Starman4308

That... actually gives me an idea. I'd been assuming a Dragon mounted in the top, with a pyrotechnically separated panel in case it needed to separate in a hurry. But what if... we bolted a Dragon to the side of the Starship? Relatively minimal modifications to the Starship, less damage to the Starship if the Dragon separates in a low-stress environment (such as if the Starship heat shield isn't fully trusted yet)... It'd play havoc with the aerodynamics, yes, but with the sheer excess of payload capacity and non-trivial gimbaling capacity on the Superheavy, it might be doable.

I'm pretty sure that's part of SolverEngines, which is bundled with Real Fuels. I'm not sure it's intended for standalone use.

The simple solution is going to be to disable CommNet in the difficulty settings. This is the WIki article on CommNet: I cannot verify whether it is accurate, since I don't play with stock or CommNet (I use RemoteTech instead). If the Wiki article says "only when close", that means it can reach when the planet is close to Kerbin, whereas if they're on opposite sides of Kerbol, it may not reach.

When optimizing for mass, frequent staging should be your preference, especially asparagus/onion staging. The quicker you shed dry mass, the better. Dry mass is the enemy of the Tsiolkovsky equation. When optimizing for cost, it gets a lot trickier, since engines and decouplers are fairly expensive relative to propellant tanks. Because of that, cost-optimized launch vehicles tend to be a bit propellant-overloaded, becoming heavier and likely incurring extra gravity losses relative to a mass-optimized vehicle. Stock is relatively forgiving in terms of how often you stage, since fuel tanks are relatively heavy and expensive. Staging "too early" helps your delta-V, ameliorating the cost of excess decouplers/engines. Staging "too late" likely means a slight propellant overload, and while fuel tanks are relatively expensive... they're still cheaper than engines.

Yeah, it's totally ridiculous. After all, we know 100-year grand tours are best done in a chair in an unpressurized 1.25m service bay. In the absence of any sort of mechanic to force players to use actual living space for their Kerbals, I suspect additional passenger compartments are a low priority for development. That said, plenty of mods do add big passenger compartments, and at least Kerbalism (and probably at least one other) forces use of something more than a chair.

While I do suspect the Starship will have a lower ballistic coefficient, that isn't strictly true. The Shuttle payload bay was (usually) empty, the crew cabin had a reasonably amount of empty space, the OMS tanks were drained, etc. Granted, none of that stopped the Shuttle from being something that didn't glide so much as it fell to the runway.

You're right. My post shall be edited to better reflect what the players want.

Stock KSP has both solar panels and RTGs (Radioisotope Thermal Generators), although I don't think they're unlocked at the start. Before that point, many liquid-fuel engines generate electricity when running, though obviously that consumes a lot of LF+O and moves your craft. Fortunately, in stock, I'm pretty sure antennae only consume electricity when transmitting science; so long as you're not reckless about turning your ship and transmitting science, battery life should be enough.

If this were true, then... Serenity is going to combine a massive bugsquash, introduce a high-performance rigid-body physics engine, add surface experiments and reasons to actually establish a surface presence, give Duna/Ike a barycenter, rebalance stock part costs and masses to make sense, add GP2, add a suicide burn timer and more easily accessed true-altitude and vertical/horizontal speed indicators, add larger-sized NTR/ion engines, let Kerbals know about the magic technology of "ladders" sooner, add tiny RCS blocks early-game for people who don't want to abuse the reaction wheels, and fix inconsistencies and outright weirdnesses with the aero model. EDIT: And a pony.

For P2P, you have the issue that you need a landing site that doesn't catch fire when hit by tons of burning methane... and approach control, and owners of very expensive rockets thinking "Hm, it may only cost $9 million to fly... but likely over $1 billion to replace if something goes wrong at the other end". I find it unlikely that somebody would risk landing a Starship at anywhere but a pre-built, designated landing site. For emergency humanitarian efforts, the Starship has to be compared to the dozens of air transports that can be there in less than 24 hours (substantially less if they start nearby), can literally drop non-fragile cargo onto any reasonably open area, and can be based from anywhere with a sufficiently large runway. For more regular commercial traffic, you have to compare it to aircraft that can launch from and land on a very wide selection of runways, can be refueled with standard jet fuel (no cryogenic methane and oxygen), have well-understood all-weather capabilities, and are presumably easier to load and unload. There's also the issue of payloads. Again, this is going onto an enormously expensive launch vehicle, so payloads will need to be vetted for "will not destroy the Starship". That's true of aircraft as well, but A, that's much better understood, and B, there are many payloads that will survive low-pressure, low-altitude flight... but not outright vacuum. Shaving some hours off transit is all nice and well... but air transport is much better understood, has a lot more infrastructure behind it, is more flexible, and a big chunk of the total shipment time is in handling, not in transit.

