLS when the actual payload capability will be 95+ mT. I think NASA should be more clear about what the actual capabilities of that first version of the SLS to fly will actually be. Saying it will do 70 metric tons to LEO is misleading as to what that first version can actually do.

According to the reports that first version to fly will even have an interim cryogenic upper stage, and at quite low cost by the reports if the Delta IV derived one is used. Presumably, this will improve the LEO capability, perhaps to the 100 to 105 metric ton range.

A launch capability this high raises the possibility of even doing lander missions not just lunar flyby's. This is important because it means we will have the capability of doing lunar lander missions not just in 2030 when the full SLS comes on line but just in 5 years.

This becomes even more important when you realize the necessary stages, the Centaurs, already exist to make the Earth departure/lander stages. ULA has written numerous reports on markedly reducing boiloff in the Centaurs so that we can consider that to be well understood, and essentially solved.

It has been complained that the SLS has no mission. NASA being direct, so to speak, about what the actual capabilities of that first version of the SLS to fly will make clear that the SLS does have an important mission, and in the very near term and at (comparatively) low cost: Return to the Moon.


A Simple, Low Cost Upgrade.

 A question asked about the SLS is that if the Block 0 is derived from the space shuttle system that could lift 100+ mT to orbit when you include both the orbiter and payload, then why could the Block 0 only lift 70 mT to orbit? The answer is that for the shuttle the SSME engines only took the orbiter to a highly elliptical orbit whose perigee lied well within the Earth's atmosphere. This ensured the external tank after being jettisoned would reenter the atmosphere and break up on return.

 The shuttle would then use one or two OMS burns to raise the perigee and circularize the orbit. These OMS burns typically only totaled 90 m/s or less. Note that the total thrust of these OMS engines for the 100 mT+ shuttle was only about 6,000 kgf. This thrust is less than that of a single RL-10 engine. Then a way to recover the full mass to orbit of that of the shuttle system is by using a small propulsive stage to provide the same low amount of extra delta-v as provided by the shuttle's OMS engines.

 The shuttle orbiter with payload and with OMS fully fueled can mass 120 mT. An OMS burn of 90 m/s is less than 1/3rd the total OMS delta-v available of 305 m/s. So much of the OMS propellant of 12.8 mT will remain, with the remaining gross mass of the orbiter at the end of the OMS burn being above 100 mT.

 This delta-v change for a 100 mT payload can be done by just a cryogenic stage at only 1/10th the size of a Centaur upper stage, one of only 2 mT size. The Centaur has better than 10 to 1 mass ratio. But mass ratio gets better as you scale up or said another way gets worse as you scale down.

 The 'Golden Spike' paper on a commercial return to the Moon plan gives estimated sizes for some smaller cryogenic stages than the Centaur in a table on page 13. One at a 2,172 kg propellant load is given a dry mass of 445 kg. This could provide a 90 m/s delta-v to a 105 mT payload with a RL-10 engine at 451 s Isp:

451*9.81ln(1 + 2.172/(.445 + 105) = 90 m/s.

 Note this is just for Block 0. But the actual first version to be launched will be the Block 1 with 25% greater size and thrust on the SRB's and 33% greater size and thrust on the core stage. Then also using a small cryogenic stage the payload would be at least 25% greater than the 105 mT amount and probably closer to 30% greater since the upper stage that actually reaches orbit has a greater influence on payload than a lower stage.

 Even 25% greater would put the payload at 130 mT. This matches the payload of the expensive Block 2 SLS but only requiring a small cryogenic stage a fraction of the size  of a Centaur, and would be available by the 2017 first launch of the SLS.

Return to the Moon Architecture.

 In the post "SLS for Return to the Moon by the 50th Anniversary of Apollo 11" I suggested the Space Exploration Vehicle(SEV) be used alone as the single crew module for a lunar mission following the Early Lunar Access architecture. However, the Orion capsule has had billions of dollars spent on it and therefore has a lot of political capital attached to it. So I'll show we can also have a design that uses the Orion for the traverse from Earth orbit to lunar orbit and the return, with the SEV just for the trip from lunar orbit to the lunar surface. Using all cryogenic propulsion this will be doable using the likely 95 mT or higher payload first version of the SLS scheduled to launch in 2017. Using both the Orion and the SEV is in the plan NASA is considering for asteroid missions. I'm suggesting it also be used for lunar missions to get a lightweight architecture, rather than using some analogue of the quite heavy Altair lander (45 metric tons, really??).

 Use the delta-v's for the Earth-Moon system shown here:

Delta-V budget.
Earth–Moon space.

Currently existing cryogenic stages for simplicity and low cost: for the SEV lander use the Ariane H8 LH2/LOX upper stage. It had a 9,687 kg gross mass and 1,457 dry mass, and 443 s Isp. I'll round off the H8 mass values to 9,700 kg and 1,500 kg in the calculation. Use 4 mT for the crewed mass of the SEV, then:

 443*9.81ln(1 + 8.2/(1.5 + 4)) = 3,970 m/s, sufficient for the flight to and from the lunar surface from low lunar orbit.

 For a stage to insert the Orion+SEV lander into lunar orbit and return the Orion to Earth from lunar orbit, use the Ariane H10-3 LH2/LOX upper stage. This stage has a gross mass of 12,310 kg and dry mass of 1,570 kg, at a 445 s Isp. I'll round off the mass values to 12,300 kg and 1,600 kg, so 10,700 kg of propellant.

 The delta-v to insert into lunar orbit is 900 m/s, and the translunar injection(TLI) delta-v is 3,140 m/s making up the 4,040 m/s delta-v to go from LEO to low lunar orbit(LLO), as shown in the table above.

 Use 9 mT for the crewed mass of the Orion, and 13.7 mT for the SEV plus lander. Now burn only 6.9 mT of propellant for the lunar insertion, retaining 3.8 mT of the propellant after the lunar orbit insertion in order to be able to return Orion back to Earth. Then:

445*9.81ln(1 + 6.9/(1.6 + 9 + 13.7 + 3.8)) = 960 m/s, sufficient for lunar orbit insertion.

