The design of the Delta IV

[Lobo]

Over in another thread this subject got hit on.  Why did Delta IV get designed like it did?  Why was it so different than Delta II and Delta III?  Both of those used a kerolox RS-27A sustainer engine with a cluster of GEM boosters.  However, with Delta IV they went with a brand new hydrolox engine, new larger GEM boosters as options (rather than required like Delta 2/3), a much wider core at 5m, etc.

Conversely, it looks like Atlas 2 to Atlas 3 to Atlas 5 was a much more synergetic progression.  Keeping kerolox, and just improving the engine and widening up the core some. (and I’m sure other improvements in avionics and such).

So what was McDonnel-Douglas/Boeing thinking on this upgrade?  I’m sure they have great engineers, and there was probably good reasons for it, I’m just not seeing them.  Plus, there’s a reason that hydrolox is normally only used for an upper stage, or systainer stage with some other higher thrust boosters.  It’s just not the greatest 1st stage booster propellant, and it’s the most difficult propellant to handle.  So to make that issue larger by making the booster all hydrolox, as well as just the upper stage?  I’m guessing there’s a reason that hadn’t been done before?
From my uneducated view I wonder about a few things.

a)   Why didn’t they stick with RS-27A, and widen up the core and go with more of them.  2, 3 or four depending on how powerful they wanted to go?  RS-27A is based on the H-1, which was supposed to be a pretty cheap and simple engine.  Was cost an issue with RS-27A?  It’d probably have to be a large issue, to warrent development of a new hydrolox engine, even one that had already had development done on it (as I understand the RS-68 had already had development on it).  Seems like a more incremental step would be increasing the Delta 2/3 2.4m diameter kerolox core with something like 3-3.5m  And then have the GEM-60’s developed and put a cluster of them on it as well?  Or just have like four RS-27A’s and go with something more F9 and Atlas V diameter.  3.6-3.8m?
b)   If they wanted to stick with just a single core engine, instead of multiple RS-27’s, why not have PWR update the F-1A, instead of the RS-68?  They are both kerolox gas generators.  With the 5m core that Delta IV got, they could mount a single F-1A on it, and probably just have a stretched DCSS to compensate for the lower ISP of the F-1A vs. RS-68.  I’d assume that updating the F-1A would be roughly the same price, or cheaper, than finishing development on the RS-68?  Or is that incorrect?  Just that single F-1A could boost a 4.5-5m core a stretched DCSS and have performance probably in the D4H range, but with no more actual hardware and costs than a Delta IV-medium.

I mean, I’d assume that most of the cost of an upper stage is in the systems, not the physical length of the cylindrical tanks.  The physical amount of alloy is a pretty small fraction of the cost I think.  If it needed two RL-10’s, that’d be a little more.  But RL-60’s were like 90% developed anyway.  That might be a good engine for such an LV. 
A 5m kerolox core would look a lot like Atlas Phase 2 in size.  Just with a single F-1A engine.  It would have the same footprint, so the launch facilities at LC-37 wouldn’t be any larger.  And the tri-core heavy version would probably not even be needed by USAF/DoD, given the capacity of it.  If a little more performance was needed, it could mount GEM-60’s just like the 5m Delta IV does anyway. 

Seems like you’d have an EELV-heavy launch vehicle, for the approximate costs of a single stick Delta IV-medium without SRB’s.  Seems like a more obvious way to go.

Or just go with the scaled up Delta 2/3, a little wider diameter core with like two RS-27A’s and a cluster of 9 GEM-60’s instead of 9 GEM-46’s. 

Seems like either would have been a more incremental evolution of the Delta rocket.
Instead they take quite a turn to go from 2.4m core to 5m core, hydrolox instead of kerolox.  From about a 25m long core to a 40m long core.  Form integral SRB’s to optional SRB’s.  Just wondering why.

And sorta thinking that if Delta IV had been a 5m wide LV with a single F-1A engine, that that might have been a building block LV that NASA may have evaluated more closely during ESAS.  It should have gotten Orion to LEO by itself.  And stretching the core and putting two F-1’s on it would make something like the Dynetics booster, and a tri-core version of that would have made a pretty good 1.5 launch architecture, with a lot of synergy with EELV’s. 
The tri-core would have just probably needed a new larger upper stage and/or EDS, which NASA could have made themselves.

