Why was Delta IV less successful than Atlas V?

[nicp]

It's pretty clear that Atlas V has been more successful than Delta IV, it has flown more than twice as many times (100 vs 45) and is not done yet - but it is out of production.
So why is A-V so much more successful?

An obvious reason is of course the use of hydrogen on D-IV - deeply cryogenic, metal embrittlement potential etc, which all adds to cost.
After some years it almost seemed that the Delta IV was irrelevent - except for the heavy variant. Atlas V could do the rest.

I'm curious if anyone has anything to add beyond 'hydrogen is a royal pain in the rear' for Delta IV's comparative lack of success.

[DanClemmensen]

It's pretty clear that Atlas V has been more successful than Delta IV, it has flown more than twice as many times (100 vs 45) and is not done yet - but it is out of production.
So why is A-V so much more successful?

An obvious reason is of course the use of hydrogen on D-IV - deeply cryogenic, metal embrittlement potential etc, which all adds to cost.
After some years it almost seemed that the Delta IV was irrelevent - except for the heavy variant. Atlas V could do the rest.

I'm curious if anyone has anything to add beyond 'hydrogen is a royal pain in the rear' for Delta IV's comparative lack of success.

I think it's better to exclude the Delta IV heavy from the comparison, since its 16 launches carried payloads too heavy for Atlas V. Thus, we have 29 Delta IV medium versus 100+ Atlas V.

One advantage of Atlas V is the ability to configure from zero to five SRBs. Thus, cost varies more finely with payload requirements. Delta IV had only 0,2, or 4 SRBs.

[dj_fan]

Atlas V used the crown jewel of the USSR Cold War - the RD-180.  Delta IV used RS-68A, which was the largest hydrogen-fueled rocket engine ever flown (LH2).  You could even probably find a Youtube video of Senator Bill Nelson extolling the benefits of the RD-180s in Senate testimony.

[Jim]

Delta IV had weight growth and RS-68 didn't meet target
The plain single core could not meet performance requirements and needed SRMs. 
To meet performance requirements, the core could not generic and was designed specifically to each configuration: single core, core with 2 SRMs, core with 4 SRMs, heavy core, heavy right booster and heavy left booster.  This drove up costs and reduced flexibility.   Regardless of performance, the concept of horizontal integration with tail service masts fixed at the pad meant that the side boosters of the heavy were mirror images of each other.  Atlas and Falcon had ways to work around this.

Atlas V cores were identical and could do any configuration including the heavy,

Delta IV also needed more hands on

[edzieba]

Atlas V cores were identical and could do any configuration including the heavy

I thought the triple-core configuration never made its way past the proposal stage, was there actual hardware (or vestiges) of that in some Atlas V cores?

[AmigaClone]

I suspect the cost might have been a factor in Atlas V launching more than three times that a Delta IV. The bulk of the price difference for a particular mission likely depended on the difference in cost of the respective first stages as well as the boosters used by each vehicle.

[nicp]

Atlas V cores were identical and could do any configuration including the heavy

I thought the triple-core configuration never made its way past the proposal stage, was there actual hardware (or vestiges) of that in some Atlas V cores?

I don't know - it does seem odd that Atlas V Heavy never got to fly, especially as Jim points out they could do this with little or no modification - something I didn't know.
Perhaps GSE updates, or ensuring two dissimilar systems... Again.

[Coastal Ron]

Atlas V cores were identical and could do any configuration including the heavy

I thought the triple-core configuration never made its way past the proposal stage, was there actual hardware (or vestiges) of that in some Atlas V cores?

I don't know - it does seem odd that Atlas V Heavy never got to fly, especially as Jim points out they could do this with little or no modification - something I didn't know.
Perhaps GSE updates, or ensuring two dissimilar systems... Again.

Delta IV Heavy already existed, and was certified for all USAF payloads, so why spend money on bringing Atlas V Heavy online?

The USAF would not spend the money to create and certify Altas V Heavy unless there was a payload they planned to build that would not fit on Delta IV Heavy, and apparently there never was such a payload.

Don't look at these types of situations from an engineering standpoint, you have to look at it from a money standpoint - who is paying for what, and why?

[ccdengr]

I thought the triple-core configuration never made its way past the proposal stage, was there actual hardware (or vestiges) of that in some Atlas V cores?