 Now for the return of the Orion, we have:

445*9.81ln(1 + 3.8/(1.6 + 9)) = 1,340 m/s, sufficient to go from low lunar orbit back to LEO, according to the table above. (Actually other sources give the required delta-v to break lunar orbit as only 900 m/s, same as to enter orbit, so it may be possible to make this stage even smaller.)

  Now we need a stage to do the translunar injection(TLI), requiring 3,140 m/s delta-v. The Centaurs have the best Isp and mass ratio of any upper stages so we'll use those. You could use two of them firing together in parallel or get better mass to TLI by firing them serially.  For simplicity I'll use the twin, parallel Centaur format. Rounding off, the Centaur has 21 mT propellant and 2 mT dry mass, with 451 s Isp. So two together would be 42 mT propellant and 4 mT dry mass. The Orion, SEV, and cryogenic stages together mass 35 mT. Then:

451*9.81ln(1 + 42/(4 + 35)) = 3,230 m/s, sufficient for TLI.

 Then the total mass that needed to be lofted to orbit would be 81 mT. The leeway between this and the 95 mT, and likely higher, payload capacity of the SLS would probably allow even hypergolics to be used at least for the departure stages, both from the lunar surface and from lunar orbit.

Increasing Mass Ratio to Improve Performance.  

 An even better option than the twin Centaurs would be to use the proposals of ULA (United Launch Alliance) to scale the Centaurs up larger, widen their diameters, and use lightweight aluminum-lithium instead of the steel now used. ULA suggests by doing this their mass ratio can be increased from 10 to 1 to 20 to 1. This is discussed by Jon Goff on his site, Selenian Boondocks.

 Scaling a rocket stage up is known to increase mass ratio. Widening them improves mass ratio because the closer a tank is to sphere the better the storage efficiency, a sphere having the best mass efficiency. And in regards to strength compared to weight, Al-Li can be as much as twice as good as steel. 

 These weight saving methods should also be applied to the smaller cryogenic stages to improve their performance. For instance the Ariane cryogenic stages I used above may be able to reach 10 to 1 mass ratios by following this. ULA has discussed improving mass ratio as the best way to improve performance at the NewSpace 2012 conference.

 A Centaur-style stage with these weight saving techniques applied at a 40 mT propellant load and 2 mT dry mass using the best vacuum Isp for a RL-10 series engine at 465.5 s can transport 5 mT from LEO to the Moon and back as a single stage:

465.5*9.81ln(1 + 40/(2 + 5)) = 8,700 m/s, sufficient for the round-trip according to the table above.

  Actually since the delta-v of a launch to LEO is just a little more than this delta-v for a round-trip lunar mission, I like to think of this example as a stealth SSTO. ULA in maximizing the mass ratio of a Centaur-style stage while at the same time using the highest Isp engine would unwittingly also create a SSTO, capable of significant payload to orbit.

 For this SSTO to have an engine that can operate at sea level, the nozzle extension would have to be retracted at launch and extend while the engine is firing. According to Henry Spencer, this has already been successfully tested.

RL-10-B2 with nozzle retracted.

2001: A Space Odyssey.

 Another version of this high mass ratio upper stage would put it in the from of a sphere. Since a sphere has the best mass efficiency for a tank this would get an even better mass ratio, and could carry more payload. This would be most useful for the lunar transport case since you would not have to worry about the high air drag of a spherical launcher as in the ground launched case.

Aries Lunar Shuttle.
 This would be interesting since it could serve as an homage to 2001: A Space Odyssey.

Bob Clark

Update, Sept. 28, 2013:

 Finally, NASA has acknowledged that the Block 1, first version of the SLS to launch in 2017 will have a 90+ mT payload capacity not the 70 mT always stated by NASA:

SLS Dual Use Upper Stage (DUUS).

 This is important since it means we will have the capability to do manned lunar landing missions by the 2017 first launch of the SLS:

SLS for Return to the Moon by the 50th Anniversary of Apollo 11, page 5: A 90+ metric ton first launch of the SLS.

Wednesday, December 19, 2012

"Golden Spike" circumlunar flights.

Copyright 2012 Robert Clark

"Golden Spike Co" has released a paper describing their return to the Moon plans:

An Architecture for Lunar Return Using Existing Assets.
by James R. French et. al.

 It gives several different architectures and types of missions. But on page 8 it gives the payload capability of the Falcon 9, presumably the new version Falcon 9 v1.1, as 16,700 kg. However, on the SpaceX site it's given as 13,100 kg:

 Interestingly at the 16.7 mT number you can do a manned circumlunar mission on a single Falcon 9 + Dragon, even including a launch abort system(LAS), by using a half-size Centaur as the in-space stage. But at the 13.1 mt number it becomes more problematical .
 Such a mission would be very important to accomplish. Recall the Apollo 8 mission was a manned lunar flyby that served as the prelude to the Apollo 11 mission. It is regarded then as being a part of the costly Apollo program, requiring the expensive Saturn V launcher.
 The skepticism among many about the Golden Spike plan or other commercial lunar plans is the idea it would require large, highly expensive Saturn V class launchers. However, if the manned flyby could be done by a single launch by what is still just a medium size launcher in the Falcon 9 v1.1 it would show that by going small and following a low cost, commercial approach, that a low cost return to the Moon is feasible.
 The Falcon 9 v1.1 will cost in the $60 million range, and we might estimate the half-size Centaur to be in the $15  million range. So the launch cost for such a mission might be in the $75 million range.
 As I discussed before in regards to using the first test flights of the Falcon Heavy for unmanned BEO test flights, the test flights of the Falcon 9 v1.1 could serve for unmanned test flights for this lunar flyby. Since SpaceX needs to do such tests anyway most of the cost of the Falcon 9 and Dragon capsule would be borne by SpaceX. Then you could have Golden Spike only paying ca. $15 million for the half-size Centaur.
 There would be some development cost of course beyond that for this half-size Centaur. For one thing you would have to make the cryogenic propulsion undergo less boiloff for 1 to 2 week missions. ULA has done studies on this so should be doable but still it has to be carried out in practice. An advantage of this would be that this half-size Centaur is about the size you need for the lander. So the lander could be derived from this, and the development cost for the two stages could be reduced.

