Quote from: whitelancer64 on 09/28/2022 08:22 pmQuote from: Kaputnik on 09/28/2022 08:01 pm*snip*And today we can develop engines faster and cheaper than before. In the long run it could be a better choice.I don't know if that's true. 6-10 years from concept to first flight attempt has been approximately how long it takes to develop a rocket engine since the 1960s. I think the perception that we can develop engines faster comes from all the small launch companies. I think the reality is that we've always developed smaller engines faster.
Quote from: Kaputnik on 09/28/2022 08:01 pm*snip*And today we can develop engines faster and cheaper than before. In the long run it could be a better choice.I don't know if that's true. 6-10 years from concept to first flight attempt has been approximately how long it takes to develop a rocket engine since the 1960s.
*snip*And today we can develop engines faster and cheaper than before. In the long run it could be a better choice.
The engines have such different roles that it is a bit of a compromise to force one engine to do everything.
Quote from: Kaputnik on 09/28/2022 08:01 pmThe engines have such different roles that it is a bit of a compromise to force one engine to do everything.Actually, during the Investor Day presentation, Beck seemed to be saying that the requirements for a second-stage engine and a first-stage engine capable of propulsive landing are similar. E.g., if you've given your first-stage engines enough throttle range that you can land the stage by lighting up one of the nine at minimum throttle, that's about what you'd need for the second-stage engine anyway.
But fundamentally one engine design is being reused and the other is being expended. In the emerging launch market the disposable upper stage cost is going to be a huge driver for Neutron's costs, and it makes sense for its engine to be as cheap as possible. A constraint which does not apply to the first stage.
Quote from: Kaputnik on 09/28/2022 10:18 pmBut fundamentally one engine design is being reused and the other is being expended. In the emerging launch market the disposable upper stage cost is going to be a huge driver for Neutron's costs, and it makes sense for its engine to be as cheap as possible. A constraint which does not apply to the first stage.Wait a sec, there is an expendable upper stage on Neutron?
Quote from: matthewkantar on 09/29/2022 04:08 amQuote from: Kaputnik on 09/28/2022 10:18 pmBut fundamentally one engine design is being reused and the other is being expended. In the emerging launch market the disposable upper stage cost is going to be a huge driver for Neutron's costs, and it makes sense for its engine to be as cheap as possible. A constraint which does not apply to the first stage.Wait a sec, there is an expendable upper stage on Neutron?That was RL plan from day one but they are not discounting reuseable US. The problem with reuseable US in this LV class high payload penalty, suddenly a 13t LV becomes 10t at a guess. If $kg is $2k then that is $6M of lost performance which could pay for expendable US that doesn't need refurbishing.
Of course, not every payload uses the full capacity of the rocket. And rockets are priced (for customers) per launch, not per kg -- you don't get a discount for using less than full capacity.
Quote from: trimeta on 09/29/2022 06:31 amOf course, not every payload uses the full capacity of the rocket. And rockets are priced (for customers) per launch, not per kg -- you don't get a discount for using less than full capacity. Not always true. SpaceX charged NASA $50.3M to launch IXPE, $11.7M less than the standard price of $62M.https://www.nasa.gov/press-release/nasa-awards-launch-services-contract-for-groundbreaking-astrophysics-mission/
Quote from: Steven Pietrobon on 09/29/2022 06:40 amQuote from: trimeta on 09/29/2022 06:31 amOf course, not every payload uses the full capacity of the rocket. And rockets are priced (for customers) per launch, not per kg -- you don't get a discount for using less than full capacity. Not always true. SpaceX charged NASA $50.3M to launch IXPE, $11.7M less than the standard price of $62M.https://www.nasa.gov/press-release/nasa-awards-launch-services-contract-for-groundbreaking-astrophysics-mission/People often cite that example, but that feels a lot more like "specifically trying to undercut (or at least, go head-to-head with) Pegasus XL" than "a trend of SpaceX offering a discount for reduced payloads." And before you mention SpaceX's bid to use Starship for NASA's TROPICS missions, that one was basically "let us carry an experimental payload on Starship's first launch": they needed to launch Starship anyway, might as well see if someone would pay them for it. Again, not representative.