Or…for more synergy, a 2-launch architecture with two tri-core heavies of the single engine cores.  That should have gotten something like 120-140mt to LEO I’d think.  It would have basically been Atlas Phase 2, but already developed an flying.  And with a storied US built engine.
 

[simonbp]

Delta IV is a direct descendant of NLS III.

The NLS program was built around the STME, which was a cost-reduced expendable version of SSME. The NLS I used an ET core with STMEs and two shuttle SRBs. NLS II was smaller version with ET tank and no SRBs and old Atlas-style stage and-a-half. NLS III was an even smaller version with one STME and 5.5 m tanks.

Boeing continued to develop the concept after NLS was canned and proposed it for the competition that became EELV, with a an ablative nozzle version of the STME called RS-68. When Boeing and McDonnell Douglas merged, the rocket became known as Delta IV.

EDIT: This explains the logic a bit:

http://books.google.com/books?id=eWbO9JJoAogC&lpg=PA258&ots=9_LV1L0-ID&dq=%22NLS%20III%22%20stme&pg=PA259#v=onepage&q=%22NLS%20III%22%20stme&f=false

[Jim]

Delta IV is a direct descendant of NLS III.

The NLS program was built around the STME, which was a cost-reduced expendable version of SSME. The NLS I used an ET core with STMEs and two shuttle SRBs. NLS II was smaller version with ET tank and no SRBs and old Atlas-style stage and-a-half. NLS III was an even smaller version with one STME and 5.5 m tanks.

Boeing continued to develop the concept after NLS was canned and proposed it for the competition that became EELV, with a an ablative nozzle version of the STME called RS-68. When Boeing and McDonnell Douglas merged, the rocket became known as Delta IV.

Correction, it was all McDonnell Douglas.  Heritage Boeing was never involved in Delta IV, it only became Boeing through the merger.  Boeing lost out of the EELV competition when it went from 4 to 2 contractors.  Boeing then developed Sealaunch to get in the launcher business.

There are photos of Boeing's EELV proposal on L2.
http://forum.nasaspaceflight.com/index.php?topic=29906.0

[Lobo]

Delta IV is a direct descendant of NLS III.

The NLS program was built around the STME, which was a cost-reduced expendable version of SSME. The NLS I used an ET core with STMEs and two shuttle SRBs. NLS II was smaller version with ET tank and no SRBs and old Atlas-style stage and-a-half. NLS III was an even smaller version with one STME and 5.5 m tanks.

Boeing continued to develop the concept after NLS was canned and proposed it for the competition that became EELV, with a an ablative nozzle version of the STME called RS-68. When Boeing and McDonnell Douglas merged, the rocket became known as Delta IV.

Correction, it was all McDonnell Douglas.  Heritage Boeing was never involved in Delta IV, it only became Boeing through the merger.  Boeing lost out of the EELV competition when it went from 4 to 2 contractors.  Boeing then developed Sealaunch to get in the launcher business.

There are photos of Boeing's EELV proposal on L2.
http://forum.nasaspaceflight.com/index.php?topic=29906.0

Very interesting Jim.  Thank you.

So Delta 4 was the result do MD work on NLS or something?

I'm still a little confused.  delta 2 was Boeing's design right?

So was Delta 4 completely a MD design with no relationship to Delta 2/3, but they just called it a Delta anyway?

[Galactic Penguin SST]

Delta IV is a direct descendant of NLS III.

The NLS program was built around the STME, which was a cost-reduced expendable version of SSME. The NLS I used an ET core with STMEs and two shuttle SRBs. NLS II was smaller version with ET tank and no SRBs and old Atlas-style stage and-a-half. NLS III was an even smaller version with one STME and 5.5 m tanks.

Boeing continued to develop the concept after NLS was canned and proposed it for the competition that became EELV, with a an ablative nozzle version of the STME called RS-68. When Boeing and McDonnell Douglas merged, the rocket became known as Delta IV.