 The Golden Spike landing plan specifies using two Falcon Heavy's even though it uses a Dragon sized capsule. This is more than 100 mT to LEO. This is puzzling since the advantage of using a lightweight capsule is that it should require smaller amounts to be launched to orbit, known as IMLEO, initial mass to low Earth orbit. For instance the Early Lunar Access plan only requires 52 mT to orbit using a small two-man capsule. However, I believe the Golden Spike paper by French et. al. explains where the discrepancy arises.
 On page 13 is given a table of some masses for different possible propulsive stages. The mass for the Dragon with trunk and crew and supplies is 8,853 kg, well above the given dry mass of the Dragon capsule at 4,200 kg. The trunk section is less than 1,000 kg and the propellant for the Dragon is at 1,290 mT. The mass for crew and supplies in the Golden Spike paper is given as 300 kg. Evidently then the extra mass to get to a 8,853 kg mass is coming from the launch abort system (LAS).
 In any case a 8,853 kg mass would be at the mass of the Orion capsule and we would lose any advantage of a lightweight architecture. Then I suggest an alternative to the SpaceX LAS that has the LAS permanently integrated into the capsule.
 We could use again a tower type LAS that would be jettisoned prior to reaching orbit. To estimate its mass we might make a comparison to the Orion LAS. The Orion LAS is at 6,000 kg. The Dragon is at half the mass of the Orion capsule. Then we can estimate the mass for a tower type LAS for the Dragon as 3,000 kg.
 This is also a high additional mass. However, typically a tower LAS is jettisoned soon after first stage separation. So we can estimate how much this will subtract from the payload to orbit by making a comparison to how much the payload is reduced for an increased dry mass to the first stage. A rule of thumb is that every kilo of mass added to the first stage dry mass subtracts off 1/10 of a kilo from the payload. So we can estimate 300 kg being subtracted off the payload.
 Now we'll estimate the size of a cryogenic stage needed to take the Dragon to a circumlunar mission. In the table on page 13 in the Golden Spike paper is given a cryogenic, LH2/LOX, stage at 1,196 kg dry mass and 7,534 kg propellant mass. This is a mass ratio of 7.3 to 1. Notably this is less than that of current Centaur upper stages at about 10 to 1. This is because mass ratio improves as you scale up your rocket stages.
 This rocket stage would be sufficient to carry the Dragon's 4,200 kg dry mass plus the 300 kg for crew and supplies using RL-10 engines. The delta-v for trans lunar injection(TLI) is 3,150 m/s. Using a 451 s Isp for the RL-10 engines we get a delta-v of:

451*9.81ln(1 + 7534/(1196 + 4500)) = 3,700 m/s.

 But because of the loss of payload capacity due to the LAS from  SpaceX's cited payload to LEO of the Falcon 9 v1.1 of 13.1 mT, this would be slightly more mass than can be carried to LEO. So we'll use a slightly smaller stage. Let the propellant mass be 7,000 kg. Keeping the same 7.3 mass ratio, this corresponds to a dry mass of 1,100. Then the delta-v will be:

451*9.81ln(1 + 7000/(1100+ 4500)) = 3,600  m/s, still sufficient for the TLI.

  Bob Clark

Saturday, December 1, 2012

SpaceX Dragon spacecraft for low cost trips to the Moon, page 3: Falcon Heavy for BEO test flights.

Copyright 2012 Robert Clark

The Falcon Heavy is planned to be tested by SpaceX by 2014. By using the Early Lunar Access (ELA) architecture we could have a return to the Moon by 2019, using either the Falcon Heavy or the SLS, as described in the blog post "SLS for Return to the Moon by the 50th Anniversary of Apollo 11". Note this would also be by the 2020 timetable set by the Vision for Space Exploration(VSE).
 This ELA architecture could be implemented either using either the Dragon capsule or the NASA Space Exploration Vehicle (SEV). The SEV is intended to be used for BEO missions perhaps for mission durations up to 28 days long with two crew members. Then the first Falcon Heavy missions would provide means for testing unmanned the SEV for BEO missions to the lunar surface, the Lagrange points or to near Earth asteroids.
 SpaceX needs to perform the test flights for the Falcon Heavy so they would pay for the costs of the launch themselves. For the lunar flights as discussed in the "SpaceX Dragon spacecraft for low cost trips to the Moon" post, the two Centaur-like upper stages might require in the range of $30 million dollars each. There would be an extra cost for the SEV but for these first test missions we might only use the prototype test vehicles now undergoing field tests with NASA's Desert RATS program. These prototype vehicles only cost in the few hundred thousand dollar range. This would put the cost to NASA in the low cost Discovery-class mission range.

2011 Desert RATS Overview
 Since these are to be unmanned test flights, ideal would be to "man" them with two Robonauts. This would dovetail nicely with the Johnson Space Center's Project Morpheus plan to send Robonaut to the lunar surface.  In regards to NEO missions, this is one of the planned uses for the SEV. For the manned asteroid flights NASA was considering, the mission time ranges were above 90 days. That would be unrealistic for flights just using the SEV alone. However looking at the Near-Earth Object Human Space Flight Accessible Targets Study (NHATS) page there are NEO missions with duration times of 34 days or less at stay times of greater than 8 days that could correspond to under 28 day mission times if we limit the stay time to a day or so.

      Bob Clark

Monday, October 29, 2012

SLS for Return to the Moon by the 50th Anniversary of Apollo 11.

