DBowman

I was thinking the same thing. I don't know if one could improve things by spinning up the bubble and/or putting some charge on it and on other inner/outer metalized bubbles - anything to move to better than spherical. I don't know enough about what shape you'd want your telescope to be. Sounds like a lot of infrastructure to produce the mirror ... shame to make just one ... inteferometry of 100 m mirrors? nice I think the longer wavelength for radio makes it 'much less demanding' than optical frequencies. The optical ones have big evacuated tubes (long) with mirrors that they use to adjust the 'optical distance' within wavelength scale distances. Oh interesting idea, look down, for sure lots of photons...

You can cheat the law kind of using interferometery - the light from two 0.1 m mirrors 10 m apart can be combined to give the diffraction limited resolving power of a 10 m mirror. They still only have the light gathering power of two 0.1 m mirrors but you can get the same image as the 10 m mirror if you look for 'long enough'. This approach has problems: 'station keeping' & combining the light correctly is 'fiddly', probably the more so the more mirrors you combine using long exposures is not good for transient events e.g. star transiting behind planet atmosphere, planets rotate, stars rotate, planets move - some of these you might be able to compute away after increasing exposure even more It seems like on wavelengths that pass the atmosphere / don't have interference modern ground based telescopes with adaptive optics and computation can out deliver orbital systems. It's just so much cheaper to make them and service / update them. Once we start doing techniques like putting a star blocker some kilometers in front of the telescope (so the planet light does not have to 'compete' with the star light) space systems will have an advantage again. These guys have plans to inflate 'soap bubbles' (some kind of uv setting resin) hundreds of meters in diameter and deposit metal vapor in them. Maybe going huge will put the advantage back on orbital systems.

The original prototypes were restartable, and I think the amount of propellant you have to use to flush 'decay heat' is pretty minimal - you have to budget for that non linearly decaying isp / deltav though as part of manoeuvre planning. The need for radiators with nukes in KSP is 'game balance' driven not reality driven. Anytime you want high thrust (compared to ions, enough to do reasonable Pe kicks like >= 0.1 m/s^2 say) and high delta-v then nukes are good. If it just did the trans Mars injection (say) then you might as will take it with you to use for insertion at the other end - per NASA DRA5 nuke option (which even with extra insulation and cryo chillers is lighter than the chemical option). Apparently a significant problem with solar ion as a LEO - HEO tug is that the Van Allen Belt radiation will 'quickly' degrade the solar arrays, I guess the avionics can be adequately shielded. I'm not sure what quickly means. Tethers Unlimited has a plan to 'drain the electron radiation out of the inner belt' using tethers ...

I think there are two main design dimensions for 'space only' craft; can/cannot aerocapture - it's a huge propellant saving once you figure it out and make it safe, but it's also an 'intrusive' design requirement. 'high' (low) / very low thrust - if it has to get into a low orbit of a deep gravity well in a reasonable time it has to Pe kick with 'reasonable thrust' - maybe 0.1 g but if the vessel is large that still adds up. If it never goes near a gravity well or doesn't care how long it take to spiral up/down then it can be 'even more flimsy'. For the non aerocapture vessels I agree they'd start out ISS / NautilusX like; modular and bolted/docked together, trusses etc. When we started constructing in space it might get more interesting; using large temporary inflatable structures as a scaffolding / mould, layering materials inside and outside, additive and subtractive processes, inflating things within frameworks etc - ending up with some kind of 'foamy' structure. Also I think the idea that it's 'a' vessel will go away - it will be a pod / school / fleet of craft. ISS already has a couple of arms that can crawl all over it and specialist crew/cargo vehicles docked. If something is going to be operational for years / decades like an asteroid 'processing' factory ship would be then it's going to have lots of these robo arm 'creatures' crawling all over it, some of them will be space manoeuvrable (e.g. running out tethers etc). To plan it's operation on approach it will have a vanguard of pods of mapping and surface exploration drones/probes. To carry out operations it would have a variety of crawling and flying heavy bot/drones for setting up, moving, and maintaing machinery and materials. Also I think there would be a lot (1000s * regular ships) of very small dumb ships / smart projectiles. In a lot of places it will be easiest to 'fire' chunks of product wrapped in an aerocapture shell at/to a planet. These cargos would spend years in transit on low energy orbits, aerocapture and then be collected.