Breakthrough Starshot uses speculative, untested technology to claim they can hit 0.2c. Nothing based on well-developed technology is going to get us to 0.01c. It's not charting a course that's hard. Space is rather empty and positions are well-known. It's getting to the staggering velocities required that is the issue. While there are mods that present players with what are functionally warp drives, the stock game restricts itself to chemical, nuclear-thermal, and ion propulsion. The Dawn engine has a specific impulse of 4200s. To send a 1 kg probe to velocity of 0.01c with that specific impulse, you would need about 4*1031 kg of xenon propellant as a hard lower bound. That is 20 solar masses. The highest specific impulse ion thrusters yet built top out around 20000s. That would require a mere 4,400,000 kilograms of propellant... which is more mass than the entire Saturn V rocket, nevermind the question of how you get 4,400 tonnes to orbit (since ion engines have far too little thrust to launch themselves to orbit).

As per usual when I take a break from KSP, I think "I can totally just catch up on the WDYDIKSP thread then post". I've... given up on that. So: backlog of imagery, from an RP-0 campaign, then the RP-0 campaign after I realized I didn't have part unlock costs, then RP-1 after I decided to check that out. A totally safe vehicle returns to Earth from a suborbital flight. Incidentally, on one of these, I accidentally steered south, so I can only assume the pilot was defecting to Cuba. A historically implausible rocket: an RD-108 powered core (straight out of the R-7 Semyorka), an AJ-10 powered upper stage (similar to the Able), and a half-dozen A-9 boosters. This is a Soviet core stage, with an American upper stage, and boosters made from an American improvement over a... German rocket. I'm lucky that, even in its few flights, I had only one loss-of-performance on an A-9 booster. Those are not exactly paragons of reliability. My first lunar impactor of the campaign. Vernier engines were used in place of anything restartable, since I didn't have any high-thrust restartable engines at the time. This rocket, which is clearly not a ripoff of the R-7, separates its RD-107 boosters from the RD-108 core stage in what is absolutely not a Korolev cross.

To the best of my knowledge, those are still random Internet commenters and space enthusiasts with no more authority on the subject than you or I. Preaching to the skies about "airliner-like reliability" won't make Starship as reliable as an airliner. Flying it tens of thousands of times will... and there just aren't enough payloads lined up for that in the forseeable future.

Right now, PF is distressingly large for orbital launch vehicles. There is no indication Starship/Superheavy will be any different. And separating (PF,PA) just makes things needlessly confusing. For the purposes of this discussion, "failure" is failure to protect passengers. Furthermore, there are reasons beyond that simple mathematical equation for having or skipping safety features: 1) The safety feature is expensive. This is most likely the case for something like a cargo aircraft: the probability of failure is so low that nobody's willing to shell out for expensive ejection seats. 2) The safety feature has an excessive penalty to performance. For example, by the time you add ejection seats to an attack helicopter, you might have added so much mass that you've made it more dangerous by reducing its ability to not get shot down in the first place. 3) Cultural/bureaucratic inertia. See the lack of seatbelts in many buses. 4) Marketing (which is a bit part of Tater's point). 5) There is sufficient uncertainty in failure probability to confidently skip safety measures... like an experimental vehicle that most closely resembles a set of vehicles with an awful habit of exploding. I'm also pretty sure that, like most probability-based things, it's going to be a product, not a summation. Probability of increased failure caused by a Starship LES should be fairly small if it is kept simple and you're not the 1960s Soviets who have been launching orbital vehicles for less than a decade. Ejection at supersonic velocities is very unsafe... as is ejection at hypersonic velocities in near-vacuum. I do not consider them a viable LES. Also, Gemini would've been a better example... for which the ejection seats were considered borderline useless. [Citation needed]

Man, that better not be the case. If that's true, then it might be the next President who gets to unfairly claim credit for a project started by the prior President, itself a near carbon copy of a project started by the President before that, attempting to use 1970's era hardware in a way best designed to funnel money to 1970's era contractors. And that would be... bad? Anyways, I seriously hope nobody dies as a consequence of the SLS, which is composed of so much pork that it could probably win Midwest hog-raising competitions without the judges any wiser.