Copyright 2012 Robert Clark

Very interesting report about using NASA's proposed Space Exploration Vehicle for cislunar space exploration:

Lunar Surface Access from Earth-Moon L2.
A novel lander design and study of alternative solutions.
1 October 2012 | Washington, DC

 The report proposes using the lightweight SEV, at only a 3 mT empty weight, and all cryogenic propulsion as a shuttle between the L2 space station NASA has recently discussed and the lunar surface. However it could also be used as the crew capsule between LEO and the Moon's surface.
 The architecture discussed is very interesting in that the SEV would be used as the single crew module to carry the crew all the way from the L2 station to the lunar surface and back again, i.e., no separate lander crew module. There would also only be a single propulsive stage to carry the SEV from low lunar orbit to the lunar surface and back to lunar orbit, i.e., no separate lunar descent and ascent stages.
 This has similarities to the architecture for the Early Lunar Access(ELA)[1] proposal of the early 90's. This also used all cryogenic space stages to save weight, only 52 mT required to LEO. ELA also saved weight and cost by using a single crew capsule for the entire flight from LEO to the lunar surface and back again. It also used a single propulsive stage for lunar descent and ascent. But instead of linking up with a stage waiting in lunar orbit for the return, the ELA proposal was to have this single lander stage return all the way back to LEO.
 An alternative architecture discussed on page 23 in this report on using the SEV for cislunar travel does not use the method of first stopping in lunar orbit, then having a separate lunar lander stage. Instead it uses the "direct descent" method of descending directly to the lunar surface. This landing method is analogous to that used in the ELA proposal to save propellant. Interestingly the SEV report on page 23 gives the delta-V for the direct descent method as 2,610 m/s. This compares to the 760 m/s + 2,150 m/s = 2,950 m/s for the method that first stops in lunar orbit, then descends to the surface as indicated in the image above. So according to this report a savings of 300 m/s in delta-V for the trip from L2 to the Moon is possible using direct descent, a significant savings.
 I had wondered if it was possible to save delta-V and propellant in this blog post 'Delta-V for "direct descent" to the lunar surface?'[2]. The SEV report suggests it may be possible to save in the range of 300 m/s by the direct descent method.
 The only technical complaint raised against the feasibility of the ELA proposal back in the 90's was the suggestion of getting a 2-man crew capsule at only a 3 mT empty weight. So the fact the SEV is expected to have this low an empty weight is important, since it suggests the possibility with just the 70 mT first version of the SLS of a manned lunar lander mission using currently existing cryogenic stages.
 Actually the 70 mT payload of the SLS is so much better than the 52 mT needed for ELA that likely we could even use a heavier hypergolic stage for the lunar ascent stage. During the early planning of the Apollo program when the possibility an engine might not ignite was regarded as a definite possibility, it was decided to use hypergolics, which ignite on contact, for the lunar ascent stage. At this point though the cryogenic RL10 engines have had decades of use and are regarded as highly reliable.
 Still for these first versions of these new lunar landers we might still want the certainty of using hypergolics for the ascent stage. I suggest using the engine and propellant tanks of the shuttle orbiter OMS pods for the purpose. This would be quite appropriate actually since the OMS pod engines were derived from the Apollo lunar lander engines. By the Astronautix page on the OMS pods[3], they are each about 10 mT propellant mass and 1.8 mT dry mass. Then using its 316s Isp, one of them would suffice for the ca. 2,740 m/s delta-V to go from lunar surface to LEO even with a 4 mT crewed and supplied mass for the SEV with plenty of margin: 316*9.81ln(1 + 10/(1.8 + 4)) =  3,100 m/s.
 The first version of the SLS, called Block 1, is expected to launch by 2017. I would expect a test lunar lander mission, especially if using all cryogenic in-space propulsion, to be done first before a crewed mission is sent. But certainly by 2019, the 50th anniversary of Apollo 11, a crewed mission could be sent. This is in contrast to a post-2030 proposed time frame for a crewed lunar landing using the full 130 mT version of the SLS when it first becomes available.
 There is the cost issue of mounting a manned lander mission. Oddly, the high cost of the SLS might be helpful in this regard. The cryogenic Centaur-like upper stages are already available at a cost in the range of $30 million [4], so the modifications there would be comparatively low cost, compared to the already high cost of the SLS. As for the development cost of the SEV, I suggest use of NASA's commercial crew program's financing procedures. SpaceX was able to develop the Dragon as largely privately financed for reportedly $300 million. And Boeing is paying much of the cost of the development of the CST-100 capsule. It is highly dubious they would be spending a billion dollars of their own money for its development. Then likely its total development cost is in the few hundred million dollar range. Therefore it is likely the development cost of the smaller SEV under commercial crew procedures would also be in the few hundred million dollars range, again comparatively low cost compared to the SLS.
 As I discussed in the blog post "SpaceX Dragon spacecraft for low cost trips to the Moon", SpaceX will also be able to mount a manned lunar landing mission using the 53 mT Falcon Heavy by following, it turns out, the ELA architecture. This will be much cheaper than using the SLS launcher, perhaps only in the few hundred million dollars range cost. But you would have to get private financing for that, since NASA would not fund it as it would undercut NASA's own program.
 In contrast, NASA using the SLS in such an early time frame for a manned return to the Moon would provide further support for continuing the SLS funding. No longer would the SLS be referred to as "a rocket to nowhere".

  Bob Clark

Update, Sept. 28, 2013:

 Finally, NASA has acknowledged that the Block 1, first version of the SLS to launch in 2017 will have a 90+ mT payload capacity not the 70 mT always stated by NASA:

SLS Dual Use Upper Stage (DUUS).

 This is important since it means we will have the capability to do manned lunar landing missions by the 2017 first launch of the SLS:

SLS for Return to the Moon by the 50th Anniversary of Apollo 11, page 5: A 90+ metric ton first launch of the SLS.


1.)Lunar Base Studies in the 1990s. 
1993:  Early Lunar Access (ELA). 
by Marcus Lindroos 
(Note a typo on this page: the payload adapter mass should 
be 2,000 kg instead of 6,000 kg.) 

2.)Delta-V for "direct descent" to the lunar surface?

3.)Encyclopedia Astronautica.
Shuttle Orbiter OMS.

4.)Encyclopedia Astronautica.
Centaur IIA.

