I don’t really understand the mechanism of helium ingestion he’s talking about. Was the slosh so bad that helium bubbles made it all the way through the downcomer?
Angry Astronaut was wondering about the helium too, even speculating it might be a root cause for SN8 and SN9 failures.
If the slosh is that bad, and they can’t remediate, then the whole landing profile is out the window. However, I imagine there are other solutions than helium.
What I suspect they’ll end up doing is just adding better baffles to the header tanks, plus keeping two engines running, and perhaps using more control feedback on the throttles so that the engines would quickly compensate for any helium-induced drop in thrust.
Another possible major redesign would be to use cylindrical tanks with a piston to keep gas and liquids separate. Generally that’s not done because such a tank can’t have internal reinforcements, plus problems of a sliding piston at cryotemperatures, and extra weight and complexity.
You could, however, drive the piston with a big ball screw. I ran some numbers the other night and it would probably take about 300 HP to provide 3 or 4 atmospheres of pressure at the flow rate used in landings, so it’s within the range of what a Tesla motor could produce. But the piston would of course stop at the bottom of the header tank, so you’d still have to have a different method to keep the downcomer pressurized, otherwise it becomes unusable volume. The downcomer issue also afflicts other mechanical ways to pump fluid from the header tanks.
I kind of like the idea of using an electrically driven pressurization system for long duration missions, because the battery can be kept charged and they won’t be wedded to a supply of helium that they’re not going to have in situ. On the other hand, there would be far more critical failure modes, since anything from the battery to the piston or pump motor could fail.
But I’m still wondering about the cause of the excess flames running up the side of the rocket during the final descent, the extra flames that they had to put out after it landed, and the failure of the landing legs to lock in place. Sloshing wouldn’t explain any of that.
Didn’t fuel sloshing issues kill the Venture Star?
That was the least of the issues that killed VentureStar.
About Elon’s “Catch the Starship, too” tweet, I’ve been thinking about how you grab a big vehicle without putting large local loads on the structure.
The answer is, don’t concentrate the loads. If you can keep the main tanks pressurized to even just 10 psi at landing, the grabber can 3/4 encircle the stage and then inflate a bunch of thick-skinned balloons to just 1 psi. With millions of square inches of contact between the “hand” and the rigidly pressurized tanks, millions of pounds of friction will ensure that the vehicle is going nowhere, yet the alignment can be very loose and any delicate features on the outside such as thermal tiles and cable raceways would see forces smaller than the airloads imposed during flight. Softly, softly catchee monkey.
Even though the balloons would put some squeeze on the Starship, the pressure in the tanks is vastly greater than that- the tank walls would remain in tension, with no danger of buckling. It’s lot like trying to crush a sealed soda can in your hand- you’ll hurt your hand first.
This system could work for both stages, too, no need to put large loads on the SH grid fins.
I’ve had rocket grabber ideas since 1988 with Hummingbird. Nice to see someone with the resources to actually do it.
Correction- a mitt wrapping 3/4 around. and 50 feet along Starship’s tanks, would only have about half a million square inches of contact.
Give it *2* psi.
Elon speculated that perhaps the helium was being siphoned/wicked down via the baffle arrangement somehow.
“There were baffles, but one may have acted like a straw to suck bubbles in from above liquid/gas level
Something similiar happened on an early Falcon 1 flight, resulting in unexpectedly high liquid oxygen residuals at main engine cutoff.” — Elon Musk on Twitter.
With helium essentially a non condensable, non combustible gas it certainly would play havoc with pump and turbine. Continuous pressurization with methane to maintain NPSH as the bubble collapses from sloshing liquid is probably the answer. Combo COPV and auto-gen. Or shift the header to the nose like LOX. Don’t hear LOX header problems
I don’t really understand the mechanism of helium ingestion he’s talking about. Was the slosh so bad that helium bubbles made it all the way through the downcomer?
Angry Astronaut was wondering about the helium too, even speculating it might be a root cause for SN8 and SN9 failures.
If the slosh is that bad, and they can’t remediate, then the whole landing profile is out the window. However, I imagine there are other solutions than helium.
What I suspect they’ll end up doing is just adding better baffles to the header tanks, plus keeping two engines running, and perhaps using more control feedback on the throttles so that the engines would quickly compensate for any helium-induced drop in thrust.
Another possible major redesign would be to use cylindrical tanks with a piston to keep gas and liquids separate. Generally that’s not done because such a tank can’t have internal reinforcements, plus problems of a sliding piston at cryotemperatures, and extra weight and complexity.
You could, however, drive the piston with a big ball screw. I ran some numbers the other night and it would probably take about 300 HP to provide 3 or 4 atmospheres of pressure at the flow rate used in landings, so it’s within the range of what a Tesla motor could produce. But the piston would of course stop at the bottom of the header tank, so you’d still have to have a different method to keep the downcomer pressurized, otherwise it becomes unusable volume. The downcomer issue also afflicts other mechanical ways to pump fluid from the header tanks.
I kind of like the idea of using an electrically driven pressurization system for long duration missions, because the battery can be kept charged and they won’t be wedded to a supply of helium that they’re not going to have in situ. On the other hand, there would be far more critical failure modes, since anything from the battery to the piston or pump motor could fail.
But I’m still wondering about the cause of the excess flames running up the side of the rocket during the final descent, the extra flames that they had to put out after it landed, and the failure of the landing legs to lock in place. Sloshing wouldn’t explain any of that.
Didn’t fuel sloshing issues kill the Venture Star?
That was the least of the issues that killed VentureStar.
About Elon’s “Catch the Starship, too” tweet, I’ve been thinking about how you grab a big vehicle without putting large local loads on the structure.
The answer is, don’t concentrate the loads. If you can keep the main tanks pressurized to even just 10 psi at landing, the grabber can 3/4 encircle the stage and then inflate a bunch of thick-skinned balloons to just 1 psi. With millions of square inches of contact between the “hand” and the rigidly pressurized tanks, millions of pounds of friction will ensure that the vehicle is going nowhere, yet the alignment can be very loose and any delicate features on the outside such as thermal tiles and cable raceways would see forces smaller than the airloads imposed during flight. Softly, softly catchee monkey.
Even though the balloons would put some squeeze on the Starship, the pressure in the tanks is vastly greater than that- the tank walls would remain in tension, with no danger of buckling. It’s lot like trying to crush a sealed soda can in your hand- you’ll hurt your hand first.
This system could work for both stages, too, no need to put large loads on the SH grid fins.
I’ve had rocket grabber ideas since 1988 with Hummingbird. Nice to see someone with the resources to actually do it.
Correction- a mitt wrapping 3/4 around. and 50 feet along Starship’s tanks, would only have about half a million square inches of contact.
Give it *2* psi.
Elon speculated that perhaps the helium was being siphoned/wicked down via the baffle arrangement somehow.
“There were baffles, but one may have acted like a straw to suck bubbles in from above liquid/gas level
Something similiar happened on an early Falcon 1 flight, resulting in unexpectedly high liquid oxygen residuals at main engine cutoff.” — Elon Musk on Twitter.
With helium essentially a non condensable, non combustible gas it certainly would play havoc with pump and turbine. Continuous pressurization with methane to maintain NPSH as the bubble collapses from sloshing liquid is probably the answer. Combo COPV and auto-gen. Or shift the header to the nose like LOX. Don’t hear LOX header problems