Correction, it was all McDonnell Douglas.  Heritage Boeing was never involved in Delta IV, it only became Boeing through the merger.  Boeing lost out of the EELV competition when it went from 4 to 2 contractors.  Boeing then developed Sealaunch to get in the launcher business.

There are photos of Boeing's EELV proposal on L2.
http://forum.nasaspaceflight.com/index.php?topic=29906.0

Very interesting Jim.  Thank you.

So Delta 4 was the result do MD work on NLS or something?

I'm still a little confused.  delta 2 was Boeing's design right?

So was Delta 4 completely a MD design with no relationship to Delta 2/3, but they just called it a Delta anyway?

The whole Thor-Delta series from 1960 was done by Douglas, which became McD, then merged into Boeing in 1997. So your 2nd question is wrong, while the other two are correct. 

[Jim]

Delta IV is a direct descendant of NLS III.

The NLS program was built around the STME, which was a cost-reduced expendable version of SSME. The NLS I used an ET core with STMEs and two shuttle SRBs. NLS II was smaller version with ET tank and no SRBs and old Atlas-style stage and-a-half. NLS III was an even smaller version with one STME and 5.5 m tanks.

Boeing continued to develop the concept after NLS was canned and proposed it for the competition that became EELV, with a an ablative nozzle version of the STME called RS-68. When Boeing and McDonnell Douglas merged, the rocket became known as Delta IV.

Correction, it was all McDonnell Douglas.  Heritage Boeing was never involved in Delta IV, it only became Boeing through the merger.  Boeing lost out of the EELV competition when it went from 4 to 2 contractors.  Boeing then developed Sealaunch to get in the launcher business.

There are photos of Boeing's EELV proposal on L2.
http://forum.nasaspaceflight.com/index.php?topic=29906.0

Very interesting Jim.  Thank you.

So Delta 4 was the result do MD work on NLS or something?

I'm still a little confused.  delta 2 was Boeing's design right?

So was Delta 4 completely a MD design with no relationship to Delta 2/3, but they just called it a Delta anyway?

Boeing had nothing to do with Delta.  Delta was Thor heritage which was produce by Douglas pre 1968 and MD there after.  The 4m upperstage on Delta 3 became the 4m upperstage on Delta IV.  MD wanted to leverage the LH2 experience in the industry and the large diameter from its Titan IV fairing work and have common propellants for all stages for Delta IV

[catdlr]

This may help some: Rocket Genealogy

[Lobo]

Boeing had nothing to do with Delta.  Delta was Thor heritage which was produce by Douglas pre 1968 and MD there after.  The 4m upperstage on Delta 3 became the 4m upperstage on Delta IV.  MD wanted to leverage the LH2 experience in the industry and the large diameter from its Titan IV fairing work and have common propellants for all stages for Delta IV

Ok...so Douglas created Delta, the MD after that. 
So that still leaves my original question, why wasn't Delta 4 an evolved Delta 2/3 instead of such a different animal altogether?

MD wanted to leverage LH2 experience in the industry?  Weren't The complex reusable RS-25 and the upper stage only RL-10 the only LH2 engines in the US then?
Why wouldn't they want to leverage existing gas generator kerolox experience they already had with RS-27?   Or larger with the F-1A?
Or the RS-84 in development that was based on the new ORSC tech from the Russians?

It still really seems like an unusual left turn suddenly.  Especially since LH2 isn't a particularly great booster propellant compared to Kerolox...(or even solids)
I mean there was a reason that the S-1C was kerolox, instead of he Saturn V being all hydrolox and the shuttle had solid boosters rather than being all hydrolox. 
Ditto for Araine 5 and other rockets around when Delta 4 was designed.

[edkyle99]

a)Why didn’t they stick with RS-27A, and widen up the core and go with more of them.  2, 3 or four depending on how powerful they wanted to go? 