here is some Skylab footage of what you have in mind. Seems like a good idea if you have the space. Also stuff like throwing heavy medicine balls to each other would be good inertia driven exercise.

I think it would be cost effective to ship water from Mars moons to LEO - but I guess you mean all the way to Earth surface to compete there. Also buying H2O in LEO assumes we've already broken the catch-22 and have some kind of big economy off earth. I think the closest thing to a viable product from space is electricity from orbital solar. There are lots of reasons OS is thought to be impractical or un economic but there is a huge and growing for the foreseeable future market for low CO2 emission power. If the problems could be overcome then there would be a huge amount of activity in earth space and that could be a potential market for some off earth resources.

good point - storable combustion propellants will give you only around isp 320 vs 412 for H2O NTR. NH3 NTR isp 520 - which is okay if the propellant is lifted from Earth but you probably have other uses for the N if you were getting it from ISRU. CH4 NTR isp is 640 and is easy to ISRU on Mars. In fact (once the infrastructure was built) it would be cheaper to ship CH4 from Mars to LEO than to lift it from Earth.

this says that HD disassociates at much lower temperatures than H2 resulting a propellant with lower average molecular mass than H2. I couldn't find anything to substantiate that though. I read this as saying the disassociation energy is about the same - but since it's heavier maybe the same energy is reached at a lower temp? Anyway another interesting fact if you disassociate low pressure H2 via electric discharge and it has trouble reassociating giving you long enough to do something with it maybe (like heat and squirt it out a nozzle).

ISP is equivalent to exhaust velocity. In a hot gas there is a distribution of velocity of the molecules, some are faster than others. More mass per molecule 'squashes' the distribution toward the slower end. there is a couple of graphs here So for a given temperature a 'lighter' propellant will have more molecules going fast. But it's not just proportional to mass or H20 NTR would be 1/9 of 900 not a little less than half of it. I guess since it's double density the 30% extra tank mass for keeping the H2 cryonic would be effectively reduced (halving the length of the cylindrical tank say). Maybe somewhat better since the gas would be 'slower' and not need to be so cold? (edit: 162 vs 71 kg/m^3 ) I don't have the math to to work out how much you lose from reduced exhaust velocity. The 'most probable' molecular speed would be 70% of plain old H2 but the distribution has a long tail at the fast end so it's 'less bad' than that I think.

yep H2O NTR about 412 vs LH2 LOX cryonic abut 450

Mars ISRU concrete would be apropos for Science and Spaceflight - sub zero temperature and the water will sublimate right out of it ...

For sure, LH2 is the ideal fuel (or dissociated atomic H if they figure out how to make it happen; isp 1600). I'm just pointing out that there is some flexibility so you can take advantage of different ISRU opportunities. H2O gives you near cryonic isp, but storable. project rho project rho again... ah thanks - that makes more sense - putting the LH2 through the engine slows em and makes more thermal neutrons and amps up the reaction? Re Chemical vs Nukes - I think the last NASA Design Reference Architecture for Mars (DRA5) had both a chem and a nuke option speced out. I get the impression they's love to go nuke but don't think it will get support. I also think this makes a lot of sense.