Wednesday, October 24, 2012

SpaceX Dragon spacecraft for low cost trips to the Moon, page 2: Comparison to 'Early Lunar Access'.

 Copyright 2012 Robert Clark
 Early Lunar Access lander stage.

 The Early Lunar Access [1],[2], proposal of General Dynamics came as quite a surprise to those in the industry when it was first proposed in the early 90's. It suggested manned lunar missions at half the mass needed to LEO and at 1/10th the cost of the Apollo missions.
 It was based on using existing launchers with small upgrades to keep costs low. The only part of it that was technically doubtful at the time was that you could get the lightweight 2-man capsule they were proposing at only a ca. 3.7 mT crewed mass.
 Based on such a small sized capsule, they were able to get a manned mission to the Moon at only about 52 mT required to LEO using all cryogenic space stages. However, the 7-man Dragon capsule at a ca. 4mT dry mass suggests this is indeed feasible.
 It is also interesting the architecture they were proposing for low costs was similar to what I suggested for the SpaceX Dragon via the Falcon Heavy launcher. It would use a single capsule to take the crew all the way from LEO to the Moon's surface and back again, i.e.,no separate lunar crew module. Also it would use as I suggested a single lander stage to take the crew capsule from low lunar orbit to the Moon's surface and then all the way back to LEO, rather than linking up with a return stage waiting in lunar orbit for the return.
 This gives further confidence in the feasibility of the lunar lander plan using the Dragon with Centaur-style stages launched on the 53 mT Falcon Heavy.

  Bob Clark

1.)Encyclopedia Astronautica
Early Lunar Access.

2.)Lunar Base Studies in the 1990s.
1993:  Early Lunar Access (ELA).
by Marcus Lindroos
(a typo on this page: the payload adapter mass should
be 2,000 kg instead of 6,000.)

Sunday, October 14, 2012

Re: On the lasting importance of the SpaceX accomplishment.

Copyright 2012 Robert Clark 

Congratulations to SpaceX on their second successful flight to the ISS. However, it is disturbing that there have been engine anomalies on all the flights, the last being the most serious:

 It is reassuring that the mission was able to be completed even with one engine shut down. However, I don't think that would be an acceptable state of affairs for manned flights to have an expectation that during any flight at least one engine would malfunction and need to be shut down, including to the extent that that engine would be destroyed, shedding debris in the process.
 I think SpaceX should investigate the possibility of producing a larger version of the Merlin to reduce the number of engines required. It's been reported also that NASA is not too sanguine on the possibility of using so many engines on a manned vehicle.
 There was a discussion of this possibility on the forum:

Should SpaceX aim for a 330,000 lbs engine rather than am F1 class engine?

 The idea was generally disparaged on that forum, but I think it is a good idea. SpaceX was considering building a 1.5+ million pound thrust engine referred to as the Merlin 2 as part of a proposal to NASA for a heavy lift vehicle. They estimated a $1 billion development cost for the engine. Based on thrust size, we might estimate the development cost for this smaller upgrade at 1/5th of this so only $200 million. Given the billion dollar contracts SpaceX already has for NASA and commercial satellite launches, there is little doubt that SpaceX could again get private financing for the development of this engine.
 SpaceX has shown that it is able to cut development costs when it follows a private financing path. I think that would be the ideal approach to follow in this case as well.
 If they did develop the 330 klb. engine, that would still require 5 engines for the Falcon 9 v1.1 first stage. My preferred solution then to minimize the number of engines at an affordable cost would be to go for a 500,000 pound thrust engine. Again estimating based on thrust size, this would be a ca. $300 million development cost, not too much more than the 330 klb case. But in this case you would only need three engines.

   Bob Clark

Saturday, September 15, 2012

Delta-V for "direct descent" to the lunar surface?

Copyright 2012 Robert Clark

 I was trying to get a lower roundtrip delta-V for lunar missions by flying directly to the lunar surface rather than going first into lunar orbit then descending, the "direct descent" mode. Here's a list of delta-V's of the Earth/Moon system:

Delta-V budget.
Earth–Moon space.

If you add up the delta-V's from LEO to LLO, 4,040 m/s, then to the lunar surface, 1,870 m/s, then back to LEO, 2,740 m/s, you get 8,650 m/s, with aerobraking on the return.
I wanted to reduce the 4,040 m/s + 1,870 m/s = 5,910 m/s for the trip to the Moon. The idea was to do a trans lunar injection at 3,150 m/s towards the Moon then cancel out the speed the vehicle picks up by the Moons gravity. This would be the escape velocity for the Moon at 2,400 m/s. Then the total would be 5,550 m/s. This is a saving of 360 m/s. This brings the roundtrip delta-V down to 8,290 m/s.
I had a question though if the relative velocity of the Moon around the Earth might add to this amount. But the book The Rocket Company, a fictional account of the private development of a reusable launch vehicle written by actual rocket engineers, gives the same amount for the "direct descent" delta-V to the Moon 18,200 feet/sec, 5,550 m/s:

The Rocket Company.

Another approach would be to find the Hohmann transfer burn to take it from LEO to the distance of the Moon's orbit but don't add on the burn to circularize the orbit. Then add on the value of the Moon's escape velocity. I'm looking at that now.

Here's another clue. This NASA report from 1970 gives the delta-V for direct descent but it gives it dependent on the specific orbital energy, called the vis viva energy, of the craft when it begins the descent burn:


The problem is I couldn't connect the specific orbital energy it was citing to a delta-V you would apply at LEO to get to that point. Any suggestions on how to accomplish that are appreciated.

  Bob Clark

Sunday, August 26, 2012

The Coming SSTO's: Applications to interplanetary flight.

Copyright 2012 Robert Clark

Credit: NASA image of an Orbital Transfer Vehicle with aerobrake. From David S.F. Portree's page: Shuttle-Era Manned Mars Flyby (1985).