Quote from: JEF_300 on 09/28/2022 08:24 pmQuote from: whitelancer64 on 09/28/2022 08:22 pmQuote from: Kaputnik on 09/28/2022 08:01 pm*snip*And today we can develop engines faster and cheaper than before. In the long run it could be a better choice.I don't know if that's true. 6-10 years from concept to first flight attempt has been approximately how long it takes to develop a rocket engine since the 1960s. I think the perception that we can develop engines faster comes from all the small launch companies. I think the reality is that we've always developed smaller engines faster.Smaller engines somewhat faster, larger engines take longer, yeah. As examples, both the RL-10 and the J-2 took about 6 years from start of development to first launch. The F-1 took about 10 years. The SpaceX Merlin 1A was 6 years. Rocket Lab's Rutherford took 4 years, as an example of a smaller engine taking less time.
Quote from: sevenperforce on 09/27/2022 03:41 pmSlight modification to my earlier pixel counting approach, now with estimated tank volumes.The volumetric split on the upper stage is assumed to be equal to that on the booster, and all spherical caps are assumed to be perfect spherical caps.I hope you're accounting that the curvature adius of the caps along the vehicle axis is 373/2 while sideways it's 617/2I.e. I hope you're assuming the caps are actually not spherical but ellipsoidal, based on a an rotationally symmetric ellipsoid with the short diameter 373cm and both identical long diameters of 617cm.The values are based off your pixel counts.
Slight modification to my earlier pixel counting approach, now with estimated tank volumes.The volumetric split on the upper stage is assumed to be equal to that on the booster, and all spherical caps are assumed to be perfect spherical caps.
Unless they want to develop two totally different thrust class engines no easy way to avoid F9 and Electron's well proven 9+1 combination. The other alternative is engine that can throttle down to 20-30% but that is big ask for large booster engine.
Quote from: sebk on 09/28/2022 10:53 amQuote from: sevenperforce on 09/27/2022 03:41 pmSlight modification to my earlier pixel counting approach, now with estimated tank volumes.The volumetric split on the upper stage is assumed to be equal to that on the booster, and all spherical caps are assumed to be perfect spherical caps.I hope you're accounting that the curvature adius of the caps along the vehicle axis is 373/2 while sideways it's 617/2I.e. I hope you're assuming the caps are actually not spherical but ellipsoidal, based on a an rotationally symmetric ellipsoid with the short diameter 373cm and both identical long diameters of 617cm.The values are based off your pixel counts.How sure are you that the caps are ellipsoidal rather than spherical? There appears to be a pretty sharp angle at the seam between the cap and the ring section.But yes, if they are ellipsoidal caps then my tank volumes will be slightly low.Quote from: TrevorMonty on 09/28/2022 02:25 amUnless they want to develop two totally different thrust class engines no easy way to avoid F9 and Electron's well proven 9+1 combination. The other alternative is engine that can throttle down to 20-30% but that is big ask for large booster engine.Even if you do a landing engine that can throttle down to 20-30%, it's still going to be overweight for your upper stage.The real alternative, if you're willing to dev two different thrust class engines, is to use 1-2 large engines on your booster and 1-2 smaller engines on the upper stage, with 1-2 upper-stage engines acting as first-stage verniers on ascent and as landing engines. For example, SpaceX could have fielded a "Falcon 4" with a booster powered by two Merlin 1Cs and two sea-level optimized Kestrels on the booster and an upper stage powered by two Vacuum-optimized Kestrels. Or if ULA wanted to make a reusable, stretched Delta IV (which is obviously dumb but for other reasons) they could upgrade the DCSS to use two RL-10s and also attach a pair of RL-10A-5s to the sides of the booster to provide extra umph on ascent and landing thrust on landing.