Given the limitations imposed by the original EELV requirements, McDonnell Douglas would have needed 6 to 8 of those H-1 type engines on each core and a bigger upper stage powered by more than one RL-10 engine.  A Heavy then would have ended up with 18 to 24 booster engines.  Those things weren't cheap.  Delta would have flat out lost the proposal to a staged combustion Atlas.

Strap-on solids would have made the task far easier, but weren't allowed.  General Dynamics showed how during the 1980s with its original "Atlas II/Centaur G-PRIME" proposal for CELV (won by Titan IV).  That proposal used five H-1D engines on a 200 inch core augmented by four 67 inch diameter solids motors (SRB-A class). 

McDonnell Douglas did contemplate a fatter Delta.  During the late 1980s or early 1990s it studied a 2xRS-27A core with 12 strap-on solid motors, topped by a bigger upper stage.  That led to Delta III, with more powerful solid motors taking the place of the dual core engine approach.

 - Ed Kyle

[simonbp]

Given the limitations imposed by the original EELV requirements, McDonnell Douglas would have needed 6 to 8 of those H-1 type engines on each core and a bigger upper stage powered by more than one RL-10 engine. 

Well that sounds familiar.

[Lobo]

Given the limitations imposed by the original EELV requirements, McDonnell Douglas would have needed 6 to 8 of those H-1 type engines on each core and a bigger upper stage powered by more than one RL-10 engine. 

Well that sounds familiar.

Heheheheh....nice point!

But I thought that H-1's were supposed to be relatively cheap?
Or could they have been more streamlined and automated if there'd been a relatively large production run of them?
I mean, aren't they sort of a larger Merlin 1?  The economics of needing nine of those on an EELV-class launcher seem to be working out so far for SpaceX.  ;-)

But, Merlin 1's were designed some 50 years after the H-1 and a modern production facility was put together for them, so perhaps the H-1's couldn't be made in a mass produced, cost effective way by PWR for McDonnell-Douglas?
I don't know, but I'd think they could, maybe?

[Lobo]

a)Why didn’t they stick with RS-27A, and widen up the core and go with more of them.  2, 3 or four depending on how powerful they wanted to go? 

Given the limitations imposed by the original EELV requirements, McDonnell Douglas would have needed 6 to 8 of those H-1 type engines on each core and a bigger upper stage powered by more than one RL-10 engine.  A Heavy then would have ended up with 18 to 24 booster engines.  Those things weren't cheap.  Delta would have flat out lost the proposal to a staged combustion Atlas.

Strap-on solids would have made the task far easier, but weren't allowed.  General Dynamics showed how during the 1980s with its original "Atlas II/Centaur G-PRIME" proposal for CELV (won by Titan IV).  That proposal used five H-1D engines on a 200 inch core augmented by four 67 inch diameter solids motors (SRB-A class). 

McDonnell Douglas did contemplate a fatter Delta.  During the late 1980s or early 1990s it studied a 2xRS-27A core with 12 strap-on solid motors, topped by a bigger upper stage.  That led to Delta III, with more powerful solid motors taking the place of the dual core engine approach.

 - Ed Kyle

I'm not at all familiar with the EELV requirements, but the basic Atlas V only has 860klbs thrust.  Four H-1B's wou'd have 820klbs, and five would have 1025klbs.  Why would MD need 6-8 H-1's?
That's be Saturn 1 class rather than Atlas V-401 or Delta IV-Medium class, which are the basic versions of those EELV's.

And the initial RS-68 only put out 668klbs of thrust.  Three H-1B would get you close to that.

And that's assuming the H-1 couldn't be tweaked to put out a little more thrust.

And it could have used SRB augmentation for more performance, because Atlas V and Delta IV both do.  So I'm assuming that was ok.

Now, the H-1's aren't as efficient as either the RD-180 or RS-68, but the Merlin 1's are GG kerolox as well, and they seem to work pretty good.  Falcon doesn't have great BLEO performance, but I sort of assumed that had more to do with the kerolox upper stage than the kerolox GG booster?
So why couldn't Delta IV have basically been a Falcon 9, but with five H-1 engines and a stretched Delta III upper stage?  And if an RL-10 wouldn't be enough, what about an RL-60?
Again, as I understand, the RL-60 was 90% developed anyway.
Was RL-10 a common engine prior to the EELV's, in that in was desirable to keep using it? 
I think it was only used on Centaur on certain Titan IV launches...right?
It wasn't quite the mainstay upper stage engine it's been since the EELV's started flying back then, was it?

