United Launch Alliance CEO Tory Bruno announced Thursday night their planned No Earlier Than date for the first flight of Vulcan Centaur, while acknowledging that additional delays are always possible, Tory Bruno announced the date as...
May the 4th - usually said in Blog space along with the phrase, "be with you."
Bruno said the rocket's current "pacing item" for the debut launch is some final work qualifying the BE-4 rocket engines for flight. Blue Origin delivered two flight engines to ULA last fall, however, each of these machines had only undergone a fairly brief round of tests, known as acceptance testing. After this, two virtually identical BE-4 engines were sent from Blue Origin's factory in Washington to Texas. These "qual" engines have been undergoing a much more rigorous series of tests known as qualification testing.
The idea was to push the qualification engines through their paces, and beyond their expected flight environment, to find any flaws. During this series of tests, Bruno said, the oxygen pump on one of these engines has consistently produced about 5 percent more oxygen into the engine than expected. This fell outside the bounds of nominal performance but had only been observed in this engine.
Bruno said that after considerable investigation that they've concluded it's unit to unit variation, although it has only been seen in this one engine. None of the four BE-4 engines currently on Vulcan have shown this, nor have the other engines in the qual. test process. Bruno doesn't say how many man hours went into diagnosing that, how many other pumps have been tested, or any other details, but one out of a handful being an outlier in one metric compared to the others makes me wonder how different that 5% increase is. Are the others all closer to within 1% or closer to that 5%? How many standard deviations? Not that my question matters - it's their data. They expect to resume testing soon and expect the rest of the testing to consume six weeks.
With six weeks more testing in mind, it's conceivable that the launch date for the Certification-1 mission could be moved up to into mid-April, but the NET May 4 date seems good. The last piece I posted on the first flight for the Vulcan said an apparent "long pole for the tent" appeared to be the Astrobotic Peregrine lunar lander which has apparently still not had its engines installed. Bruno said he always has "backup plans for his backup plans" in case the payloads won't be ready by May, but he doesn't expect to need them.
Following this Cert.-1 mission, ULA has a Cert.-2 mission tentatively set for later this year which will lift the Sierra Space Dream Chaser spacecraft on a test flight. It's not certain at this point that Dream Chaser will be ready, either, but Bruno wants to fly a second mission by this fall so he can complete the military's certification process. One would assume his "backup plans for his backup plans" cover an alternative to Dream Chaser, although that isn't stated.
If they get that military certification, they can conduct Vulcan's first national defense mission for the US Space Force before the end of '23. ULA won the 2020 National Security Space Launch Phase 2 launch services procurement contract, specifying they would launch 60 percent of the US Space Force's payloads in the next few years while SpaceX would launch 40%. That 60% won't go anywhere until they achieve that military certification. Between those Space Force launches, Amazon's Kuiper Internet satellite constellation and using up the last of their Atlas Vs, ULA plans to be getting very busy.
"We have to ramp up," Bruno said. "Before the end of 2025 we expect to be really at a tempo, which is flying a couple of times a month, every two weeks."
25 launches/year would be unprecedented for ULA, even going back to its heydays of flying Atlas and Delta rocket variants.
As for ULA's talk about recovering its BE-4 engines for re-use, Bruno implied that was pretty low on his priority list and not being considered for years.
"In terms of our engine recovery, that is going to happen within a handful of years," Bruno said. "I don't want to say exactly when because it's part of the contract we have with one of our customers at this time, and we're not releasing the details of that. But it will take a couple of years to actually be reusing the engine."
Vulcan's core stage for the Cert.-1 mission being lifted into a processing facility at Cape Canaveral Space Force Station in January. Image credit United Launch Alliance.
A 5% pump volume output difference sounds huge, how could that be caused? 5% different RPM in the pump's rotor? 5% different volume in the pump's cavity? Distortion from bad slicing in the 3D printer? Rollers slipping in the printer used to print out the blueprint glued to the billet block in one of those third world metalworking videos, where the machinists are standing in a pile of chips in sandals?
ReplyDeleteAh, BO produces... variances. Interesting. A 5% variance will be catastrophic. Either using up fuel too quickly or causing a rapid disassembly of the affected piece, which will result in a loss of vehicle. That's not good. ULA is trusting BO way too much, but, really, BO has ULA backed up against the wall and there's not much ULA can do about it.
ReplyDeleteAs to not recovering the engines any time soon, well, SpaceX did the same, threw lots of engines away while perfecting them. We'll see what happens. And we'll see if ULA can afford to toss engines away.
Makes me wonder if the BE4s are actually reusable after all. It's all fine to do test stand starts and restarts, even the F1s went through a huge start, restart cycling and they were meant to be thrown away. But to make a 'reusable' engine that isn't really reusable? Is that where the BE4s are at now?
But to make a 'reusable' engine that isn't really reusable? Is that where the BE4s are at now?
DeletePersonally, I'm at the point of wondering if they're actually engines that can be used once. That 5% variance bothered me a lot. There's a lot I'd like to know, but I'm sure it's well-buried so that only their staff can get to it. How big is the sample? How much variation exists among those engines they're running qual test on?
Let's say there's eight engines, two Vulcan's worth, which I seriously doubt. What's the variance between them? Are seven engines within +/- 1% and this one at 5%? If the seven were within 4.9% instead of 1% that makes 5% a whole lot easier to live with. If 5% is 1 sigma from the mean, that's extremely different from 6 sigma. If it was my butt riding that engine, I'd sure prefer it to be closer to the mean.
Can that variance be accommodated by a control loop that "does something" to make up for 5% variance? It's a methalox engine, and we know from watching the earliest SpaceX Raptors that if the fuel mix isn't just right, they'll "burn green", burning away the metal in the combustion chamber away. Does 5% make that more likely or less?
With no way to answer those questions, we just have to wait and watch.