DIRECT v3.0 - Thread 2

[adamsmith]

Ok, I’m going to step in here before things get out of hand and say something, and I hope this helps clear some of this up. We designed the DIRECT architecture to be a tool of the policy. We created an architecture that accomplishes the goals of the VSE as they were laid out by President Bush and adopted by the Congress. DIRECT, and the Jupiter launch vehicle, is not the goal; it is a tool of national policy as established by the President...

If you recall the charter of the Commission, it was to examine Human Spaceflight as a national policy...

If the final report does not change the overall direction of the VSE and its goals remain essentially intact, then we believe that the DIRECT architecture, which the Commission does have in its possession, and which uses the Jupiter launch vehicle, is the best tool available to implement that policy and achieve those goals within a reasonable time frame and for reasonable expenditures. We are confident that reasonable policy makers are smart enough to see that. Remember - we elected these people to make these kinds of decisions. We elect people who we believe are smart enough to surround themselves with smart people. The ballot box is no small thing.

Now it’s vital to remember that DIRECT is a tool, not an end in and of itself. I say that because there is a very real chance that the final result of all this is that the policy might change wrt HSF... The purpose of national policy is not to provide a program for the Jupiter. It is to establish and set a direction for American Human Spaceflight going forward, and to select appropriate tools to help meet the goals of that direction. DIRECT is one such tool, out of many, designed to serve a very specific policy.

Nobody, except maybe Ross, is more hard-core DIRECT than me, but even we recognize that DIRECT, as good as it truly is, only has value to the nation so long as it serves national policy better than the other tools.

Let’s all just take a step back and wait to see what direction the policy makers want to take American HSF. Then, and only then, can we determine if DIRECT remains viable or not. It’s all a matter of policy...

For the record, and in spite of some appearances to the contrary and hand waving by a series of bloggers, we believe that the men and women of the Commission and the employees of the Aerospace Corp are doing the best they can with the mountains of data they have and with the very short time frame they were given. Will mistakes be made? Undoubtedly; it could not be otherwise under these circumstances. Is it possible for the Commission members to be misled by slanted data that comes from sources other than Aerospace? Sure it is. Is it possible that DIRECT could be swallowed up by one of these mistakes? Yes, but that’s the way these kinds of things have always gone. History is littered with good ideas that got overlooked by mistakes. It’s entirely possible for that to happen again here. Let’s just hope that, for the sake of the nation and American HSF (not for DIRECT’s sake) that such a mistake is not made. But if it is, then it is. We do not believe that anyone at Aerospace or on the Commission is deliberately skewing the data to reach a predetermined conclusion. They are all working very hard, and with great integrity, to provide the President with good data from which to form a good national policy. Let us all wish them well and just wait to see what they come up with...

My question is this, Does the Direct Architecture make sense without Jupiter, or assuming Propellant Depots, as appears likely, can it be recreated assuming some other Launchers?

Stanley

[A_M_Swallow]

My question is this, Does the Direct Architecture make sense without Jupiter, or assuming Propellant Depots, as appears likely, can it be recreated assuming some other Launchers?

Stanley

If the other launches have a similar payload to the Jupiters then the Direct Architecture can be reused with minor modifications but watch out for differences in timescales.

Something like all EELV would need a different architecture but the Direct documents can be raided for good ideas, particularly the out of LEO sections.

[fotoguzzi]

I thought end of September was always the Committee deadline, but I guess it has become more official, now:

http://www.nasa.gov/offices/hsf/related_documents/what-the-committee-is-doing.html

[MP99]

My question is this, Does the Direct Architecture make sense without Jupiter, or assuming Propellant Depots, as appears likely, can it be recreated assuming some other Launchers?

Stanley

DIRECT's Lunar mission is based on a coincidence that TLI with H2/O2 is approx 50:50 fuel & burnout mass.

The EDS launches with the fuel, but is part of the burnout mass, so must be very light but with max fuel load. TLI mass is maximised by using an over-sized upper stage (ref Saturn V & Ares V as well as Jupiter).

