On what mission are you basing this statement? No actual HLV-oriented missions have been flown since Apollo!
I'm sorry; I should have clarified that I was including non-flown studies.
Primarily Constellation, although Apollo is still somewhat applicable to the discussion.
Then, each on-orbit docking and propellant transfer introduces more potential points of failure -- what if, for example, one vehicle crashes into another? You reliability goes down sharply with the number of operations involved.
It would be interesting to collate the data on the failure rates of orbital rendezvous and docking operations. I suspect (without any data to back up my assumption) that such operations are a good deal more reliable than launch vehicles. The reliability of docking procedures may go down if automated procedures are used, but gaining a great deal of experience with docking operations would be quite advantageous, considering that rendezvous and docking could be critically important to crew safety elsewhere in exploration programs.
Further, there is another problem. If a HLV launch fails, the payload will crash into the ocean, and the mission is simply written off. But if launch #5 of a 5-launch campaign fails, then what? You're left with 80 tons of junk in Low Earth Orbit. Have a backup launcher ready? Extra costs. Deorbit controllably? You must design for such facility from the beginning.
I can't see how having the remaining hardware intact and in orbit is a
negative attribute. Obviously that hardware has a limited lifetime, most likely defined by the propellant boiloff rate, but given a low enough boiloff rate, a fast enough turnaround time to launch a replacement component, and/or the ability to top off propellant on-orbit, the mission could be salvaged. It may be challenging to design for such mission recoverability, but it's less challenging than designing a payload to withstand a range safety detonation.
Having an extra launcher on hand to launch a backup component isn't that costly if said launcher is set to launch the next payload in line anyway. Deorbiting things from LEO is relatively easy, and such a spacecraft stack ought to have some form of propulsion onboard that would be capable of performing a deorbit burn if necessary.
Do you know that launch vehicles have volume constraints in addition to size constraints?
Yes. Launch vehicles already make use of hammerhead fairings to accommodate high-volume payloads. Wider fairings could potentially developed to carry even wider payloads. From the
Atlas 5 User's Guide, page 263;
Payload fairings as large as 7.2m (283 in.) in diameter and up to 32.3m (106 ft) in
length have been considered.
Such a fairing could easily enclose an S-IVB. Further discussion of such a fairing is found on page 349, where it is stated that this fairing configuration is limited to flying on configurations with up to 4 boosters (though the Heavy variant is not discussed). Larger hammerheads are, at the very least, a considerable possibility.
In addition, the wider the vehicle's body, the wider, theoretically, a hammerhead fairing can be. The Delta IV already sports a 5 meter fairing without a hammerhead; applying a similar ratio to that of the current Atlas (about 1.4) gives a 7 meter size, and applying a similar ratio to that of the 7.2 meter fairing would give a fairing in excess of 9 meters in diameter. Note that this is merely a simple extrapolation; such a fairing size may or may not be possible, depending on various factors.
A medium or heavy lift launch vehicle may be wider still; the Saturn IB was 6.61 meters in diameter and required no hammerhead to carry an S-IVB stage. No vehicles with a similar diameter are currently operational or in development, however. The Atlas V phase 2 evolution would utilise a 5-meter core using Delta tooling, as well as a 7.2 meter fairing and variants with payload capacity from 9 to 75 metric tons;
Except nobody has such skill at the moment. The question is if developing this skill will cost less than going the HLV route.
It probably would be. Whether the cost of a program utilising EELVs with propellant storage and transfer would be less than that of the HLV route is less certain.
Besides, can you name an obvious commercial or military application of this technology?
In-flight refuelling or limited in-flight repair of satellites would be pretty useful. Granted, satellites use hypergols rather than cryogens, but some of the developments for cryogenic propellants may be useful there as well, as would some of the technology developed for mitigating cryogen boiloff.
Beyond that, a space tug, or in-flight refuelling for upper stages might be beneficial.
Regarding unmanned exploration, launch costs are not the primary cost driver there. A Discovery class mission has a budget of $1B, out of which the LV costs $0.2B -- i.e. 20%. Most of the budget is the development of unique mission hardware and keeping the scientific staff on payroll for several years.
The cost reduction may not be revolutionary but it'd still be present. There's a benefit in different programs sharing costs.
