Studying the issues with LV13.1 (Ares I classic)

[Downix]

Earlier on the SD-HLV two stage thread, got a bit riled up at the claims made about the J-2X, that it was designed with the purpose of being both an ideal US engine and EDS engine, so went and dove into the ESAS to uncover the goal, and the end result.  What I discovered was that within the ESAS, it dismissed the replacement of SSME with the J-2S+ (the engine which eventually became the J-2X), stating that such a move would torpedo both cost and timeline, which is exactly what happened.  I also saw the target specifications for the J-2S+, and the J-2X fails to meet the target for EDS, sacrificing them to fulfill the new role on Ares I.

So, this got me to think, what alternative did they have in 2005?  This got me to wracking my brain, they did not want to put a new engine, as the J-2X was, on the long pole.  Then it hit me, there was an air-startable engine already in service at that time, not that pricey an engine, with a long track record including being used for manned flight.

The LR-87.

The desire for an LH2 engine could be addressed, as the LR-87 has been flown with both hypergolics and RP-1, and done full testing with LH2.  The performance for two of the LH2 version would have higher thrust than the SSME, with lower isp, and the Hypergol would be even better for thrust, again at a loss of isp.  Two LR-87-11's would fit under the Ares I Classic tank, and together they could produce 2,185.6 kN of thrust vac, 1,721.7 kN SL.  It also gives a buffer, should the engine effort fall behind schedule, you can load the AIUS with Hypergolics instead of LH2 and run the engines as/is.  Using the existing 4-seg SRB, the LH2 version lifts 25mT and the Hypergol version can lift 23.4 mT.

So, what you'd have is a viable launch vehicle, even if not ideal due to the use of an SRB as first stage, but a workable one without a long-pole.  Fly the hypergol version first, waiting for the LH2 version to come online.  The VAB is dealing with hypergols already from the OMS, so this is a non-issue.  The total GLOW is even light enough to allow it to be carried to the pad w/o fueling once on the pad.  While, yes, Hypergol's are not the ideal launch fuel, a bird in the hand is worth two in the bush.  What killed Constellation was putting too many long-poles up front.

Ah, the fun of heated discussions making me ponder the situation, what went wrong, and what could have been done instead.

[Jim]

The OMS is basically empty in the VAB and loaded at the pad.  The proposed amount hypers would not be allowed in the VAB, much less be loaded there.

Production capability of the LR-87 was shutdown long ago, and it would be almost a new start like the J-2

[beb]

Production capability of the LR-87 was shutdown long ago, and it would be almost a new start like the J-2

Trust Jim to throw cold water on a hot idea.

So it would appear that Constellation was doomed from the start by being given an impossible short deadline when the U.S.  simply lacked the kinds of engines needed for the mission.

[Downix]

Production capability of the LR-87 was shutdown long ago, and it would be almost a new start like the J-2

Trust Jim to throw cold water on a hot idea.

So it would appear that Constellation was doomed from the start by being given an impossible short deadline when the U.S.  simply lacked the kinds of engines needed for the mission.

There was one last engine, as I mentioned a few times: AJ-26.

*edit* also checked, there are several LR-87-AJ11's in storage which could be used for initial flight testing.  I even located one of the LH2 modified units.  And I also checked, as of 2001 the LR-87 was in production, and there is nothing to discuss discontinuing of production until 2004, less than a year before we are discussing.  Restarting production of an existing engine within such a short period is not a game breaker, and there would be existing engines to utilize in the interim until production was ready.

But yes, Constellation was doomed due to the inability of management to read their own report.  Reading the ESAS report, there was a chain of options for crew launch, based on time to delivery, development schedule, and cost as well as crew safety and reliability.  The chain went from LV13, the 4-seg w/ 1 SSME Ares I to an Atlas V HLV as/is, to Delta IV Heavy as/is, to Delta IV Heavy w/ a new J-2S+ based upper stage, to the Atlas V HLV w/ the new J-2S+ based upper stage, to *then* to the LV17 5-segment w/ J-2S+ upper stage.  They even gave a breakdown as to why LV17 and LV19 (the two 5-segment based designs in the study) were so far down the chain, listing concerns over 5-segment vibration along with the time and cost to develop not just one, but two new engines before even basic flight testing could be started.  They also voiced concerns that the development may not give the performance necessary, and that such performance loss would result in even further delays as the CEV would need to be re-designed.  They even discussed an evolution type path, to prevent such a critical loss of access, that if LV13 would not work, to use the Atlas V HLV as/is, developing the future LV17 upper stage so that it could be fitted up to the Atlas V as an initial upgrade, and then if the 5-segment SRB can meet performance, replacing the three CCB's with the 5-segment only at that time.  This would have eliminated the long-poles immediately from the path, while still giving the potential for the future for growth.