So, I'm wondering why not a modernized H-1 engine, maybe boosting it's thrust to around 225klbs and put four of them on a 3.6m core (1Mlbs total) with a stretched Delta III upper stage with an RL-60 engine?
Wouldn't that have been as competative of a proposal as developing the RS-68, and more synergetic with the Delta family?
And have various optional GEM-60 SRB augmentation options as well?
Or a tri-core heavy?

[Jim]

MD wanted to leverage LH2 experience in the industry?  Weren't The complex reusable RS-25 and the upper stage only RL-10 the only LH2 engines in the US then?
Why wouldn't they want to leverage existing gas generator kerolox experience they already had with RS-27?   Or larger with the F-1A?
Or the RS-84 in development that was based on the new ORSC tech from the Russians?

It still really seems like an unusual left turn suddenly.  Especially since LH2 isn't a particularly great booster propellant compared to Kerolox...(or even solids

And most of the people in the engine industry had LH2 experience.  SSME and RL-10 had/were going through various upgrades.  There was only  sustaining engineering of the RS-27.

MD wanted a common propellant also

[Jim]

And it could have used SRB augmentation for more performance, because Atlas V and Delta IV both do.  So I'm assuming that was ok.

Not for the original USAF requirements.  Commercial requires drove the addition of solids.  Reliability and cost drove the single engine design.

[edkyle99]

I'm not at all familiar with the EELV requirements, but the basic Atlas V only has 860klbs thrust.  Four H-1B's wou'd have 820klbs, and five would have 1025klbs.  Why would MD need 6-8 H-1's?

Specific impulse differences.  The H-1 gas generator cycle was less efficient than the RD-180 staged combustion cycle.  Less efficient meant more propellant which meant more thrust.  The effect cascades.  The engines were 15% less efficient, but the rocket would have to weigh 35-40% more.

A bigger upper stage would allow for a smaller first stage and use of only six H-1 engines, but that bigger upper stage, which would have to carry twice as much propellant as Centaur, would need more RL10s (probably two for GTO and four for LEO missions).  Five H-1 engines simply can't lift enough to meet the EELV Medium GTO requirement, no matter how big the upper stage.  Unless, of course, McDonnell Douglas had added a third stage - a move adding more costs.

 - Ed Kyle

[Chris Bergin]

Terrible thread title resolved!

[Lobo]

Terrible thread title resolved!

Thanks Chris!

[Lobo]

MD wanted to leverage LH2 experience in the industry?  Weren't The complex reusable RS-25 and the upper stage only RL-10 the only LH2 engines in the US then?
Why wouldn't they want to leverage existing gas generator kerolox experience they already had with RS-27?   Or larger with the F-1A?
Or the RS-84 in development that was based on the new ORSC tech from the Russians?

It still really seems like an unusual left turn suddenly.  Especially since LH2 isn't a particularly great booster propellant compared to Kerolox...(or even solids

And most of the people in the engine industry had LH2 experience.  SSME and RL-10 had/were going through various upgrades.  There was only  sustaining engineering of the RS-27.

MD wanted a common propellant also

Hmmm...

Didn't most people in the industry also have RP-1 experience?

And I can understand a common propellant.  The Russians and SpaceX went that route too. 
But...take Atlas V.  You have three liquids.  LH2, LOX, and RP-1.

If you want to remove one to make things more simple and streamlined, it doesn't seem like removing the most simple and easy to pump, transport, and store really helps things much.  You'd think they'd remove the most difficult by far to work with...LH2.

I mean, I understand the higher performing LH2...but if you are going for simplicity, it seems a little counter intuitive to cut out the most easy liquid to work with and make it so you need even more of the most difficult liquid to work with.
Especially since a lot of the advantages of the ISP of LH2 isn't really realized until after the "booster" phase of ascent. 
Using LH2 to get off the pad is a little like trying to start a car on an uphill incline in 5th gear...is it not?

