Whats interesting is that if they switch to one engine to land then having the 3 engines in a row might make more sense so that one engine is dead center in the middle. That way the one engine does not need to gimbal off center.
Hmmmmm. Just realized we don't know if the tanker will also be updated to have three SL Raptors. I'd expect the answer to be yes if the main purpose is increased reliability and granularity. With deep throttle capabilities of 20%, three Raptors rated for 1700kN (380 klbf), and a dry mass no more than 85t (187393 lb),
the minimum thrust with three engines firing is 228 klbf. With two, it's 152 klbf. So a three engined BFS definitely could hover briefly with 5-10% of fuel retained for landing, especially with a dedicated a considerably lighter tanker version.
When an airliner runs out of fuel, it's one thing... But if you're 100 gallons short on a VTVL landing, engine redundancy is not going to be helpful...
Quote from: rsdavis9 on 10/17/2017 11:24 amWhats interesting is that if they switch to one engine to land then having the 3 engines in a row might make more sense so that one engine is dead center in the middle. That way the one engine does not need to gimbal off center.No, it would just make the engine-out redundancy much harder, because the 2 other engines would be much further away from the centerline.
If the 'hover-slam' is a solvable control problem with a single engine, it isn't much of a stretch to envision a 'hover-tip' maneuver with multiple engines firing off-center. The stack is tilted during descent to prevent XY acceleration, with Z velocity and now also deviation from vertical reaching 0 precisely at touchdown. It's obviously harder, with additional trades needed WRT landing gear, but not unsolvable.
Quote from: leetdan on 10/17/2017 06:07 pmIf the 'hover-slam' is a solvable control problem with a single engine, it isn't much of a stretch to envision a 'hover-tip' maneuver with multiple engines firing off-center. The stack is tilted during descent to prevent XY acceleration, with Z velocity and now also deviation from vertical reaching 0 precisely at touchdown. It's obviously harder, with additional trades needed WRT landing gear, but not unsolvable.Coslne losses suggest it might use more fuel than is absolutely necessary though, no?
Quote from: Lar on 10/17/2017 06:57 pmQuote from: leetdan on 10/17/2017 06:07 pmIf the 'hover-slam' is a solvable control problem with a single engine, it isn't much of a stretch to envision a 'hover-tip' maneuver with multiple engines firing off-center. The stack is tilted during descent to prevent XY acceleration, with Z velocity and now also deviation from vertical reaching 0 precisely at touchdown. It's obviously harder, with additional trades needed WRT landing gear, but not unsolvable.Coslne losses suggest it might use more fuel than is absolutely necessary though, no?The gimbal angle can't be all that big and cos(12 degrees) is 0.978 so the losses are a few percent or less.
Take a brand new BFS (ship A) to Mars, in 2022 or 2024. That's what, 150 days travel then it sits while being unloaded, habitats established and proven, after that, it returns to earth. That's what, another 150 days. The ship is now how old? At least a year old assuming establishing habitats is a priority, It could be much older. How far will SpaceX advance the design of the currently new BFS's while ship A is making this round trip? Or another way of looking at it is, "How useful is a year or more old Falcon 9 these days?" Or, "Has SpaceX ever built a rocket that didn't undergo major evolutionary changes over the span of a year's time?"In particular, we are addressing the first BFS's out of the box, not a mature, stable design as planned to exist by 2026.
Quote from: aero on 10/16/2017 01:58 amTake a brand new BFS (ship A) to Mars, in 2022 or 2024. That's what, 150 days travel then it sits while being unloaded, habitats established and proven, after that, it returns to earth. That's what, another 150 days. The ship is now how old? At least a year old assuming establishing habitats is a priority, It could be much older. How far will SpaceX advance the design of the currently new BFS's while ship A is making this round trip? Or another way of looking at it is, "How useful is a year or more old Falcon 9 these days?" Or, "Has SpaceX ever built a rocket that didn't undergo major evolutionary changes over the span of a year's time?"In particular, we are addressing the first BFS's out of the box, not a mature, stable design as planned to exist by 2026.I believe taking a look at the productuon runs of large transport category aircraft is instructive. The first few planes out of the factory are largely hand built, with tooling modified on the go and quality control standards yet to be developed. These aircraft are then added to the certification regimen and put through flight envelope testing. After a type certificate is issued by the FAA or other governmental agency, the test aircraft are refurbished and delivered to the launch customers.These initial articles are typically overweight and are susceptible to long-term chronic maintenance, yet they continue to fly for decades. When major upgrades or airworthiness directives are issued, all aircraft of the same type are upgraded to the new standard.It appears the business case for the BFR is predicated upon 1000 times reuse, meaning the development costs are spread between the number of ships built X the amount of flights they make. Counter to this would be a single flight to Mars and then becoming a museum piece in SITU. It would make much more fiscal sense to send the ship back to Earth for reuse. Even major upgrades to the engines or avionics would be far cheaper than scrapping the entire rocket. At least that is clearly true with atmospheric vehicles.