From the 2010 Atlas V User's Guide: https://web.archive.org/web/20130514051638/http://www.unitedlaunchalliance.com/site/docs/product_cards/guides/AtlasVUsersGuide2010.pdf

The Atlas V Heavy Lift Vehicle has been developed up to a Critical Design Review (CDR) level of
completeness. The completion of the design is currently on hold pending firm mission requirements for this
level of performance capability. At the time of this publication, the Atlas V HLV is approximately 30 months
from authority to proceed (ATP) to launch, but would require a 36-month integration cycle instead of the
typical 24-month integration shown for the Atlas V 400 and 500 series missions.

I don't know what would have happened if someone had ordered one.

[sdsds]

I'm going with engine choice, despite Jim's cogent point about how the DIV "Common" Booster Cores were each practically unique. For AV LM chose a reliable supplier of a high performance ORSC engine that met its performance specs from flight 1. Boeing chose a heritage supplier that didn't meet the basic spec requirement until RS-68A.

Because ... reasons ... USAF had chosen the DIVH over the AVH. NASA gave up on RS-68 and left USAF to pick up the tab for the RS-68A "upgrade."

AV could still be offered competitively today except for ... unexpected engine acceptability constraints.

[Blackstar]

I think it's better to exclude the Delta IV heavy from the comparison, since its 16 launches carried payloads too heavy for Atlas V. Thus, we have 29 Delta IV medium versus 100+ Atlas V.

I'm going to quibble over semantics (don't you just love it when people do that?):

The fact that the DIV Heavy could carry payloads that Atlas V could not was a capability, and therefore I would count it as a positive. By excluding those launches, you make DIV look even worse than Atlas V, but it's not fair to exclude it because it could do more than Atlas V.

It's now mostly forgotten now that we have Falcon 9 all the time, but there were a lot of decisions made about Atlas V and Delta IV that led to the programs that we got. There were a lot of assumptions as well that did not ultimately happen. For instance, there was an assumption that there would be a very vigorous commercial space sector that would require a lot of launches, and that the rockets would have high production rates, and that DoD would not get stuck with so much of the cost. I toured the Alabama production facility many years ago and it was huge, and built for a much bigger production run.

I guess I have a bit of a chip on my shoulder that so much of this history gets lost and ignored.

[DanClemmensen]

I think it's better to exclude the Delta IV heavy from the comparison, since its 16 launches carried payloads too heavy for Atlas V. Thus, we have 29 Delta IV medium versus 100+ Atlas V.

I'm going to quibble over semantics (don't you just love it when people do that?):

The fact that the DIV Heavy could carry payloads that Atlas V could not was a capability, and therefore I would count it as a positive. By excluding those launches, you make DIV look even worse than Atlas V, but it's not fair to exclude it because it could do more than Atlas V.

It's now mostly forgotten now that we have Falcon 9 all the time, but there were a lot of decisions made about Atlas V and Delta IV that led to the programs that we got. There were a lot of assumptions as well that did not ultimately happen. For instance, there was an assumption that there would be a very vigorous commercial space sector that would require a lot of launches, and that the rockets would have high production rates, and that DoD would not get stuck with so much of the cost. I toured the Alabama production facility many years ago and it was huge, and built for a much bigger production run.

I guess I have a bit of a chip on my shoulder that so much of this history gets lost and ignored.

I thought the partition was more useful, because DIVH was not in competition with AV and was therefore completely successful against AV, which contradicts the topic title. DIVH was an enabler for AV, because USAF forced 'all bidders" (i.e., ULA) to support all NSSL profiles. In theory ULA could not bid AV unless DIVH was available in its lineup. So yes it was absolutely a positive and critical ULA capability of the Delta IV.

[Blackstar]

In theory ULA could not bid AV unless DIVH was available in its lineup.

And that reveals something important--this was not really a competitive market, it was essentially DoD procurement, with a little bit of commercial on the side. DoD called the shots, and that determined the "success" of one vehicle over another.

I seem to remember--somebody who has more facts can step in--that the original plan was to downselect to a single launch vehicle, but the decision was made to keep both. That resulted in greater inefficiency, but more redundancy.

[tbellman]

I seem to remember--somebody who has more facts can step in--that the original plan was to downselect to a single launch vehicle, but the decision was made to keep both. That resulted in greater inefficiency, but more redundancy.

There was this tiny little industrial espionage affair that forced Boeing and Lockheed to create United Launch Alliance...