Better to start in a high torque low speed gear, no?

[Thorny]

It was a progression toward a simpler vehicle in theory. If you stick to Delta, you have...

Delta II... RP-1, LOX, Aerozine 50 and N204 (four liquids)
Delta III... RP-1, LOX, LH2 (three liquids)
Delta IV... LH2, LOX (two liquids)

[Hyperion5]

I'm not at all familiar with the EELV requirements, but the basic Atlas V only has 860klbs thrust.  Four H-1B's wou'd have 820klbs, and five would have 1025klbs.  Why would MD need 6-8 H-1's?

Specific impulse differences.  The H-1 gas generator cycle was less efficient than the RD-180 staged combustion cycle.  Less efficient meant more propellant which meant more thrust.  The effect cascades.  The engines were 15% less efficient, but the rocket would have to weigh 35-40% more.

A bigger upper stage would allow for a smaller first stage and use of only six H-1 engines, but that bigger upper stage, which would have to carry twice as much propellant as Centaur, would need more RL10s (probably two for GTO and four for LEO missions).  Five H-1 engines simply can't lift enough to meet the EELV Medium GTO requirement, no matter how big the upper stage.  Unless, of course, McDonnell Douglas had added a third stage - a move adding more costs.

 - Ed Kyle

So the ratio you're saying that if we want to launch the same amount of payload to orbit, the rough rule of thumb is that a 1% increase in booster engine Isp results in a 2.33-2.66% lighter rocket?  With regards to the Delta IV though, wouldn't the superior impulse density of the RD-180 make the Atlas V the more optimized LV? 

There's one other thing about the Delta IV that bothers me besides an all-hydrolox design.  It's the complete and utter lack of common bulkheads anywhere in the design.  I'm looking at the design overview diagram of the Ariane 5 ECA (http://en.wikipedia.org/wiki/File:Cut_drawing_of_an_Ariane_5_ECA_EN.svg), and it's clear that LV was mass-optimized.  That in turn allows Ariane to use less expensive engines for the same level of performance.  Why, if the Delta IV has to overcome the drawbacks of an all-hydrolox design, did they not follow Ariane's lead?  I've heard explanations that it was considered a 'bridge too far' and so on, but it seems like a wasted opportunity.  I'm curious, if the Delta IV did feature Ariane 5-style common bulkheads, just how much would that affect its performance and dry mass? 

[Jim]

Didn't most people in the industry also have RP-1 experience?

No, as I said, only sustaining engineering on RS-27.  There hadn't been a new RP-1 design for years.

[Jim]

And I can understand a common propellant.  The Russians and SpaceX went that route too. 
But...take Atlas V.  You have three liquids.  LH2, LOX, and RP-1.

That was already LM experience base and that was what they chose.  And they ended up with a non US RP-1 engine.  MD trades said to do otherwise.

[Jim]

1.  If you want to remove one to make things more simple and streamlined, it doesn't seem like removing the most simple and easy to pump, transport, and store really helps things much.  You'd think they'd remove the most difficult by far to work with...LH2.

2.  I mean, I understand the higher performing LH2...but if you are going for simplicity, it seems a little counter intuitive to cut out the most easy liquid to work with and make it so you need even more of the most difficult liquid to work with.

1.  LH2 was going to be needed on the upperstage to meet the USAF performance requirements for any vehicle.

2.  MD chose to simplify ground infrastructure by having one propellant combination.

[edkyle99]

So the ratio you're saying that if we want to launch the same amount of payload to orbit, the rough rule of thumb is that a 1% increase in booster engine Isp results in a 2.33-2.66% lighter rocket?  With regards to the Delta IV though, wouldn't the superior impulse density of the RD-180 make the Atlas V the more optimized LV?

My example was only for a hypothetical H-1 powered Delta IV versus Atlas V, and only for the two-stage "Medium" version, and depended on assumptions about the second stage.  It can't be a universal "rule of thumb".