Quote from: sevenperforce on 09/30/2022 12:35 amQuote from: sebk on 09/28/2022 10:53 amQuote from: sevenperforce on 09/27/2022 03:41 pmSlight modification to my earlier pixel counting approach, now with estimated tank volumes.The volumetric split on the upper stage is assumed to be equal to that on the booster, and all spherical caps are assumed to be perfect spherical caps.I hope you're accounting that the curvature adius of the caps along the vehicle axis is 373/2 while sideways it's 617/2I.e. I hope you're assuming the caps are actually not spherical but ellipsoidal, based on a an rotationally symmetric ellipsoid with the short diameter 373cm and both identical long diameters of 617cm.The values are based off your pixel counts.How sure are you that the caps are ellipsoidal rather than spherical? There appears to be a pretty sharp angle at the seam between the cap and the ring section.But yes, if they are ellipsoidal caps then my tank volumes will be slightly low.Quote from: TrevorMonty on 09/28/2022 02:25 amUnless they want to develop two totally different thrust class engines no easy way to avoid F9 and Electron's well proven 9+1 combination. The other alternative is engine that can throttle down to 20-30% but that is big ask for large booster engine.Even if you do a landing engine that can throttle down to 20-30%, it's still going to be overweight for your upper stage.The real alternative, if you're willing to dev two different thrust class engines, is to use 1-2 large engines on your booster and 1-2 smaller engines on the upper stage, with 1-2 upper-stage engines acting as first-stage verniers on ascent and as landing engines. For example, SpaceX could have fielded a "Falcon 4" with a booster powered by two Merlin 1Cs and two sea-level optimized Kestrels on the booster and an upper stage powered by two Vacuum-optimized Kestrels. Or if ULA wanted to make a reusable, stretched Delta IV (which is obviously dumb but for other reasons) they could upgrade the DCSS to use two RL-10s and also attach a pair of RL-10A-5s to the sides of the booster to provide extra umph on ascent and landing thrust on landing.A short high G deceleration burn uses less fuel than long low G burn, which means more powerful landing engine the better to certain extent.
Quote from: sebk on 09/28/2022 10:53 amI hope you're accounting that the curvature adius of the caps along the vehicle axis is 373/2 while sideways it's 617/2I.e. I hope you're assuming the caps are actually not spherical but ellipsoidal, based on a an rotationally symmetric ellipsoid with the short diameter 373cm and both identical long diameters of 617cm.The values are based off your pixel counts.How sure are you that the caps are ellipsoidal rather than spherical? There appears to be a pretty sharp angle at the seam between the cap and the ring section.But yes, if they are ellipsoidal caps then my tank volumes will be slightly low.
I hope you're accounting that the curvature adius of the caps along the vehicle axis is 373/2 while sideways it's 617/2I.e. I hope you're assuming the caps are actually not spherical but ellipsoidal, based on a an rotationally symmetric ellipsoid with the short diameter 373cm and both identical long diameters of 617cm.The values are based off your pixel counts.
Quote from: sevenperforce on 09/30/2022 12:35 amEven if you do a landing engine that can throttle down to 20-30%, it's still going to be overweight for your upper stage.The real alternative, if you're willing to dev two different thrust class engines, is to use 1-2 large engines on your booster and 1-2 smaller engines on the upper stage, with 1-2 upper-stage engines acting as first-stage verniers on ascent and as landing engines. For example, SpaceX could have fielded a "Falcon 4" with a booster powered by two Merlin 1Cs and two sea-level optimized Kestrels on the booster and an upper stage powered by two Vacuum-optimized Kestrels. Or if ULA wanted to make a reusable, stretched Delta IV (which is obviously dumb but for other reasons) they could upgrade the DCSS to use two RL-10s and also attach a pair of RL-10A-5s to the sides of the booster to provide extra umph on ascent and landing thrust on landing.A short high G deceleration burn uses less fuel than long low G burn, which means more powerful landing engine the better to certain extent.
Even if you do a landing engine that can throttle down to 20-30%, it's still going to be overweight for your upper stage.The real alternative, if you're willing to dev two different thrust class engines, is to use 1-2 large engines on your booster and 1-2 smaller engines on the upper stage, with 1-2 upper-stage engines acting as first-stage verniers on ascent and as landing engines. For example, SpaceX could have fielded a "Falcon 4" with a booster powered by two Merlin 1Cs and two sea-level optimized Kestrels on the booster and an upper stage powered by two Vacuum-optimized Kestrels. Or if ULA wanted to make a reusable, stretched Delta IV (which is obviously dumb but for other reasons) they could upgrade the DCSS to use two RL-10s and also attach a pair of RL-10A-5s to the sides of the booster to provide extra umph on ascent and landing thrust on landing.