Boeing / McDonnell Douglas originally won the EELV procurement with the Delta IV rockets, leaving Lockheed with nothing.  And yes, there was some assumption going in to that procurement that the loser would survive anyway, due to expected commercial business (from e.g. the Iridium satellite constellation).  But then it turned out that the reason Boeing / McDonnell could win, was that they had had access to lots of internal Lockheed documentation.  DoD realized that, since the commercial market had failed to build up, just giving the contract to Lockheed instead would not give them the redundancy they wanted.  And then more or less forced the creation of ULA.

Here's an old article about the espionage case (written a few months before the formation of ULA):
The Seattle Times: Boeing probe intensifies over secret Lockheed papers

[sdsds]

There were also some court cases involved in the formation of ULA:

For several years before the ULA venture was
announced, Boeing and LM had engaged in bitter litigation involving
competition to provide launch services to the DOD. Lockheed Martin had
sued Boeing for alleged misconduct in competing for awards in the Air Force
Extended Expendable Launch Vehicle Program (EELV) and accused Boeing
of violations of the Racketeer Influenced and Corrupt Organizations Act, the
Florida Civil Remedies for Criminal Activities Act, the Sherman Act, and
the Florida Antitrust Act.56 In the same case Boeing filed a counterclaim
alleging that LM had engaged in unfair competition and tortious interference
with contractual relations and had violated the Lanham Act and the Florida
Unfair Deceptive and Trade Practices Act.57 The agreement to create ULA
stipulated that, upon the closing of the transaction, the companies would seek
an order to suspend their litigation in federal district court concerning the Air
Force EELV program.58

From page 11, https://scholarship.law.gwu.edu/cgi/viewcontent.cgi?article=2757&context=faculty_publications

More on the original question,

One advantage of Atlas V is the ability to configure from zero to five SRBs. Thus, cost varies more finely with payload requirements. Delta IV had only 0,2, or 4 SRBs.

Agreed. Atlas V executed on the dial-a-rocket concept really well. AV-411 was ... spectacular.
https://blog.ulalaunch.com/blog/meet-the-truly-unique-atlas-v-411-rocket

[Jim]

Atlas V cores were identical and could do any configuration including the heavy

I thought the triple-core configuration never made its way past the proposal stage, was there actual hardware (or vestiges) of that in some Atlas V cores?

It went to CDR.  Facilities were scarred for it.  Just needed to add propellant lines and tail service masts to the MLP.  Nose cone and booster attach hardware would just need to be manufactured and qualified.

[Proponent]

This is merely an impression, but ....

Somewhere there is a video of an Atlas V and a Delta IV and the trucks carrying them arriving at the Cape on the same barge. The trucks drive off. The truck carrying the Atlas looks like the sort of mass-produced vehicle you routinely see on a highways. The truck carrying the larger Delta is an over-sized machine that looks looks like it was built for some specialized application and very likely produced only in small numbers. I'm sure neither truck accounted for much of the cost of either rocket, but it gives the impression of the Delta having been designed as a military-style, cost-is-no-object vehicle.

[Jim]

The truck carrying the larger Delta is an over-sized machine that looks looks like it was built for some specialized application and very likely produced only in small numbers. I'm sure neither truck accounted for much of the cost of either rocket,

It didn't and it was used for multiple tasks like transporting payloads to the pad.  It also took the completed launch vehicle (single core or heavy) to the pad.   Atlas had a similar vehicle for only payloads.  And Trackmobiles for pushing the MLP to the pad.

So Delta had one vehicle to do everything and Atlas has 3 different ones.

but it gives the impression of the Delta having been designed as a military-style, cost-is-no-object vehicle.

Not at the beginning.

[Coastal Ron]

...
It's now mostly forgotten now that we have Falcon 9 all the time, but there were a lot of decisions made about Atlas V and Delta IV that led to the programs that we got. There were a lot of assumptions as well that did not ultimately happen. For instance, there was an assumption that there would be a very vigorous commercial space sector that would require a lot of launches, and that the rockets would have high production rates, and that DoD would not get stuck with so much of the cost. I toured the Alabama production facility many years ago and it was huge, and built for a much bigger production run.

IIRC the Delta production facility was sized for producing 40 Common Booster Core (CBC) per year, but the amount of demand never got close to that.

And yeah, the reason for why ULA came to be was, in part, due to a lack of demand for both Atlas V and Delta IV M/H. And between both Atlas V and Delta IV M/H, Atlas V was more popular with the USAF, which constituted the majority of the demand for ULA. Customer demand is a great indicator of how well your product is doing...