The crew launch is therefore somewhat less than 50% of the total mass, but that provides for the mass of the LAS, and additional margin for the crew.

Any architecture with a light EDS can perform a reasonably efficient 2-launch TLI (mass dependent on launch mass).

With an upper-stage-as-EDS and a depot, your EDS has a lot more fuel and can push a large mass to Mars / NEO's (as long as the total engine burn time (ascent + TMI + maybe MOI) isn't breached).

cheers, Martin

[Mark S]

Great.  Now we get to endure a-whole-nother month of rampant speculation and idle rumor mongering.  I thought we were going to see some results by next week.

I hope the DIRECT team is still refining their designs, such as the new thrust structure and SSME arrangement.  I think DIRECT is still the best solution, and probably the front runner due to its ability to extend ISS without busting the budget.  But that's just rampant speculation on my part.

At this rate, Gen. Bolden won't have time to read the report before he has to decide whether or not to cancel the Halloween launch of Ares-IX.

Sigh.

[adamsmith]

To those that have responded to my Architecture comment.  Thanks a lot for the insight.

Stanley

[robertross]

Great.  Now we get to endure a-whole-nother month of rampant speculation and idle rumor mongering.  I thought we were going to see some results by next week.

Sigh.

I'll second that sigh. 

What this says to me is that the numbers from Aerospace are still coming in, or waiting to come in. If we're (they're) waiting on NASA for numbers from NSC calculations & costings, it could sway things a bit (if there even is such a thing at this point). They may have also been cued by the Direct team as to their misinformation and are re-writing the whole thing for options. Who knows??

One thing I hope is that if it takes a little longer to do it right, then fine.

[Variable]

Great.  Now we get to endure a-whole-nother month of rampant speculation and idle rumor mongering.

Easy to solve.

[Mark S]

Great.  Now we get to endure a-whole-nother month of rampant speculation and idle rumor mongering.

Easy to solve.

Start my own space program?

[mjcrsmith]

I thought end of September was always the Committee deadline, but I guess it has become more official, now:

http://www.nasa.gov/offices/hsf/related_documents/what-the-committee-is-doing.html

While we are all anxious for the report, better that the committee gets as much data analyzed as possible.  An additional few weeks now, may result in a better future for HSF.  My hope is that the NSC and DIRECT data will be differentiated and compaired.  I think the advantages of DIRECT will be obvious.

[Variable]

Great.  Now we get to endure a-whole-nother month of rampant speculation and idle rumor mongering.

Easy to solve.

Start my own space program?

Oooo, can I haz rokit syenteest?

LOL  Hey, I hear they're doing great things with water Down Under!

Nah, no need to start your own program.  Just quit speculating and what-iffing.  Just wait and see what comes out in the report.

[Nathan]

One wonders whether NASA isn't running it's own budget numbers on architectures at the moment ,with some though going into Direct and some thought going into HLV. It would make sense for them to have a counter plan.

Anyone in the know able to hint at this?

[Ben the Space Brit]

This is something that has been working its way around my brain since I read the proposed timelines and missions being considered by the Augustine Commission.  Remember that I am not a professional, just an interested amateur.  I can really only respond to what I can obviously see in front of me as to technology development and costings. 

Still, disclaimer aside, this is my idea of how to modify the DIRECT 3.0 proposals for the new post-Augustine reality.

Key Assumptions
* The Ride cost figures for SDLV were exaggerated by ~50%, possibly so as to support the case for large-scale space 'privatisation' and make Ares-I look less of a train wreck;
* ESA can build one ATV or ATV-derived vehicle a year without significant (and costly) changes to their operations;
* The Russians will proceed with their next-generation crew vehicle (I call it 'Son of Soyus' or 'Orionski' depending on my mood) and the Angara LV.  These will both be operational by 2016.