 Note also a key fact about SSTO's is that the delta-V requirement for
a round-trip mission from LEO to the lunar surface is a little less
than that for flights from Earth's surface to LEO. Then if you could
do orbital refueling, you could have a single, reusable vehicle that
does lunar missions. This important capability about SSTO's is
mentioned in G. Harry Stine's very nice book Halfway to Anywhere:
Achieving America's Destiny in Space

" SSTO that is refueled in orbit has the capability to fly to the
Moon, land, lift off, and fly back without additional refueling."
Halfway to Anywhere: Achieving America's Destiny in Space, p. 220.

A table that gives the delta-V budget for trips in the Earth-Moon
system is given here:

Delta-V budget.
Earth–Moon space.

  From this you can calculate that the delta-V for a round trip from
LEO to the lunar surface is less than that for getting to LEO.
It has been argued that SSTO's are not economical. But that such a
vehicle with orbital refueling could also be used for lunar missions
changes the economic equations significantly.
 Surprisingly such SSTO's could also be used for Mars missions.
Elon Musk has argued in favor of promoting creating a self-sustaining
colony on Mars:
Elon Musk "Mars Pioneer Award" Acceptance Speech - 15th Annual
International Mars Society Convention.

 For such a colony he proposes reusable vehicles and getting propellant for
return trips from Mars. Musk proposes cutting the costs to space by two
orders of magnitude by reusability. Then there would be also a dramatic drop
in the cost to lift the large amount of propellant to space.
 So let's suppose there are propellant depots at LEO. Since Musk proposes a
self-sustaining colony on Mars, lets also suppose propellant depots in low
Mars orbit for return trips.
Here's a map of delta-v's between Mars/Moon/Earth:

 If you add up the delta-v's from low Earth orbit to low Mars orbit you get
6.1 km/s. Now use the same specifications for the Falcon 9 v1.1 first stage
as estimated before, 13 mT dry mass and 375 mT propellant load. Then
you could transport 45 mT from LEO to low Mars orbit:

311*9.81ln(1 + 375/(13 + 45)) = 6,130 m/s.

   Bob Clark

Friday, August 17, 2012

A liquid water component to clouds and fogs on Mars.

Copyright 2012 Robert Clark

Curiosity Surveys a Martian 'Mojave Desert': Big Pic.
Aug. 8, 2012 --

The panoramic image shows what appears to be "haze" at the base of the mountains in the distance in Gale crater. This was predicted prior to landing:

Pink skies, water ice haze in forecast for Curiosity landing.
12:56 PM, Aug 5, 2012
"PASADENA, CALIF. — Expect pink skies with a chance of a water ice haze over Gale Crater Monday when NASA’s Mars Science Laboratory and Curiosity rover arrive at the red planet.
"Seasonal winter temperatures are expected to be a balmy 10 degrees Fahrenheit when Curiosity touches down at 3 p.m. local Mars time."

 It is important to realize that clouds, fogs and hazes can have some proportion of liquid water even well below freezing temperature. This is well known to happen when salts are dissolved in the water through freezing point depression. But it can also happen with pure water through supercooling.
 The temperature at which supercooled liquid water can occur can even be below -40C, which,  coincidentally is also -40F:

Supercool Water.
Posted: 11/28/11
"Liquid water as cold as minus 40 F has been found in clouds. Scientists have done experiments showing liquid water can exist at least down to minus 42 F."


 Noctis Labyrinthus, part of the Valles Marineris system, frequently shows dense low lying clouds/fogs that give the appearance of precipitation carrying clouds on Earth:

Clouds in Noctis Labyrinthis.
Credit: NASA, Viking orbiter image.
This image shows early morning fog in the Noctis Labyrinthis, at the westernmost end of Valles Marineris. This fog, which is probably composed of water ice, is confined primarily to the low-lying troughs, but occasionally extends over the adjacent plateau. The region shown is about 300 kilometers (186 miles) across.

Noctis Labyrinthus, labyrinth of the night.
Mars Express
European Space Agency
30 November 2007

 Here's another great image showing dense clouds/fogs in Valles Marineris somewhat further west of Noctis:

taken from this ESA report:

Adsorption water driven processes on Mars.
D. Möhlmann
21-25 February 2005, ESA/ESTEC

The author reaches these conclusions:

Adsorption water in the upper martian surface is an actual challenge
to martian surface chemistry and possibly also to exobiology:
* Adsorption water makes possible and/or supports a martian surface
chemistry, also at present: These processes are energetically driven
by photons (UV). Current martian surface chemistry is mainly (non-
thermal) photo-chemistry.
* Existing iron oxides (as haematite), UV and adsorption water are a
cause for the production of oxidizing OH-radicals, which are expected
to contribute to the oxidation of organics (Methane, carbonaceous
* Adsorption water mobilizes acids (as sulfuric acid), which can
modify earlier formed carbonates (surface cover by sulfates, e.g.).
* Adsorption water covered catalytic surfaces of minerals are expected
to be essential agents in non-thermal photo-chemical processes. Photon
driven non-thermal redox-processes on catalytic surfaces might
together with atmospheric CO2 cause a non-biogenic production of
organics (?). Related experiments are in preparation.
* Adsorption water deposits also on the surfaces (cell walls) of
microbes etc. There, it can be a source of water for the microbial
metabolism. Physico-chemical processes can be supported by adsorption
water. To study the relevance of adsorption water for life-processes
is a current challenge to exobiology. Related experiments are in

 This Mars Express image of Valles Marineris with the dense fogs was taken May 25, 2004 in mid southern Autumn on Mars at a time approaching Mars aphelion.
 Equatorial clouds are known to be seasonal on Mars, frequently occurring near aphelion. It is now nearing the end of southern Winter on Mars, at the time of the Curiosity landing. It would be interesting to find out if higher resolution imaging by Mars Reconnaissance Orbiter also detects these dense low lying clouds/fogs during the next southern Autumn on Mars.
 Some MRO images near the location of the image with the dense clouds/fogs:

HiRISE | Latitude/Longitude Search Results.
Search by latitude and longitude range.
Latitude from: -25 to -5
Longitude from: 290 to 310 (Note: this is measured in east longitude.)