[Downix]

This morning I had a bit of epiphany.  One of the concerns I've had with the Ares I design, a concern which was reinforced by the Ares IX flight, is that the residual fuel within the SRB casing could add a pulse of thrust after staging, and push the second stage into the upper before the engines had fully engaged.

This morning I looked at the old N-1 and Soyuz, and it struck me a solution so brilliant yet simple. 

[Downix]

Not saying to be exact, but the idea was to enable the upper stage engine(s) to fire while still attached to the first stage, enabling the venting of exhaust.  Could be through blow-away panels or some such to use as a channel.

[Jim]

They didn't reuse the lower stages.  Ares I was to use ullage and retrorockets to prevent recontact.

[simonbp]

Not saying to be exact, but the idea was to enable the upper stage engine(s) to fire while still attached to the first stage, enabling the venting of exhaust.  Could be through blow-away panels or some such to use as a channel.

Um, they tried that for CLV/Ares I. IIRC, problem was the solid first stage accelerated so fast that the lateral aero loads at max Q crushed the open truss interstage. Thus, they went with a much more beefy solid interstage.

[Downix]

They didn't reuse the lower stages.  Ares I was to use ullage and retrorockets to prevent recontact.

You'd need a retrorocket with as much power as the SRB in the last 10% of it's burn, nothing I've seen in any proposal addresses that with the 5-segment.

Let us compare.  To stop a 4-segment SRB at the recommended staging point, with 10 seconds of potential burn remaining for safety (typically it is 6 seconds, but I like margin).  At this point, we would need between 218kN and 667kN.  Not horrible, could use a few of the high-thrust STAR's with the burn time reduced, not a game-stopper.

Now, let's do the same on the 5-segment, using ATK's own test data.  The first test, you would have needed 1,733 kN.  The second, you would have needed 1,848 kN.  Now you're in trouble.

[Jim]

Let us compare.  To stop a 4-segment SRB at the recommended staging point, with 10 seconds of potential burn remaining for safety (typically it is 6 seconds, but I like margin).

It was staged at burnout

[Downix]

Let us compare.  To stop a 4-segment SRB at the recommended staging point, with 10 seconds of potential burn remaining for safety (typically it is 6 seconds, but I like margin).

It was staged at burnout

You can't do that, because burnout is inconsistant.  Studying the now 262 SRB burnout times (not counting Challenger for they did not burn to burnout), burnout occurs between 126 seconds and 132 seconds without any pattern.

[ugordan]

What do you mean you can't do that? If staging is triggered when the chamber pressure drops below a certain threshold, what difference does it make if it happens a few seconds earlier or later?

[Downix]

What do you mean you can't do that? If staging is triggered when the chamber pressure drops below a certain threshold, what difference does it make if it happens a few seconds earlier or later?

Because out of those 262 SRB burnouts, the pressure has dropped then raised again 114 times within the last 10 seconds of burn, with three of those being reported as 0 pressure before reappearing for one last burst.  You must stage before SRB burnout or else you endanger the crew.

[kch]

What do you mean you can't do that? If staging is triggered when the chamber pressure drops below a certain threshold, what difference does it make if it happens a few seconds earlier or later?

Because out of those 262 SRB burnouts, the pressure has dropped then raised again 114 times within the last 10 seconds of burn, with three of those being reported as 0 pressure before reappearing for one last burst.  You must stage before SRB burnout or else you endanger the crew.

Were there any instances of that last burst persisting past the 132 second point?  If not, would it be reasonably safe (I know, nothing's 100%) to delay staging (coast) to around 135-140 seconds?  Roughly how much performance penalty would be incurred by doing so?  Just wondering if that's a workable procedure.

[Downix]

What do you mean you can't do that? If staging is triggered when the chamber pressure drops below a certain threshold, what difference does it make if it happens a few seconds earlier or later?

Because out of those 262 SRB burnouts, the pressure has dropped then raised again 114 times within the last 10 seconds of burn, with three of those being reported as 0 pressure before reappearing for one last burst.  You must stage before SRB burnout or else you endanger the crew.

Were there any instances of that last burst persisting past the 132 second point?  If not, would it be reasonably safe (I know, nothing's 100%) to delay staging (coast) to around 135-140 seconds?  Roughly how much performance penalty would be incurred by doing so?  Just wondering if that's a workable procedure.

The longest burn I saw ended at 132.1 seconds.  If you staged at 135 seconds, you'd loose some momentum.  But, there is a way to help offset this.

Pondering it a bit, using the ullage discussion a bit.  If they used a few small solids attached to the outside of the upper stage, with ~5 second burn times, they may be able to off-set this.  Looking at ATK's catalog, 8 Star 8's would do the job.