Spacecraft to develop
* J-130 (LEO crew & cargo);
* J-241/2 (the number of upper stage engines TBD based on eventual J-2X performance figures);
* Orion CEV;
* ATV-derived cargo tug for SSPDM (aka 'Robot Resupply Barge' or RRB);
* ATV-derived storage/lab module (aka ATVLab); This is to also be compatible with new Russian crew vehicle;
* Altair-lite (4 crew for 4 days on surface or unpressurised one-way cargo hauler);
* 2 x CRS and Commercial Crew Transfer (CCT) vehicles with associated LVs;
* Possible development of 5-segment RSRM depending on payload requirements.

Timeline (All Dates are Calendar not FY)
Thanksgiving 2009 - Project kick-off
Mid-2010 - First Jupiter CCB ground test unit completed
2011 - STS Retirement, MAF converts to building Jupiter CCB
2013 - CRS to begin
mid-2015 - CCT to begin test flights; Jupiter-130X test flight; JUS development begins at MSFC
mid-2016 - CCT replaces Soyuz as default US access to ISS
early-2017 - Jupiter-130 IOC with first flight of RRB (no more than 1 flight per 2 yrs)
late-2017 - Orion flight test program begins
mid-2018 - Jupiter-130 FOC/Orion IOC. 
mid-2019 - Orion FOC; Test flights to ISS to practice rendezvous systems & 6-month uncrewed 'hibernation' mission in LEO to test untended loiter software
early 2020 - J-241 IOC; JUS launched on free-return translunar flyby to test vacuum start and RCS
mid-2020 - ISS retired; Possible Orion missions to return some items from the station such as EXPRESS pallets; De-orbit by modified ATV; Altair-lite development begins in earnest
late-2020 - Orion/ATVLab missions in LEO as stop-gap until J-241 FOC; Launch of cryo depot test unit 1; First CRS-F commericial propellent launch
~2021 - First JUS refuelling from cryo depot; Uncrewed Orion translunar mission to test lunar return re-entry and measure cabin conditions beyond the magnetosphere.
~2022 - J-130/WBC used to launch first Orion crewed translunar flyby.
~2023 - J-241 FOC - Orion/ATVLab flyby to NEO supported by cryo depot; Orion/ATVLab LEO missions to continue at rate of 2-3 flights/year when NEO windows not open
~2024 - Orion/ATVLab launched to either LLO or EML-1 for long-term experiments on lunar orbital environment; Either depot-refuelled J-241 or 2 x J-130 (one with Orion/WBC the other with ATVLab)
~2025 - Altair-Lite IOC; crewed LEO and uncrewed LLO tests
~2025-26 - Altair-Lite FOC; First lunar landing (Phase C) supported by cryo depot
~2026-28 - Phase D landings with cargo Altair pre-positioned to increase surface endurance

"Direct 4.0" Developments
* J-241N with NTP/NEP upper stage
* Reusable Lunar Access Vehicle (RLAV)
* Hypergolic depot/crewed EML-1 transfer station
* CRS-L and CCT-L programs to supply transfer station and later lunar outpost
* Reusable MTV for long-haul Inner Solar System crewed missions to include own MPS and inflatable TPS for Earth return aerobraking

[EDIT]
Added one mission

[MP99]

You could do it w/o depot if you did a 246 + a 130 I believe, with the 246 carrying all but the crew module, the 130 a crew module + the fuel needed, based on my math. (which, I am no rocket scientist, so if I'm wrong let me know)

That'd probably work, especially since according to Ross, Direct can do that anyway without a depot.  The J-130 lifting the CSM and LSAM, each doing their own LEO circularization burn, and the J246 launches the EDS, and the two vehicles rendezvous with it and go to TLI.
So they why would you need the Depot?

That works OK for cargo missions, but the margins are probably too low to comfortably do it for crewed missions (has to launch mass of both Altair & Orion). Assumes standard Orion & Altair, of course.

cheers, Martin

[DonEsteban]

Lobo -

I think that the main difference is that there are only 2 Mission Critical launches - The J24x and the Crew Launch.  There may indeed be as many as 8 to 10 total launches, but if one or two Depot launches go awry, the mission is not at risk.  And with more and more launches (as mentioned many times before) the reliability of the launch system to the Depot increases and the cost per launch theoretically decreases.