This report suggests clouds may be harder to form on Mars than thought previously:

NASA Study Reveals Less Water in Mars' Clouds.
Dec. 6, 2007

MOFFETT FIELD, Calif. – Martian clouds may contain less water than previously thought, according to a new NASA study.
New NASA laboratory measurements of simulated martian clouds reveal that scientists may have been overestimating the amount of water in the planet's atmosphere.
"The martian clouds we are studying are composed of water ice, like some clouds on Earth. However, they are forming at very cold temperatures, often below minus 100 degrees Celsius (minus 148 degrees Fahrenheit)," said Tony Colaprete, a planetary scientist at NASA's Ames Research Center, Moffett Field, Calif. "What we have found in our laboratory studies is that it is much harder to initiate cloud formation at these cloud temperatures than what we thought," he explained.

The last statement in the NASA news release is misleading:
The amount of water in the martian atmosphere varies greatly in spaceand time," Colaprete observed. Clouds in the atmosphere largelycontrol the amount of water that comes off of the north pole andmigrates to the south pole."If all the water in the atmosphere were to freeze out to the surface,it would make a layer of ice about one-fifth the thickness of a humanhair, according to Colaprete."Cloud mass is typically only 10 to 20 percent of the total watercontent. However, the thin martian atmosphere is much more sensitive/reactive to the influence of these clouds," he said.
 Since the water vapor content on Mars is known to be so low that implies that the water content in any cloud must be even lower. But actually it is because overall the cloud cover of the entire planet is relatively low.
 But the water content in precipitation clouds can be much higher than the water vapor content in the surrounding area. For instance during a storm you can have many inches of rainfall or snowfall. But the water vapor content on Earth is at most 5 to 6 precipitable cm, about 2 to 2.5 inches (the amount of water vapor in an atmospheric column if it were condensed.)

The NASA report focused on clouds at very cold temperatures -100C. But it is known there are daytime clouds/fogs very close to the surface on Mars where the temperatures will be much higher than this, frequently above -40C for instance. As mentioned above, this is a temperature at which even pure water in clouds can undergo supercooling to remain in liquid form. Indeed supercooling is a major part of cloud formation on Earth:


 In any case clouds at such low temperatures have been observed on Mars. Also even at such low temperatures it is still possible some proportion of the condensed water in the clouds is in liquid form. For instance actual measurements of Polar Stratospheric Clouds on Earth show that liquid water aerosols with nitric and sulfuric acid can be liquid down to -80 C:

Polar Stratospheric Clouds.
Type I a (Nitric acid trihydrate particle - NAT)
crystalline particles forming at 195 K,
Type I b (Supercooled ternary solution - STS)
spherical liquid particles forming at 193 K,
Type II (Water ice) ice crystals forming below 188 K.

Chemical Analysis of Polar Stratospheric Cloud Particles.
Jochen Schreiner, Christiane Voigt, Andreas Kohlmann, Frank Arnold, Konrad Mauersberger, Niels Larsen
Science, 12 February 1999: vol. 283 no. 5404 pp. 968-970.
A balloon-borne gondola carrying a particle analysis system, a backscatter sonde, and pressure and temperature sensors was launched from Kiruna, Sweden, on 25 January 1998. Measurements within polar stratospheric cloud layers inside the Arctic polar vortex show a close correlation between large backscatter ratios and enhanced particle-related water and nitric acid signals at low temperatures. Periodic structures in the data indicate the presence of lee waves. The H2O/HNO3 molar ratios are consistently found to be above 10 at atmospheric temperatures between 189 and 192 kelvin. Such high ratios indicate ternary solution particles of H2O, HNO3 [nitric acid], and H2SO4 [sulfuric acid] rather than the presence of solid hydrates.

 Because these Earth clouds are stratospheric, they occur at pressures near those on the surface of Mars. Then low lying fogs or clouds on Mars would occur at similar pressures and temperatures to the liquid water containing PSC's on Earth.

 Also recent work by Bogdan suggests some liquid water in clouds could remain down to -140C(!)

New Observations On Properties Of Water.
ScienceDaily (Dec. 13, 2006) -- Recent research on the properties of
water reveals information relevant for cloud physics and even
cryopreservation science.
Experimental studies conducted by Ph.D. Anatoli Bogdan at the
University of Helsinki, Finland, have received broad interest in the
scientific world, as the results might have applications even in the
cryopreservation of cells and tissues. Bogdan's results show that
mixture droplets consisting of sulphuric acid and water can be slowly
cooled down to-140 degrees Celsius and then heated again without ice

Reversible Formation of Glassy Water in Slowly Cooling Diluted Drops.
J. Phys. Chem. B, 110 (25), 12205 -12206, 2006. 10.1021/jp062464a S1520-6106(06)02464-3
Web Release Date: June 6, 2006

 This might make it possible for even lower temperature Polar Mesopheric Clouds, also called noctilucent clouds, on Earth to contain liquid water. The Aeronomy of Ice in the Mesosphere (AIM) satellite was recently launched to study such clouds on Earth. If such clouds are also found to contain some proportion of liquid water, that would greatly increase the range of possibilities for liquid water in clouds on Mars.
 The term "noctilucent clouds" comes from the fact they often have a luminous appearance at night:

The secrets of night shining clouds.

Meteor Smoke Makes Strange Clouds.
August 7, 2012:  Anyone who's ever seen a noctilucent cloud or “NLC” would agree: They look alien.  The electric-blue ripples and pale tendrils of NLCs reaching across the night sky resemble something from another world.

 The NLC's often have a bluish tint. Interestingly it has often been noticed both from ground-based observations and from Mars spacecraft that there are frequent bluish clouds on Mars:

Mars Pathfinder.
Dr. Mark Lemmon, University of Arizona
Mars Pathfinder Imaging Team

These clouds from Sol 15 have a new look. As water ice clouds cover the sky, the sky takes on a more bluish cast. This is because small particles (perhaps a tenth the size of the Martian dust, or one-thousandth the thickness of a human hair) are bright in blue light, but almost invisible in red light. Thus, scientists expect that the ice particles in the clouds are very small. The clouds were imaged by the Imager for Mars Pathfinder (IMP).

