In Tuesday's post about the BE-4 engine explosion during Acceptance Testing, there was an item that ULA's Tory Bruno was going to do an update press conference today with more details on Vulcan Centaur. That conference was held today and reported by Eric Berger of Ars Technica. Bruno began by saying that the tank failed due to "higher-than-anticipated stress near the top of the liquid hydrogen propellant tank and slightly weaker welding."
The Centaur V upper stage was destroyed during pressure testing at Marshall Space Flight Center in Alabama on March 29. Bruno said this was the 15th test in a series of 45 tests to qualify the Centaur stage for all potential mission profiles. However, about halfway through the test the hydrogen tank started leaking, and over the course of four and a half minutes the leak expanded.
During this time hydrogen leaked into a confined area of the test stand, a relatively enclosed space. After it reached a flammable concentration and found an ignition source, the hydrogen caught fire. This seriously damaged the test stand as well as the hydrogen tank and liquid oxygen tank—loaded with liquid nitrogen for this test.
It was observed from the start of the failure analysis that the dome had failed, and by matching pieces of the shattered metal, they were able to deduce exactly where the failure started. Given that, by doing high-fidelity modeling of the stresses, the team found there were unexpectedly higher loads there. Next, the team analyzed the strength of the welds nearby and found they were not as high as previously stated. Bruno went on to say:
"The two things together, higher loads and somewhat lower strength in the weld, are what caused the crack to begin," he said. "The other thing I would ask you to appreciate, since we're being completely transparent here, is how we were already 15 tests in, which is considerably more testing and exposure to many more pressure cycles and lots and lots of more time with the structure sitting under pressure than would ever happen in any single flight."
Bruno then described the relatively straightforward fix. The area near the failure will be strengthened with an additional ring of stainless steel and strips of metal. These corrective actions will add about 140 kg to the mass of the upper stage, which seems to be a negligible 0.5% loss of payload capacity. The payload capacity is rated at 27 metric tons to low-Earth orbit; this change drops it to 26.86 metric tons.
Given how far along the test and qualification process this tank was before failure (15th of 45 tests), the new tank that's being built will be subjected to a handful of additional pressure tests to verify its predicted behavior. As has been reported, the upper stage for the first flight test, Cert-1, has been de-stacked and returned to the company's factory in Decatur, Alabama for similar modifications.
Bruno said performing the final qualification tests for the Centaur V anomaly and modifying the flight version of the tank are the final two steps needed before Vulcan can launch. He said he was pleased with the performance of the rocket's first stage during a recent "flight readiness firing" test, when the rocket's BE-4 engines ignited for a few seconds.
ULA is focused on getting the two certification flights completed so that they can start working through their enormous list of sold missions. Bruno said he anticipates a period of four to six weeks between the completion of the qualification tests and the launch of Vulcan's Cert-1 mission. The BE-4 that self destructed during its ATP on June 30 was scheduled to be built into the Cert-2 booster. Blue Origin's production line has other engines in process, partly for just such a problem, and it looks feasible that both certification missions could be concluded fairly early in '24, perhaps the end of March. Given that, it seems that DOD launches could begin by the summer.
We can't call it Elon Standard Time but considering any predicted launch time as being a bit on the optimistic side, especially the launch of a vehicle that has never flown, is prudent.
The orange cloud on the left is the anomaly that started this whole series of events, resulting in the modifications talked about here. Most of us would call that anomaly an explosion.
So... will ULA be able to trust the BE4s considering one blew up? Don't they have to wait to find out why it blew before they proceed?
ReplyDeleteAs to Centaur, sometimes shaving too much weight is a bad thing. It's one of the joys of Starship that loss of 0.5% is far less important on a 300 ton ship than on the smaller Vulcan.
Or, in spokesholespeak:
ReplyDelete"The engine over-performed, but unfortunately the tank was underbuilt."
In flight crew lingo:
"The engineers have now designed the tank to fail at a later mission waypoint."