[hoku]

...
It's now mostly forgotten now that we have Falcon 9 all the time, but there were a lot of decisions made about Atlas V and Delta IV that led to the programs that we got. There were a lot of assumptions as well that did not ultimately happen. For instance, there was an assumption that there would be a very vigorous commercial space sector that would require a lot of launches, and that the rockets would have high production rates, and that DoD would not get stuck with so much of the cost. I toured the Alabama production facility many years ago and it was huge, and built for a much bigger production run.

<snip>
Atlas V was more popular with the USAF, which constituted the majority of the demand for ULA. Customer demand is a great indicator of how well your product is doing...

True, though with the retirement of Delta II ULA might have forced some of its customers to use an (overpowered) Atlas V. Else WorldView-3/4, Landsat 9, or NOAA-21 might have been launched by a Delta II.

[TheKutKu]

"EELV Program Background
Cost Perspective
Curt Khol"

A short PDF to remember the economic assumptions behind the EELV program.

[Jim]

Atlas V was more popular with the USAF, which constituted the majority of the demand for ULA. Customer demand is a great indicator of how well your product is doing...

No.  Delta IV was primarily a DOD vehicle. Only 7 of the first 20 Atlas V were DOD. 15 of the first 20 Delta IV were DOD.  Only 7 of the total 45 Delta IVs were non DOD.  56 of the 101 Atlas V were DOD.

[Jer]

Atlas V cores were identical and could do any configuration including the heavy

I thought the triple-core configuration never made its way past the proposal stage, was there actual hardware (or vestiges) of that in some Atlas V cores?

I don't know - it does seem odd that Atlas V Heavy never got to fly, especially as Jim points out they could do this with little or no modification - something I didn't know.
Perhaps GSE updates, or ensuring two dissimilar systems... Again.

It never flew because it would've been a deadly threat to Ares I.

[Proponent]

Delta IV Heavy was chosen over Atlas V Heavy in the 1990s, years before the Constellation program began.

[sdsds]

DIV-H was selected by the USAF for NSSL missions. DIV-H was selected by Lockheed-Martin for the first flight of an Orion capsule. AV-H was never selected by NASA because its capabilities were so similar to Ares I, and Ares I provided commonality with (pre-paid some development costs of) Ares V.

[Jim]

Atlas V cores were identical and could do any configuration including the heavy

I thought the triple-core configuration never made its way past the proposal stage, was there actual hardware (or vestiges) of that in some Atlas V cores?

I don't know - it does seem odd that Atlas V Heavy never got to fly, especially as Jim points out they could do this with little or no modification - something I didn't know.
Perhaps GSE updates, or ensuring two dissimilar systems... Again.

It never flew because it would've been a deadly threat to Ares I.

It was way before Ares I existed

[Jim]

DIV-H was selected by the USAF for NSSL missions. DIV-H was selected by Lockheed-Martin for the first flight of an Orion capsule. AV-H was never selected by NASA because its capabilities were so similar to Ares I, and Ares I provided commonality with (pre-paid some development costs of) Ares V.

AV-H wasn't wasn't there for NASA to select. 

[Jer]

DIV-H was selected by the USAF for NSSL missions. DIV-H was selected by Lockheed-Martin for the first flight of an Orion capsule. AV-H was never selected by NASA because its capabilities were so similar to Ares I, and Ares I provided commonality with (pre-paid some development costs of) Ares V.

AV-H wasn't wasn't there for NASA to select.

That's my point, that if AV-H was already flying in the mid 00s it would've put a target on the back of Ares I. There would've been no rhyme or reason for Ares I to exist.

A 5-meter core (using Delta IV tankage) version of the AV-H was presented to the Augustine Commission. ULA was even suggesting a possible 8.4m super heavy lift version. If Congress was gonna throw money at Boeing and Lockheed, they should've at least thrown on something good. Imagine SLS powered by a cluster of RD-180s..

[edzieba]

DIV-H was selected by the USAF for NSSL missions. DIV-H was selected by Lockheed-Martin for the first flight of an Orion capsule. AV-H was never selected by NASA because its capabilities were so similar to Ares I, and Ares I provided commonality with (pre-paid some development costs of) Ares V.

AV-H wasn't wasn't there for NASA to select.