 It would be interesting to find out if the NLC's on Earth contain the sulfuric acid content to allow the NLC cloud particles to remain partially liquid down to the extremely low temperatures suggested by Bogdan It would also be interesting to find out if the bluish clouds on Mars also have this sulfuric acid component.

   Bob Clark

About the life on Mars question.

Copyright 2012 Robert Clark

 I was watching “This Week at NASA” after the Curiosity landing. I was interested in how the voice-over describing the landing phrased the life on Mars question. It said Curiosity will try to determine if the conditions are right for microbial life to exist on Mars:

Curiosity Has Landed! on This Week @NASA.

It was notable to me this was phrased in the present tense, not for microbial life to have existed on Mars, but to exist on Mars. Since Viking with the general consensus that the current life on Mars question was answered in the negative, usually NASA missions were described as only determining if life could have existed in the past on Mars, not the present.
On the “NASA360″ episode shown this week, the NASA scientist interviewed Dr. Bruce Jakosky of the Curiosity and upcoming MAVEN Mars missions described them also as determining if conditions are right for life to exist on Mars, present tense:

NASA 360 Season 3, Show 19.

Bob Clark

Saturday, August 4, 2012

The Coming SSTO's: Falcon 9 v1.1 first stage as SSTO.

Copyright 2012 Robert Clark 

  I've been arguing that SSTO's are actually easy because how to achieve 
them is perfectly obvious: use the most weight optimized stages and 
most Isp efficient engines at the same time, i.e., optimize both 
components of the rocket equation. But I've recently found it's even 
easier than that! It turns out you don't even need the engines to be 
of particularly high efficiency. 
SpaceX is moving rapidly towards testing its Grasshopper scaled-down 
version of a reusable Falcon 9 first stage: 

Reusable rocket prototype almost ready for first liftoff. 
Posted: July 9, 2012 

SpaceX deserves kudos for achieving a highly weight optimized Falcon 9 
first stage at a 20 to 1 mass ratio. However, the Merlin 1C engine has 
an Isp no better than the engines we had in the early sixties at 304 
s, and the Merlin 1D is only slightly better on the Isp scale at 311 s. 
This is well below the highest efficiency kerosene engines (Russian) 
we have now whose Isp's are in the 330's. So I thought that closed 
the door on the Falcon 9 first stage being SSTO. 

However, I was surprised when I did the calculation that because of 
the Merlin 1D's lower weight, the Falcon 9 first stage could indeed be 
SSTO. For the calculation we'll need the F9 dry mass and propellant 
mass. I'll use the Falcon 9 specifications estimated by GW Johnson, a 
former rocket engineer, now math professor: 

Reusability in Launch Rockets.

The first stage propellant load is given as 553,000 lbs, 250,000 kg, 
and the dry weight as 30,000 lbs, 13,600 kg. 

I'll actually calculate the payload for the first stage of the new version of 
the Falcon 9, version 1.1. The Falcon Heavy will use this version's first stage 
for its core stage and side boosters. SpaceX expects the Falcon 9 v1.1 
to be ready by the end of the year. 

Elon Musk has said version 1.1 will be about 50% longer: 

Q&A with SpaceX founder and chief designer Elon Musk. 
Posted: May 18, 2012 

I'll assume this is coming from 50% larger tanks. This puts the 
propellant load now at 375,000 kg. Interestingly SpaceX says the side 
boosters on the Falcon Heavy will have a 30 to 1 mass ratio. This 
improvement is probably coming from the fact it is using the lighter 
Merlin 1D engines, and because scaling up a rocket actually improves 
your mass ratio, and also not having to support the weight of an upper 
stage and heavy payload means it can be made lighter. 

So I'll assume for this SSTO version of the Falcon 9 v1.1 the mass 
ratio is 30 to 1, which makes the dry mass 13 mT. 

To estimate the payload I'll use the payload estimation program of 
Dr. John Schilling:

Launch Vehicle Performance Calculator. 

It actually gives a range of likely values of the payload. But I've found 
the midpoint of the range it specifies is a reasonably accurate estimate 
to the actual payload for known rockets. 

Input the vacuum values for the thrust in kilonewtons and Isp in 
seconds. The program takes into account the sea level loss. SpaceX 
gives the Merlin 1D vacuum thrust as 161,000 lbs and vacuum Isp 
as 311 s: 


For the 9 Merlins this is a thrust of 9*161,000lb*4.46N/lb = 6,460 
kN. Use the default altitude of 185 km and select the Cape Canaveral 
launch site, with a 28.5 degree orbital inclination to match the 
Cape's latitude. 

Input the dry mass of 13,000 kg and propellant mass of 375,000 kg. 
The other options I selected are indicated here:

Then it gives an estimated 7,564 kg payload mass:

Launch Vehicle: User-Defined Launch Vehicle 
Launch Site: Cape Canaveral / KSC 
Destination Orbit: 185 x 185 km, 28 deg 
Estimated Payload: 7564 kg 
95% Confidence Interval: 3766 - 12191 kg 

This may be enough to launch the Dragon capsule, depending on the mass 
of the Launch Abort System(LAS). 

Bob Clark

UPDATE, Sept. 26, 2013:

 See more accurate calculations using Dr. John Schillings Launch Performance Calculator here:

The Coming SSTO's: Page 2.

UPDATE, August 26, 2014:

 This blog post actually used the estimated specifications for the Falcon Heavy side boosters, as this was supposed to have an even better mass efficiency than the core stage. However, Elon Musk gave a speech where he gave some values for the Falcon 9 v1.1 first stage that allow us to estimate the propellant and dry masses for the stage. Then we can calculate the payload capability of a F9 v1.1 core stage SSTO itself. I estimate it as ca. 5,000 kg:

The Coming SSTO's: Falcon 9 v1.1 first stage as SSTO, Page 2.