TOG

Yes, yes, I know all about that.  I was just expanding on the other post by saing a launch is a luanch.  You don't have a 1/2 launch, and you don't have prollant that magically appears in your depot.  They are all launches.  And even propellant launches are "mission critical" because if one is aborted, or delayed, you don't have enough prollant to get to the moon.

No, no, no. You have excess propellant at the depot. If one or few propellant launches are delayed, there is less excess propellant left at the depot after you fuel-up for Moon. If you want to mitigate that, launch more propellant flights. But one-few propellant flights are definitively not "mission critical". The added flexibility and robustness is significant.

I'd call that "mission critical".  It's just that there's a certain amount of flexability in that you can send up a replacement launch to fill the manifest.  HOwever, on that same vein, you could send up a 2nd J-246 with EDS if something happened to the first, since that would be launched before the crew anyway. 

But the key is the cost difference between a replacement propellant flight (cheap, possibly low-reliability rocket, low-cost cargo) vs the second J-246 with EDS (expensive rocket and cargo).

So the only true LOM failure is the J-246 carrying the crew and LSAM, and that wouldn’t change from baseline to depot.  A the EDS J-246 would cost more than a single failed EELV launch, granted.

But, without X successful propellant launches to a depot, there is no mission.  If there’s a problem with one, you either replace it or scrub the mission, so in that sense each –is- mission critical. 

To have a serious problem with propellant launches (= many launches lost, after several successful launches establishing the capabilities) is quite low probability. Much lower then a loss of low-flight rate (and thus inherently either overengineered/costly or unreliable) J-246.

Plus you have to coordinate with other entities which as we found out with the ISS is not always as easy as it sounds.

ISS is not a good example - the partners are delivering one-off complicated items. Delays are natural/typical in such case.

With propellant delivery you sign contract with somebody who has already proven some capability, and you want them to just keep repeating that.

 
So, just pointing out that a lot of PD advocates tend to speak about it like you are reducing rather than increased the number of successful launches that the architecture depends on.

Yes, you are increasing the number of successful launches you depend on. But you are decreasing the fraction of the launches that need to be successful. For example, company A has capacity to launch 12 'fuel trucks' per year. Their 'expected reliability' is 90% (quite low, but they are start-up), and you need 8 fuel trucks for a lunar mission. So you ask them to start filling the depot (for simplicity and optimistically assuming zero-boiloff) a year ahead and even if their actual reliability is piss-poor 70%, they will still manage to deliver enough fuel for you. In fact, while doing that many launches, I would expect them to debug their system and end-up with much more reliable launcher then your low-flight rate overengineered/human rated? HLV.

I understand that certain economic factors weight towards PD’s despite their increased launch dependency.  But it always bugs me a bit this misconception that with a depot, we have single launch lunar architecture instead of 7-8 launch lunar architecture.  And that architecture is only economically cheaper if you have others paying for most of those launches.

Just sayin’….
If you can get someone else to pick up the tab for all those propellant launches, then it’s a economically more viable despite being logistically more complex.

The point of depots is that you can launch most of the mass on cheap per kg launcher (high production/low development cost - no overengineering due to  'we can't afford a launch failure).  Another thing is operational flexibility - it is much easier to grow your departure/lunar lander stages. Once those get past single launch and you need docking, then refueling becames cheaper and more reliable as well.

If others pick up the tab for the propellant launches, all the better. But a PD architecture should be better even without them.

[StuffOfInterest]

With propellant delivery you sign contract with somebody who has already proven some capability, and you want them to just keep repeating that.

I presume you would pay for the delivery and not the launch.  Make the contract say "for delivery of up to XXX kg of fuel at a rate of YY per kg delivered".  This puts the onus on the delivery company to get their reliability as high as possible as they are only getting paid if the fuel goes through the meter on the other end.