There's one other thing about the Delta IV that bothers me besides an all-hydrolox design.  It's the complete and utter lack of common bulkheads anywhere in the design.

Common bulkheads cost more, and in the age of composites don't save as much weight as they once did.  Like Delta IV CBC, Atlas V CCB also uses separate bulkheads.  Centaur uses common bulkheads because its design dates from the 1960s, when NASA was trying to squeeze maximum payload from an Atlas that only produced 367,000 lbs of liftoff thrust. 

I'm looking at the design overview diagram of the Ariane 5 ECA and it's clear that LV was mass-optimized.  That in turn allows Ariane to use less expensive engines for the same level of performance.  Why, if the Delta IV has to overcome the drawbacks of an all-hydrolox design, did they not follow Ariane's lead?

First, USAF emphasized cost savings as a primary goal for EELV, so it is only natural that McDonnell Douglas would have followed that lead.  Second, and I'm thinking out loud here, Delta IV had to be designed to cover a lot more payload range than Ariane 5.  It had to have structural margin to handle the triple-barrel Heavy loads, etc.  Third, Ariane's core is naturally more mass sensitive than Delta IV CBC because the EPC (core) burns almost to orbit and it depends on the solids for the initial boost.  CBC is a traditional first stage.  The comparison isn't Vulcain 2 versus RS-68.  It is Vulcain 2 plus two EAP-E boosters versus RS-68.

 - Ed Kyle

[Lobo]

Didn't most people in the industry also have RP-1 experience?

No, as I said, only sustaining engineering on RS-27.  There hadn't been a new RP-1 design for years.

That's a fair point.  But what about RS-84?  Wasn't it in development while Delta 4 was?  Or was it later?

And could PWR modernize the F-1 as they might do now for Dynetics?
Not a new design, but a design they already had with several working units of the F-1 and F-1A.
Or was the deign that became the RS-68 farther along that a modernized F-1 would have been?

[Nick L.]

RS-84 was later I think - it was still being developed after the Delta IV's first flight.

[Proponent]

And the initial RS-68 only put out 668klbs of thrust.  Three H-1B would get you close to that.

This is a nitpick, but 205,000-lb-thrust H-1 engines would have to be H-1Cs or H-1Ds, both of which also had 200,000-lb-thrust versions.  The H-1Cs where the central four on the S-IB stage and the H-1Ds were the outer four.  Details can be found in the Rocketdyne document attached to this post.

[Prober]

What type of core design does the Delta IV have.  What LV can it be traced to?
 
Understand the Atlas V has some iso grid back to the Titan.  What of the Delta IV?
 
 

[Hyperion5]

So the ratio you're saying that if we want to launch the same amount of payload to orbit, the rough rule of thumb is that a 1% increase in booster engine Isp results in a 2.33-2.66% lighter rocket?  With regards to the Delta IV though, wouldn't the superior impulse density of the RD-180 make the Atlas V the more optimized LV?

My example was only for a hypothetical H-1 powered Delta IV versus Atlas V, and only for the two-stage "Medium" version, and depended on assumptions about the second stage.  It can't be a universal "rule of thumb".

Alright, I understand that, but your example does help to explain how methalox needs only a 7-8% increase in Isp over kerolox to compensate for lessened impulse density (~16% I believe).  There might not be a magic number ratio, but it does make it far easier to understand why t/w ratio often plays second fiddle to an engine's impulse density & Isp. 

There's one other thing about the Delta IV that bothers me besides an all-hydrolox design.  It's the complete and utter lack of common bulkheads anywhere in the design.

Common bulkheads cost more, and in the age of composites don't save as much weight as they once did.  Like Delta IV CBC, Atlas V CCB also uses separate bulkheads.  Centaur uses common bulkheads because its design dates from the 1960s, when NASA was trying to squeeze maximum payload from an Atlas that only produced 367,000 lbs of liftoff thrust. 

They might cost more to design, but they don't seem to have much effect on a rocket's crucial cost-per-kg to orbit that they need for commercial success (besides reliability).  The Ariane 5 is cheaper per kg than the EELVs and features them, as does the Falcon 9.  If Spacex can pull these off on the cheap chances are common bulkheads are not that expensive of a design feature. 