@wumpus I believe the NTRs are stoppable and restartable: the fuel rods are spewing out neutrons all the time, but not enough to make a spontaneous chain reaction around the outside of the engine are some control rods that have neutron absorbing material on one side and neutron scattering material on the other if you turn the control rods absorber in then the reactor stays 'cool' if you turn the rods scatter side in then many neutrons end up back in the fuel rods and split U generating more neutrons - the 'chain reaction' and it gets 'hot' some designs had three materials; maybe low and high 'gear' apparently they are 'fiddly' to start/stop since propellant also has some effect on neutron flux ( I don't recall the details, maybe the H2 lets the neutrons 'stay fast' => more likely to split U so cooling the core also makes it more 'active'? ) Many studies plan to use cryo H2 and add 30% to tank mass to account for more insulation and 'chillers' to make it long term storable - but you can run them on many things; CO2, H2O, ...

I found this that says for NERVA-Alpha (1972 study) the core 'and hardware' ( whatever that means - not reflectors or shielding, so maybe just 'racks and cladding'? ) is 870 kg of 2250 kg engine total, about 38%. I couldn't find any info on the actual mass of fissile material it used or what level of enrichment it was. Maybe for these new tests they are just going with the easiest to get and handle fuel as a proof of concept. Do you think it's possible to choose the reactor geometry so the vast majority of neutron flux is planar? or even one dimensional (beam like)?

http://aviationweek.com/space/nasa-keeping-nuclear-thermal-option-open-mars It's an interesting read but if you are not inclined to register to get through the wall the summary is: test with depleted uranium 'dummy' core and heat it with radio frequencies to test that the core will hold together ok, subsequently test with live uranium in some closed system where they combust the H2 exhaust and check that the water isn't radio active (subtle core breakdown / erosion), not use bomb grade uranium for the core ... Bomb grade uranium? was that the original plan? I learned here that some nuke plants can use natural uranium (0.7% U235), most use 5%, bomb grade is 90%, and some modern smaller designs want higher grade. Can anyone shed any light on questions like: Is the mass of fissile U is the key? If so then using bomb grade is about 1/18 the mass of 5% U. Does this mean the U is only a minor part of the 3,000 engine mass? or that some safety considerations make the extra mass worthwhile? I guess the idea with a bomb grade nuke is to build the fuel rods so they can never get critical mass? Cladding, alloying, and geometry? if nuke weapon proliferation from space based reactors is not an issue then is there any low mass materials that the 90% enriched U can by combined with to make safe light fuel elements?

I wondered if the basic mission 'architecture' could reach anywhere else interesting - yes! as long as small rocks are interesting. Turns out there is one Near Earth Asteroid, 2000 SG344, with a 3550 m/s 370 day trip leaving 2027-2031. If you could dock a full Briz core under the baseline stack that would open up another 70 objects ( NASA NHATS ), all with trip times under 450 days and stay times 8-16 days. changes from baseline: less consumables => 'free mass', an extra Briz core can mean a lot more free mass. there is no EVA without either undocking the Soyuz or adding another hatch in the Orbital Module (I know which I'd pick) EVAs mean more N2 budget take science mass for on the spot work and swap in samples for the return The 'architecture' doesn't seem to get anywhere else.

I learned that not only will + + throw this around these per my Mars 2024 flyby it also has the deltaV and consumables to visit this an Aten class Near Earth Asteroid 2000 SG344 If you use one more to stack another Briz-M core as a booster then there are about a dozen NEAs it could visit. Some are smaller than the craft, crazy to think of finding a 3m rock in the middle of so much nothing.

good question - that article I linked earlier said maybe as seldom as once every four years - ouch. I worry about inclination and argument or periapsis (I'm trying to mean like the position of An/Dn). I can see you could setup a Lunar gateway station @ 5 degrees to the ecliptic which would be great for Lunar bound traffic. But other destinations have other inclinations Venus 3.39, Mars 1.85 - and probably with An/Dn not at Lunar An/Dn. NEOs, NEAs? low deltaV gravity assist flybys could be anything ( e.g. one I was looking at was 62.8 degrees). Has anyone seen an evaluation of how that plays out? Oh thanks, interesting. Yep the basic modular conception is the same, though I'm thinking on a smaller scale.