[ugordan]

Because out of those 262 SRB burnouts, the pressure has dropped then raised again 114 times within the last 10 seconds of burn, with three of those being reported as 0 pressure before reappearing for one last burst.  You must stage before SRB burnout or else you endanger the crew.

That's what the BDMs and ullage motors on 2nd stage were for. I am well aware SRBs don't shutdown completely and rapidly, but it's not an insurmountable challenge.

And no, I don't like Ares I either.

[Downix]

Because out of those 262 SRB burnouts, the pressure has dropped then raised again 114 times within the last 10 seconds of burn, with three of those being reported as 0 pressure before reappearing for one last burst.  You must stage before SRB burnout or else you endanger the crew.

That's what the BDMs and ullage motors on 2nd stage were for. I am well aware SRBs don't shutdown completely and rapidly, but it's not an insurmountable challenge.

And no, I don't like Ares I either.

Never said it was an insurmountable challenge.  And I'm with you.

Truth be, I don't hate Ares I, nor even really dislike it.  The *idea* has a certain appeal.  But the reality, just doesn't work.

Maybe a large segmented Hybrid motor? 

[kch]

Maybe a large segmented Hybrid motor? 

Now there's a thought! 

[Downix]

Ok, out of morbid curiosity I took one of the ESAS ideas that was implied and explored it, namely using a different US on the EELV's.  I used the LV17's US, using the J-2S+ but with the AIUS weight gain to reflect the evolutionary path. First I took a Delta CBC sans engine, put an SSME on the base, put an AIUS on top, and then added four ATK GEM-60.  It could lift over 25mT to the ISS.  When I tried with two, it did not lift off the pad, not enough ground thrust.  When I tried with an RS-68, with the four GEM-60's I got it to 23mT.  Two, 20mT.  None, it could not lift off of the pad, it was too heavy.  Then another thought, a Heavy with an SSME on the core, which hit 44mT.  Remove the US entirely, 26mT. Swap out Delta CBC's with Atlas CCB's, 29mT.

Then I tried with a stock Atlas CCB for the first stage.  As/is, it could loft 18mT.   Adding a single Atlas SRB got it to 22mT, then adding two got it to 25mT.

This tells me that the ESAS missed a whole evolutionary chain in the process.  The ESAS had mentioned launching on the EELV heavies, but with their demand for single-engine for safety they put the 4-seg+SSME ahead of them.  Here is a way to have closed the gap, and not had any long poles while giving them the "out" they wanted.

Baseline a NASA-derived EELV's for initial launch capability, with a staged evolutionary path to hit the 5-seg first stage and J-2S+ for US work. The initial launch vehicle would be the combo Delta CBC w/ SSME paired with Atlas CCB's, with no US.  This configuration would, per the rules set forth for safety in the ESAS report, bring the safety of the system from Delta IV Heavy's LOM of 1:172/LOC of 1:1,100 to LOM of 1:384/LOC of 1:1,864.  The cost, per ESAS calculation methodology, would also be 0.95, less than both LV13 and LV17.

This would give initial capability, a NASA vehicle which uses parts supplied by Lockheed and Boeing.  They would then work on the LV17 US and 5-segment booster (as the 5-segment is needed for the Ares V anyways) without the need to modify the J-2S+ for higher thrust, keeping it's higher ISP form.  If the US works out, great, can switch to a single Atlas CCB for crew launch, and same with the 5-segment SRB as first stage.  At no point do you have a critical path element tied to an undeveloped and untested element.  Flying Delta+Atlas together is unproven, so worst case switch to pure Delta w/ SSME on center configuration instead or stock Atlas V HLV.  This would also give more R&D money to key areas over the final proposal, from the RD-180/alternative engine domestic production startup.  Alternatively, and a way to keep ATK happy and still meet the demands, ATK as part of the 5-segment program made a more scalable segment design, and two two-segment SRB's, one on each side of the core, would give it more than enough push off the ground.  Or could fund development of more powerful GEM's, which would benefit both this project as well as ULA.  In addition, the J-2S+, as planned (and looked likely to hit based on later data) would have been a solid engine for the EELV's, and a cut-down form of the AIUS (could shrink it by 1/3 w/o issue based on the design) which levereged Centaur/DCSS technology (which it does if you study it) would have been a great US for both EELV's, rivaling ACES.

[Downix]

Had another thought, and instead of swapping the US engine, I re-configured the 4 SRB segments into two 2-segment SRBs (taking the middle two segments out of the 4-segment SRB, which has been proposed by Thiokol and later ATK on a few occasions, and based on data should work fine) and strapped them to the side of the ESAS LV13 upper stage.

Worked fine, lifting 26mT to the iSS>