 I'm looking at the design overview diagram of the Ariane 5 ECA and it's clear that LV was mass-optimized.  That in turn allows Ariane to use less expensive engines for the same level of performance.  Why, if the Delta IV has to overcome the drawbacks of an all-hydrolox design, did they not follow Ariane's lead?

First, USAF emphasized cost savings as a primary goal for EELV, so it is only natural that McDonnell Douglas would have followed that lead.  Second, and I'm thinking out loud here, Delta IV had to be designed to cover a lot more payload range than Ariane 5.  It had to have structural margin to handle the triple-barrel Heavy loads, etc.  Third, Ariane's core is naturally more mass sensitive than Delta IV CBC because the EPC (core) burns almost to orbit and it depends on the solids for the initial boost.  CBC is a traditional first stage.  The comparison isn't Vulcain 2 versus RS-68.  It is Vulcain 2 plus two EAP-E boosters versus RS-68.

 - Ed Kyle

Though to be fair the Delta IV regularly uses SRBs, which while not as powerful as the Ariane's, certainly have performance characteristics (launch, handling, etc) that make the two LVs similar in certain layouts.  I would think the lack of man-rating for the Delta IV also plays a prominent part in design choices.  I've noticed the Delta IV Heavy has sometimes had its exhaust flames climb back up the rocket and set the insulation alight (temporarily).  I can't recall anything like this ever happening with the Ariane 5, which was meant to be man-rated.  If the Delta IV had been meant to be man-rated and the RS-68 were more like the Vulcain 2, would this insulation issue still be happening? 

[Jim]

The Ariane 5 is cheaper per kg than the EELVs and features them, as does the Falcon 9.  If Spacex can pull these off on the cheap chances are common bulkheads are not that expensive of a design feature. 

Ariane is subsidized.  Spacex has yet "to pull" this off.

[edkyle99]

Though to be fair the Delta IV regularly uses SRBs, which while not as powerful as the Ariane's, certainly have performance characteristics (launch, handling, etc) that make the two LVs similar in certain layouts.

Delta IV was designed from the outset to lift off without solids.  Ariane 5 requires the solids, and can't lift off without them.  That's the difference. 

I would think the lack of man-rating for the Delta IV also plays a prominent part in design choices.  I've noticed the Delta IV Heavy has sometimes had its exhaust flames climb back up the rocket and set the insulation alight (temporarily).  I can't recall anything like this ever happening with the Ariane 5, which was meant to be man-rated.  If the Delta IV had been meant to be man-rated and the RS-68 were more like the Vulcain 2, would this insulation issue still be happening? 

This is not a problem.  It is a design feature!  Delta IV is designed to "set itself on fire" at T-0!  This burns off residual hydrogen from the engine start process.  I believe that the lack of a water deluge system plays a role, along with the fuel-lead RS-68 startup.  The amount of excess hydrogen in these fireballs seems to have been reduced as launch experience increases. 

NASA sought to reduce the fireball for RS-68B, but it really isn't a crew safety issue because the crew already can't be outside the spacecraft at liftoff!  (The safety issue arises if the hydrogen does not ignite.  Shuttle had this same concern.)  The hydrogen burns off in a second or two.  Boeing did a static test at SLC 37B once.  The fireball quickly dissipated and I don't recall any residual flames after shutdown.

 - Ed Kyle

[Lobo]

NASA sought to reduce the fireball for RS-68B, but it really isn't a crew safety issue because the crew already can't be outside the spacecraft at liftoff!  (The safety issue arises if the hydrogen does not ignite.  Shuttle had this same concern.)  The hydrogen burns off in a second or two.  Boeing did a static test at SLC 37B once.  The fireball quickly dissipated and I don't recall any residual flames after shutdown.

 - Ed Kyle

Could D4/D4H utilize the STS style sparklers to burn off that GH2 prior to ignition?

[Jim]

Could D4/D4H utilize the STS style sparklers to burn off that GH2 prior to ignition?

It does