Okay, let me try to clear this up a bit. @tater, would you agree in the abstract that an LES could be a bad idea if the Starship became very reliable? @sevenperforce, please try to understand we're coming from the practical assertion that the Starship has no proven reliability record and is unlikely to accrue such a record for a long, long time, even if it does actually have the reliability to hit the range where skipping the LES is a good idea. Overall, this is a thread about whether we think Superheavy/Starship will work. While "work" is ill-defined, I'll give my 2 cents on a few levels of "work". "Work" as in fly: Yes. SpaceX has not demonstrated the staggering levels of incompetence that would be required for Starship/Superheavy to be incapable of orbital flight. "Work" as in land: Probably also yes. "Work" as in become an economical launch vehicle on par with existing launch vehicles: Reasonably good odds. I strongly suspect the $9M/launch cost will be outright fantasy, but there's definitely wiggle room for unforeseen issues between the best-case scenario and "still good enough to compete". "Work" as in become this ultra-vehicle with airliner levels of reliability with multiple flights per day: Outright fantasy. Something this complicated, there will be design skeletons in design closets.

They might create the capability. Capability that spends most of its time in a hangar because there are no payloads left to launch.

The number Musk envisions is irrelevant. The number that will actually happen is the more important figure, and it's not going to hit even one per day anytime soon.

Point-to-point: it's going to be very hard to compete with airliners, which offer a much more gentle ride, are likely to continue to have much better all-weather capability, and have a very, very, very long track record of reliability. Even the Concorde was a commercial flop, and that shaved about 50% of the time off something like a 747 while being far easier to handle. Cislunar: See DDE's point about "the tiny handful of humans that would ever venture off the one planet to have an oxygenated atmosphere, plenty of tourist destinations, a nice handy ozone layer, etc, etc". Orbital hotels: See DDE's point.

Risk is risk, no matter how you slice it. It doesn't matter if the risk comes from being on an experimental vehicle or pushing a well-tested vehicle beyond its limits: an abort mode still makes risky endeavors significantly safer. If the Starship is extremely reliable, far beyond what I predict, then adding an LES might make it slightly less safe... but still well within acceptable bounds. If the Starship is not extremely reliable, then a robust LES can be quite useful. First: prove it. Second, there are unknown risks in reuse. Equipment wears out, and needs to be replaced. Metal fatigue accumulates. Foreign objects get left inside on accident. There are micro-fractures in Merlin engine turbines: what if those become worse over time? While the bathtub model does suggest a lot of failures occur at the beginning of equipment lifespan, it also suggests a lot of failures occur at the end of lifespan, and nobody yet knows where the end of that bathtub is. Yes, and the Space Shuttle is not the only OLV to have failed or to have redundant systems. Half the time, it's the things you don't see coming that cause failures. Accelerometers hammered in backwards. Foam shaking loose and punching through an aluminum wing. An unexpected interaction between your new fueling procedure and your COPVs. Redundancy is not a magic cure-all wand. The point of all of this is to put crew in a simple, robust, well-understood capsule design for escape or reentry. There's a reason why, despite the magnificently complicated engineering in many OLVs, their escape systems are designed with robustness and simplicity over cutting-edge performance. If the really complicated bit (the high-performance OLV) fails, use well-understood pyrotechnic or hydraulic separation and high-thrust, solid or pressure-fed rocket motors to get off the stack, followed by reentry in a simplified, reentry-only vehicle. Prove it. The closest analogies to the Starship are vehicles that very much need abort systems. Until you have solid Starship reliability data, I assume it'll be about as reliable as more conventionally built OLVs: not nearly reliable enough for no-abort manned missions. The Chinook and similar vehicles have plenty of reliability data, and look like vehicles that have not traditionally needed abort modes. The Starship... does not resemble something with which we have extensive experience.

