sevenperforce

I don't use MechJeb, so there's always some variation in dV budget for LKO ascent. So I typically just overengineer my first stage comfortably, and do the math for all the other stages. QuantumG's dV calculator is hugely useful, unless you're just dying to do the calculations by hand. (QuantumG is one of the regular posters over at the NSF forums.) And even if you do it partly by hand, no one calculates logarithms manually. If I have a lot of dV calcs to do, I write the formulas out in Excel...this is particularly good for optimization problems where I have to iterate over a range.

I did everyone's favorite mission -- Apollo 11. Plot twist: it was my first-ever flight in full KSP. And while this wasn't a historic launch, but rather a planned launch, here's X-37B on Falcon 9 FT, planned for August:

Working on an entry -- I went through the whole process and then all my kerbals burned up, so I had to rebuild my ejection seats from scratch to provide heat shielding.

The SLS core+boosters have plenty of margin to take 70 tonnes to a circular LEO. But since the SLS core can't restart to deorbit itself, it's much more efficient to take the 70-tonne-stack of ICPS+Orion+SM into an eccentric orbit with a 0-km Pe; this way the core can burn to depletion without wasting anything on circularization and deorbit naturally. The ICPS can raise Pe from the high Ap. I originally brought up the whole issue because I mistakenly thought the ICPS was going to be used for LEO payloads in addition to EM-1. This is, of course, incorrect; the ICPS is useless for LEO payloads because its TWR is so low.

I hate to say it, but if it goes this route, I imagine it being similar to how slave ownership was handled. For a long time, killing a slave was considered an offense against property, not an offense against life.

Intentional destruction of a sentient being.... I can see it making sense if (for example) the AIs we were finally able to make could only exist as persistent routines inside a single processing unit.

At least the fairing is large enough to send the DSG to cislunar orbit.

It's an optimization problem. You can treat each third stage as part of the payload, so that the second stage places your entire payload stack into an almost-orbit, all four payloads plus their stages are released, and each one circularizes and then does its own transfer to the desired orbit. On the other hand, if you can make propellant transfer work, then you can use a larger second stage and a single, small third stage with a heat shield. The second stage establishes a parking orbit with significant residual propellant, and then the third stage takes the first payload to its desired orbit and then does a short deorbit burn at apogee to aerobrake back into LEO. Rendezvous, refuel from the second stage, and repeat. Which option is more efficient depends on a lot of factors.

The US is awful in that regard. Only a tiny fraction of jobs in the United States offer any paid vacation, and those that do typically offer less than the average in every other developed country.

I actually figured this out. The ICPS is considered to be part of the 70-tonne LEO payload.

A hybrid rocket addresses both of these problems, actually. Hybrids are throttleable because you can control the flow of the oxidizer. And since hybrids are pressure-fed by definition, they are already going to be built much stronger than the liquid stages we are typically used to, with plenty of margin. I'm thinking a 4-core arrangement, with three strap-on boosters around a central core and the payload mounted on top. Each of the strap-on boosters would be fitted with a single tailfin for passive aerodynamic stability and roll cancellation on ascent, and they could be differentially throttled for yaw and pitch control. All would ignite and fire at full throttle on the pad, but the core would throttle down shortly before Max-Q and remain throttled down until booster burnout (similar to a Delta IV Heavy). At booster burnout, they would separate and the core would be throttled back up. The payload could be fitted with a simple HTP monoprop RCS system, providing guidance for the core during the terminal portion of the burn. It would also probably make sense to give the payload a small COTS solid-fueled kick stage for final circularization. Hybrid casings have to be strong enough to hold combustion pressures, so they are plenty strong enough to survive being chuted down. Since all four cores are essentially identical, this makes the testing process much simpler since they would be readily reusable. One cool addition would be an ablative nozzle that reshapes over the course of the burn. It could be machined out of something as simple as ordinary wood (cork, pine, or oak) and allow the nozzle expansion ratio to increase to compensate for change in pressure and thus maximize specific impulse. Solid oxidizers have much lower specific impulse than liquid oxidizers, so that's one of the issues there. That's the biggest reason why solid-fueled rockets have notoriously low specific impulse: solid oxidizers have high density and low specific energy. And the solid oxidizers that do exist do not vaporize well at all, in comparison to hydrocarbon-based solid fuels which vaporize, mix, and burn readily.

Right, this too. I can do enough programming to get by, but my degree is not in computer science and I only know a couple of programming languages. I can do more than enough math by hand to make decent estimates for amateur forum discussions, but I would need to be an expert programmer to set up and run the launch simulations that SpaceX and Blue Origin rely on.

Here's an idea that could make the space tourism thing a little more attractive. What about a lunar cycler? A big fancy space hotel with lots of room for labs and experiments, all on a regular Earth-Moon loop. It raises the required dV for launches, since the LV needs to match the cycler at perigee, but it changes the pricing considerations considerably. A space station needs regular resupply launches, which will maintain a high launch cadence. There's also a range of prices for tourism, because the cost of the launch is separate from the cost of staying on the cycler. Tourists can choose how many loops around the Earth-moon system they want to take. With costs being driven down by high launch cadence, companies can afford to send scientists up for research purposes. Separate the travel from the destination.

I only have a B.S. in physics. Not enough to do anything but sweep the floor at NASA, BO, SX, or ULA. Also I don't live anywhere near any of their offices.

Bahaha! I swear, if I could get paid to just sit around and run astrodynamics calculations/simulations all day, I'd be the happiest man in the world.