That's my point, that if AV-H was already flying in the mid 00s it would've put a target on the back of Ares I. There would've been no rhyme or reason for Ares I to exist.

Atlas V 'heavy' had ceased to be a thing before Ares-I was conceived. Time travel would need to be involved for the two to have had any influence on each other.

[Jim]

A 5-meter core (using Delta IV tankage) version of the AV-H was presented to the Augustine Commission. ULA was even suggesting a possible 8.4m super heavy lift version.

Those were just rehashes of LM Atlas V Phase 2 studies.

[sdsds]

Atlas V 'heavy' had ceased to be a thing before Ares-I was conceived.

Is there a definitive date for when AV-H was removed from the list of ULA offerings? (It was offered for awhile even after USAF selected DIV-H for NSSF missions.)

[joek]

Is there a definitive date for when AV-H was removed from the list of ULA offerings? (It was offered for awhile even after USAF selected DIV-H for NSSF missions.)

It was "offered" only in the sense that LM would build it if someone payed for it (as other variants). Given that the only likely customer was the USAF EELV program, and that program was not going to pay for it, and that no non-USG/DoD demand existed, it died on the vine.

I would estimate the effective death certificate was probably around or shortly after the launch market crash in the early 2000's, which subsequently resulted in the formation of ULA. After which itt was a moot point.

[Jim]

Atlas V 'heavy' had ceased to be a thing before Ares-I was conceived.

Is there a definitive date for when AV-H was removed from the list of ULA offerings? (It was offered for awhile even after USAF selected DIV-H for NSSF missions.)

It never was part of ULA offerings.  When the USAF made the first buy of EELVs missions in 1998, Delta IV got 19 and Atlas got 9 and Atlas V given relief from completing the heavy.  It got relief from building a west coast pad, which was changed a few years later.

[sdsds]

The period of time in question is leading up to the November 2005 announcement of the NASA Exploration Systems Architecture Study. Prior to that Sean O'Keefe reportedly liked the idea of launching the CEV on DIV-H or AV-H.

Or so says Wikipedia, the encyclopedia that's sometimes even right.
https://en.wikipedia.org/wiki/Exploration_Systems_Architecture_Study#Shuttle_based_launch_system

[Jim]

The period of time in question is leading up to the November 2005 announcement of the NASA Exploration Systems Architecture Study. Prior to that Sean O'Keefe reportedly liked the idea of launching the CEV on DIV-H or AV-H.

NASA would have had to pay for pad and GSE completion and to take the design past CDR.

[Jer]

The period of time in question is leading up to the November 2005 announcement of the NASA Exploration Systems Architecture Study. Prior to that Sean O'Keefe reportedly liked the idea of launching the CEV on DIV-H or AV-H.

NASA would have had to pay for pad and GSE completion and to take the design past CDR.

Still would've been cheaper than the cost of the Ares I programme, right?

[Spiceman]

Or at least produce a workable rocket that could launch an Orion with all the goodies: not the castrated weight-trimmed variant we got a decade late.
Goodies like
a) land-landing
b) a service module with 2000 m/s + of delta-v that could, like Apollo, geting IN and OUT of low lunar orbit.

[spacenut]

Also, Russia gave the green light permission for the US to manufacture the RD-180.  The best of the old Atlas and Delta was IMO, the Atlas V phase II, a 5 or 5.5m Atlas with two RD-180's.  Then an Atlas V phase II heavy version which could have delivered around 700-75 tons to LEO.  The ACES upper stage would have had to be developed with probably 4 RL-10 engines. 

It never was to be because neither NASA nor the Air Force needed the phase II version.  NASA because by that time Constellation came out with Ares V and Ares I, and the Air Force didn't need anything bigger for their satellites.  However history could have taken a different turn and we would already be back to the moon with Atlas V phase II heavy launches.  Don't know how SpaceX would have evolved if that happened. 

In my opinion Delta IV cost more due to using hydrogen than the less expensive Atlas V's for most satellite launches. 

[Jim]

Or at least produce a workable rocket that could launch an Orion with all the goodies: not the castrated weight-trimmed variant we got a decade late.
Goodies like
a) land-landing
b) a service module with 2000 m/s + of delta-v that could, like Apollo, geting IN and OUT of low lunar orbit.

No, not really.  AV-H could only lift 65klb to LEO.   Orion on orbit mass is 58klb, this excludes the LAS and ESM panels.