Trying to stay a bit on topic, any word on when we'll see some artwork for DIRECT 3.1, or whatever version has the clustered engine design?

[jordanch68]

Have any comparisons been made of Direct and other programs that were more promising but not developed? I'm thinking of Dyna-Soar and Big Gemini. Things would be much different indeed had those programs continued.

Also, I think Direct has far more support than any other option I've heard of.

[Lobo]

Yes, you are increasing the number of successful launches you depend on. But you are decreasing the fraction of the launches that need to be successful. For example, company A has capacity to launch 12 'fuel trucks' per year. Their 'expected reliability' is 90% (quite low, but they are start-up), and you need 8 fuel trucks for a lunar mission. So you ask them to start filling the depot (for simplicity and optimistically assuming zero-boiloff) a year ahead and even if their actual reliability is piss-poor 70%, they will still manage to deliver enough fuel for you. In fact, while doing that many launches, I would expect them to debug their system and end-up with much more reliable launcher then your low-flight rate overengineered/human rated? HLV.

<<<snip>>>

The point of depots is that you can launch most of the mass on cheap per kg launcher (high production/low development cost - no overengineering due to  'we can't afford a launch failure).  Another thing is operational flexibility - it is much easier to grow your departure/lunar lander stages. Once those get past single launch and you need docking, then refueling becames cheaper and more reliable as well.

If others pick up the tab for the propellant launches, all the better. But a PD architecture should be better even without them.

Don,

Thanks for the comments.   
Like I said, I’m well aware of the arguments of depot advocates.  I was just stating the fact that’s not a 1 launch architecture, it’s an 8 launch or so architecture.

However, you touch a point about it being cheaper even if others aren’t footing the bill.  I disagree.  I think it’d be for more costly. 
Here’s why.  For a lunar mission with current Orion and Altair, you need about 100mt of propellant for TLI according to Ross and some of the other people on hear smarter than I.  You can do this with a single J-246 launch with the EDS on fully fueled.
According to the cost numbers I’ve heard from Ross and others, we seem to be looking at roughly $350-$450 million for a launch of a J-246, maybe less with a high flight rate.   From all that I could gather, prices for commercial standard EELV class launchers like Atlas V and Delta 4 are in the $250 million range right now.  You’d need probably 7-8 of those launches to get 100mt of propellant into a depot.  You have to factor in tankage, avioniccs, docking hardware, and RCS systems, as well as probably a 2nd stage to get it into a stable, circular orbit for docking with a depot.  All that eats into the actual propellant mass you can loft, as well as increasing costs.  Each tank is it’s own little space ship.  (I know there’s an argument for tugs, but let’s set that aside right now).
So right now, you are looking at well over a billion dollars to fill that depot with US commercial EELV launchers.  Falcon could be a little cheaper, or perhaps Proton, but even Russian Protons are well over $100 million from what I’ve heard.  Even using best case scenario of using Russian Protons at a high flight rate (which I’m sure NASA would rather not do) and figure $100 million a fueling flight, you are around 700 million probably.  Still about double of a single J-246 flight. 
Also, the single J-246 means a single set of explendable tankage, avionics, docking hardware, RCS system, and 2nd stage to the depot.  But then again…why do you need the depot then, just dock Orion and Altair to the EDS and off you go as in Direct baseline.

So you see, in the respect, depot is only less expensive if other parties are footing the bill for all of those fueling flights (not to mention the cost of the depot itself).  My argument is have them “foot the bill” for that cheaper, 2nd Jupiter flight instead.  They save money (so are more likely to do it), coordination is easier, and Jupiter production is increased. 