I wish we had launched the DSG ages ago.

I suspect that for the sort of clientele who would be able to afford trips to space, marketing about safety is not going to have much of an effect. They are going to know exactly how safe it is.

Just cut out the middleman and go with a Z-pinch.

Based on these values, let's take a look at the F9's performance at the advertised maximum GTO payload, 8.3 tonnes. First stage burns all 411 tonnes of its propellant, delivering 4.14 km/s of dV. Subtracting the losses I calculated earlier, this places staging at 2.58 km/s. The second stage ignites and pushes to 7.46 km/s, requiring 4.88 km/s of dV plus the 853 m/s of gravity drag I previously calculated. This burns 97.5 tonnes of propellant, leaving the second stage with a net mass of 22.3 tonnes and just 2.03 km/s of remaining dV. This is about 11% (255 m/s) short of a full GTO injection burn. So either there is margin somewhere else (like a lower perigee or a shorter coast period), or the 8.3-tonne advertised payload is calculated for Block V booster and upper stage rather than Block III or Block IV. The Iridium-2 launch used a Block IV upper stage with a Block III first stage. Does anyone know whether the Intelsat-35e mission used a Block III or Block IV upper stage?

Decided to revisit that question of maximum GTO performance for the current incarnation of Falcon 9. As previously noted, the Intelsat-35e webcast telemetry showed startup-1 at 2.606 km/s, SECO-1 at 7.46 km/s, startup-2 at 7.362 km/s, and SECO-2 at 9.855 km/s. So the first burn provided 4.854 km/s net and the second burn provided 2.493 km/s prograde. With a vacuum isp of 348 seconds, the second stage would need to burn 11.5 tonnes of fuel for the second, prograde-only burn. This tells us that during the boost to orbit (from staging at 2.606 km/s to SECO-1 at 7.46 km/s), S2 burned 96 tonnes of fuel for an effective isp of just 296 seconds. That's not too surprising; even at GTO-mass payloads, the F9 upper stage flies a sharply lofted trajectory. The difference between 296 seconds and the true isp of 348 seconds consists of gravity losses. You can also calculate it the other way around; if you set isp at 348 seconds, you get 5.71 km/s for gravity losses of 853 m/s. I'm not sure if or when SpaceX's live-feed telemetry factors in the rotation of the Earth...the Kerbal equivalent being the switch between "Surface" view and "Orbit" view. So that's a grain of salt to take. The first stage has a vacuum isp of 311 seconds. Rather than integrating to go from surface isp to vacuum isp, let us merely treat gravity drag, aerodynamic drag, and isp drag as a single value. Using the vacuum isp of 311, burning the entire fuel capacity of the first stage delivers 4.17 km/s of dV...but since staging took place at 2.606 km/s, we can set isp drag, gravity drag, and aerodynamic drag at a sum total of 1.564 km/s. Elon said that Intelsat-35e was contracted at an apogee of 28,000 km but ultimately achieved an apogee of 43,000 km. A quick glance at an orbital solver tells us the perigee burn for a 28,000x262 km Hohmann transfer is 2,285 m/s of dV, so that's our theoretical minimum dV for insertion from a nominal parking orbit to a 28,000 km Ap GTO. For reference, the perigee burn for insertion to a 43,000x262 km Hohmann transfer is 2.541 km/s of dV, within 2% of the observed change in velocity on the webcast telemetry. (Differences are probably due to the fact that the parking orbit was not exactly circular, the final Ap was probably not exactly 43,000 km, and there was probably a slight plane change that didn't show up in the prograde dV.) The difference between these two is just 256 m/s, so that's the overperformance of the Falcon 9 in this mission. Obviously, the first stage would need much more than 256 m/s of dV reserved for a landing attempt, so it's safe to say the overperformance was not so great that it could have allowed recovery.

So...............

This is also why the Falcon 9 always needs a coast period for GTO missions.

IIRC, raising the apogee higher places the satellite's period as close to 24 hours as possible. This is known as a "supersynchronous" orbit. This way, the satellite does not have to alter its orbital period; it can simply alter its Pe and Ap as gradually as it needs to. Lowering the Ap and raising the Pe while keeping orbital period constant (or only needing to increase it slightly) is more efficient than having an Ap exactly at GEO altitude, because any burn that isn't exactly at Ap will reshape the orbit, and no satellite has enough thrust to go from GTO to GEO instantly.

Coming back to this... According to the Intelsat 32e webcast, second-stage startup took place at 9,384 km/h and 82.7 km. SECO-1 took place at approximately 26,857 km/h and 164 km. Restart took place at 26,502 km/h and 248 km. SECO-2 took place at 35,478 km/h and 262 km. Intelsat-35e massed 6,671 kg. Our best figures on the Falcon 9 FT (from here) are a 4-tonne upper stage carrying 107.5 tonnes of propellant and a 22.2-tonne first stage carrying 411 tonnes of propellant. Now if I can just crunch some numbers...

The "real" stuff, Napalm B, is impossible to procure outside of military contracting, but napalm is shorthand for any form of jellied petrol. "Some guy and his team" would need to purchase a suitable gelling agent and mix it with gasoline themselves...it's not terribly dangerous as long as you don't accidentally ignite it. In which case it's still less dangerous than HTP. The nice thing about HTP is that decomposes catalytically and, if cooled with a small tank of liquid nitrogen, the decomposition gasses can be used as the pressurant to push the liquid through the larger catalytic bed and into the combustion chamber.