I got the impression that under nominal conditions all three landing engines will be firing. If there is an engine out the remaining two would throttle up to compensate vs only one running normally and another starting up in an engine out event. Startup would take way longer than just throttling up, which is very problematic when a failure occurs at the worst possible point in the landing. So with normal operation of all three engines running a triangular arrangement is symmetrical in all directions. Under engine out, the remaining two only need enough gimble to compensate for the failed engine. Worst case you provide enough gimble to compensate for two engines out. Since you can’t predict which engine will go out you want all three engines as close to center as possible, which the triangular arrangement also provides. Triangular is also the most centrally compact arrangement. My guess is a triangular arrangement for the landing engines.Caveat: I’m only an arm-chair rocket engineer so I might have a flaw in my logic. If so, please feel free to kindly point it out. Thx.
Quote from: acsawdey on 10/17/2017 08:15 pmQuote from: Lar on 10/17/2017 06:57 pmQuote from: leetdan on 10/17/2017 06:07 pmIf the 'hover-slam' is a solvable control problem with a single engine, it isn't much of a stretch to envision a 'hover-tip' maneuver with multiple engines firing off-center. The stack is tilted during descent to prevent XY acceleration, with Z velocity and now also deviation from vertical reaching 0 precisely at touchdown. It's obviously harder, with additional trades needed WRT landing gear, but not unsolvable.Coslne losses suggest it might use more fuel than is absolutely necessary though, no?The gimbal angle can't be all that big and cos(12 degrees) is 0.978 so the losses are a few percent or less.If you're landing on fumes, it does matter. but yeah.
Quote from: Lar on 10/17/2017 08:28 pmQuote from: acsawdey on 10/17/2017 08:15 pmQuote from: Lar on 10/17/2017 06:57 pmQuote from: leetdan on 10/17/2017 06:07 pmIf the 'hover-slam' is a solvable control problem with a single engine, it isn't much of a stretch to envision a 'hover-tip' maneuver with multiple engines firing off-center. The stack is tilted during descent to prevent XY acceleration, with Z velocity and now also deviation from vertical reaching 0 precisely at touchdown. It's obviously harder, with additional trades needed WRT landing gear, but not unsolvable.Coslne losses suggest it might use more fuel than is absolutely necessary though, no?The gimbal angle can't be all that big and cos(12 degrees) is 0.978 so the losses are a few percent or less.If you're landing on fumes, it does matter. but yeah.A fallacy.Other than for maneuvering X-Y the angle of thrust is center-lined on the CG. So there is no cosine losses on engine out unless for some reason the engines are gimbaling not in unison but gimbling in opposition.
It seems hardly worth the trouble but controlling the engines to gimbal in opposition would give an extra reduction in minimum thrust. The engines throttle down to 20% then if gimbaled in opposition by 12 degrees gives the vertical force of 19.56% of thrust. Does anyone know a number for the common maximum gimbal angle of rocket engines? Or is there even such a number outside of SpaceX? One half of one percent reduction in vertical force is likely less than vertical acceleration reduction due to the added mass needed to strengthen the engine gimbal supports.
Quote from: aero on 10/19/2017 03:59 amIt seems hardly worth the trouble but controlling the engines to gimbal in opposition would give an extra reduction in minimum thrust. The engines throttle down to 20% then if gimbaled in opposition by 12 degrees gives the vertical force of 19.56% of thrust. Does anyone know a number for the common maximum gimbal angle of rocket engines? Or is there even such a number outside of SpaceX? One half of one percent reduction in vertical force is likely less than vertical acceleration reduction due to the added mass needed to strengthen the engine gimbal supports.So a whole 0.44% gain? That should make it clear for you why they are NOT doing it. If you are running that close to the margin where that makes all the difference, you will not have a reliable system.
Quote from: Lars-J on 10/19/2017 04:51 amQuote from: aero on 10/19/2017 03:59 amIt seems hardly worth the trouble but controlling the engines to gimbal in opposition would give an extra reduction in minimum thrust. The engines throttle down to 20% then if gimbaled in opposition by 12 degrees gives the vertical force of 19.56% of thrust. Does anyone know a number for the common maximum gimbal angle of rocket engines? Or is there even such a number outside of SpaceX? One half of one percent reduction in vertical force is likely less than vertical acceleration reduction due to the added mass needed to strengthen the engine gimbal supports.So a whole 0.44% gain? That should make it clear for you why they are NOT doing it. If you are running that close to the margin where that makes all the difference, you will not have a reliable system.No, not 0.44% difference but 2.2% difference and 0.44 percentage point difference.