I'd suggest a planet named Arret, about 150 Gm from Kerbol, further out than even Eeloo. Oxygen-bearing, bit of a gigantic planet with a moon bigger than Tylo. It'll be a real challenge for players to get to, land on, and return from. A more serious suggestion might be a Kuiper belt object in a highly eccentric (e > 0.5) orbit, with a very tenuous atmosphere, something akin to real-life Mars where it's enough to screw you up if neglected but far too thin for effective main parachutes. This would add two interesting new challenges to KSP. Duna's atmosphere is somewhat thin, but it's far thicker than Mars's atmosphere, permitting things like spaceplanes and parachute-based landings, even if requiring more wing/parachute than Kerbin.

On the first point: I suppose. I just get annoyed by people who assume that it will obtain that record as a matter of fact. There's very little evidence I can see to support it will reach that point. I'm not even sure we really disagree all that much: I am just more annoyed by the evidence-less claims that the Starship will get there. On the conflation of reliability and accident risk: does it matter? Orbital launch vehicles, even with very cautious flight plans, still fail much more frequently than fighter jets intentionally pushed to their limits. There is an enormous gulf between where orbital launch vehicles are and where they would need to be for skipping the LES to be a good idea. I agree in the abstract that at some point, a super-reliable vessel would actually be safer without LES. In the practical, orbital launch vehicles are assemblages of lots of high-energy propellants combined with complicated parts built to the extreme edges of engineering, exposed to extreme environments that are far less-well explored than the low atmosphere that most of humanity operates in. There's a few orders of magnitude missing in between where OLVs are now and where they would need to be for no-LES to be a good idea. Redundancy has been a feature of every orbital launch vehicle ever. Even the Space Shuttle had a fair deal of redundancy. Not enough of it, to be certain, but a fair chunk of it. Redundancy is not a get-out-of-jail-free card. It is in SpaceX's favor that they are doing their best to have redundant parts as often as possible, but that does not guarantee success. Overall, the thing Starship/Superheavy resembles most is other orbital launch vehicles. Most of the things that distinguish them from other OLVs are untested approaches like actively-cooled stainless-steel heat shielding. This does not fill me with overwhelming confidence.

Oh, there's plenty of precedent. There's all sorts of experimental vehicles that have never flown with identical reliability records... no reliability record. This is what I have a huge gripe with: people who assume that because the Starship/Superheavy is meant to be this wonderful thing that can fly over and over without fail, it will be this wonderful thing that can fly over and over without fail. Reminds me a whole lot of the Space Shuttle. Granted, SpaceX has advantages here (such as no Thiokol insisting his SRBs be used), but this is wholly speculative. Which is why, in peacetime, fighter jets have their ejection seats removed. Oh, no, wait, they don't. Because fighter jets still crash even in peacetime, and those are vastly more reliable than any demonstrated orbital launch vehicle. It will be a very, very long time and tens of thousands of flights before I could be convinced flying to orbit without an LES would be a good idea. Give it ten years, and maybe I'll be eating my words. Maybe Starship/Superheavy will work as intended and be the wave of the future; maybe it will achieve reliability on par with airliners. Until then, I'm reminded very heavily of the Space Shuttle and of the more conventional designs by much more experienced companies that still occasionally fail. More likely, I suspect flaws will come out of the woodwork, and things will not work out as well as hoped. Probably still successful... but not "airliner-like" levels of reliability and reusability. It's just too different from anything that's been done before for there not to be skeletons in the design closet.

One group of players who would very, very much not like this is anybody running with Real Fuels. It takes a few seconds for real liquid-fuel engines to spool up to maximum thrust, and only then do you release the clamps. I don't want my actual, full-scale Saturn V replica falling to the pad and then tipping over in the 5+ seconds it takes for those F-1 engines to spool up. While perhaps the editor might be made smart enough to, by default, always add clamps to the first stage, there's no reason why they should be hard-locked to the first stage. There are several valid reasons to not have clamps on the first stage, and there needs to be the option of not having them there. EDIT: Also, anybody running with any sort of parts failure mod that can trigger on engine ignition. If engine 3 fails to start... much safer to still be on the clamps and just abort the launch sequence.

I don't see your point about the LES for other launch vehicles. All of them had LES systems, even the borderline useless ejection seats of Gemini, for the most dangerous part of the mission: launch. Once again: "It has to work this way for Mission X, therefore it will work this way" is a rather poisonous attitude. There's no need for early Starship missions to carry more than a few astronauts at a time, and speculating beyond that is premature. Airliner-level reliability is nice to talk about, but none of that has yet been demonstrated. For immediately forseeable missions, Starship can easily afford to keep an LES around.