Many depot advocates seem to feel very confident that international partners will flock to buying fuel launches for a seat on a lunar mission.  If that is so, then NASA only buys one flight and the mission hardware, and saves that 2nd Jupiter.  A win-win.
Myself, I am skeptical that once our international partners get the “first German” or “first Japanese”, or “First Italian” on the moon, that they’ll be so willing to pony up that $700 million to over 1 billion per mission to get another of their countrymen on the moon.  A few may, but I think that well could dry up pretty fast.  I think they’d be much more willing if they could buy an option that was ½-1/3 the cost of all those EELV flights, in “buying” the 2nd Jupiter flight.  Not truly buying mind you, it’s NASA and they can’t really sell a flight.  That’s been made clear to me over and over.  But just like with the ISS, we “bartered” with our partners for shuttle flights.  They built ISS components, or built ATV’s, or logistics modules, etc.  This would be no different.  But if you were an international partner wanting a seat on a lunar mission, and you had Option A that was to use your own EELV or buy commercial and be dropping upwards of a billion for a flight, or barter with NASA for a J-246 flight for half to a third that cost, which would you want to do?  Which would you be most likely to do a 2nd time?  Or a 3rd?

Now, a few caviots in my hypothetical above.  If the cost of Jupiter were far greater than Ross and co have mentioned, that would effect my analysis.  If you could get EELV fuel flights drastically reduced, to say $50 million or less per mission, that would change the math too.  But Ross and co are pretty sharp and I think they are in the ballpark with their numbers, and I doubt greatly you can get an EELV class launch done for anywhere near $50 million.  Even if you cut corners and have “low reliability” like  you said.  SpaceX looks to be doing it about as good as you can, and I think their Falcon 9 projected cost per flight is in well excess of $100 million, and that’s not including payload of the mini spaceship to get the fuel to the depot and dock and transfer it. 
Rockets are just expensive, if you don’t want them blowing up on the pad all the time.  So, if you are trying to get 100mt or so into LEO, sometimes it is just cheaper and more efficient to use a launcher that can do it in one go, rather than a bunch of little launchers.  If you need to get 30 kids from point A to point B, do you use 1 schoolbus, or do you use 10 Priuses? (each needs a driver, and then could seat 3 kids safely)  1 bus getting 10 MPG is cheaper than 10 Pruises getting 45 MPG.

Tugs could reduce the costs some, as you can launch dummy tanks into orbit, and it could tug them to the depot, saving the expenditure of avionics and RCS systems every launch.  But I personally doubt you’d save enough to make it competitive with a single J-246 launch.  And tugs cost money themselves, and if there’s a glitch, you have to send up a replacement.  Another launch and another tug and their expenses.

I’m not against depots.  Many here smarter than me seem to think they are the cat’s pajamas, so I don’t want to pretend like I know something they don’t.  I just have a hard time seeing the math work unless you –are- able to get those international partners to step up and foot those fueling bills. 

[kttopdad]

Don,

Thanks for the comments.   
Like I said, I’m well aware of the arguments of depot advocates.  I was just stating the fact that’s not a 1 launch architecture, it’s an 8 launch or so architecture.

However, you touch a point about it being cheaper even if others aren’t footing the bill.  I disagree.  I think it’d be for more costly. 
Here’s why.  <snip>

Great summary.

[93143]

First, you shouldn't explicitly ignore ideas that make your problems go away.  If tugs help substantially with the cost of fueling, they should be developed.  This isn't Apollo, where the program was so short that development costs completely overshadowed operational costs - or at least, we hope it isn't...

Second, as we all know, the hope is that launch costs will be brought down substantially by the opportunity to launch huge amounts of very cheap stuff into space.  If SpaceX were to get their launch costs down to what they were hoping for originally, you could fill the depot with four F9H launches for less money than a single J-246.

This is why somebody should be paying for development of Skylon (or something comparable, naturally, if anyone has any good ideas).  If they hit their price and performance targets, they should be able to loft 100 mT of propellant for $20-50 million total (before anyone asks:  yes, the payload bay is large enough to max out the vehicle's payload capacity with liquid hydrogen, should you so desire).  And since the vehicle is basically an unmanned space shuttle, it should have no problem with automated prox ops at the depot, especially if the depot has a robotic arm.

Sure, near-term depot missions would have global cost greater than non-depot missions.  But we're trying to get the CATS ball rolling, and launching an SDLV twice as often doesn't help with that.