I've managed to convince myself that this is possible for about $500 million - a flyby in mid 2024 using 'mainly' off the shelf hardware. Too much information here

It lives! I made a spreadsheet, and checked it twice. A two Briz mission plan will flyby Venus and Mars with 680 kg propellant margin in the Briz and a full propellant load in the Soyuz! that's 480 to 510 m/s (2/3 on Soyuz main). My spreadsheet includes: Computed everything from a 200 km LEO but under-fueled the Brizs using mass info from the Proton mission planning handbook. I think this is conservative because it would be better (but harder for me to compute) to under-fuel the third stage and dump the more dry mass early. I did this for: inclination change circularisation burns Used the conservative choice where there were multiple source values for mass or ISP Ullage burns out of 205 kg high pressure propellant budget Used a detailed burn schedule, computing ullage, combined, main, tune regimes Full mass Soyuz (missing passengers & H2O2 = extra thermal protection) Used baseline InterPlanetary Module mass per my consumables calcs etc for 510 days + 26 days marshaling Added mass for solar cells etc on Briz Trajectory is courtesy of PLADs Flyby Finder: Depart 25 Aug 2023 02:24, inclination 62.8 degrees 3742 m/s (almost perfect for Baikonur, spooky) Venus @ D + 164 days 13.6 hours, 21 Aug 2024 @ 8054 km altitude Mars @ V + 197 days 12 hours @ 145 km altitude Earth @ M + 147 days 19.4 hours, 15 Jan 2025 reentry @ 12.4 km/s Flight plan uses a Marshaling Orbit Rendezvous strategy (LEO + 3150 m/s abut 13 day period): - Launch Proton + Briz + InterPanetary Module containing consumables and Soyuz supplementary services - Two Pe kicks from Auxiliary Propellant Tank & stage it (saving restarts, 3 might be ok) - One Briz core Pe kick into marshaling orbit - 'later' - launch Proton + Briz with dock welded on into IPM plane - launch Soyuz to rendezvous with dock-able Briz - Soyuz closes and docks - Three Pe kicks from APT and stage it - One core Pe kick to LEO + 2925 m/s / MO - 225 m/s - Soyuz rendezvous with IPM at Pe with a two minute burn - about 13 days to Trans Venus Injection - Soyuz closes and docks to IPM using as far as possible the 1200 kg of residual Briz propellant - up to this point the mission can scrub and return safe - TVI burn 592 m/s in 6:40 seconds - fall for 510 days, no deltaV required for the flyby, just tweaks to make sure the gravity assists are accurate - should be ample deltaV for setting up reentry timing (location) and Pe Things I couldn't compute: How much deltaV the Soyuz would use in rendezvous with IPM - cannot be much since the Briz that lifts the Soyuz has 300 m/s & 3 restarts margin and there is plenty of time as long as the 'matching manoeuvre' is okay. Briz attitude control propellant usage. It comes from the high pressure system the ullage motors use. There is lots of time to get oriented right so it should be very low. Briz usable propellant vs propellant capacity. I saw 2% residual used by a Mars Heavy Transport Architecture, I'm not sure how reliable it is. That would eat 400 kg of the 680 kg margin. Re-entry: 12.4 km/s is less than the record set by Stardust (12.9 km/s & 34 g) and the human g record on a rocket sled is 46 g (not for long I bet). You have to trade off g against peak and total heating - so who knows but could be doable. Lunar reentry is 11 km/s & there were plans to use Soyuz for that, so 'its only 13% faster'... (but the heating isn't linear)

I'd not really been following in detail - only a vague 'more big rocket = good' - but obviously that ignored opportunity costs and if the SLS budget is eating all the funding for it's payloads then it won't end well. Fair point, I just picked their 'nominal' (maybe notional) LEO capability (for what it's worth Ariane 5 has a storable stage that puts 21 ton in LEO I guess like Proton it's part of the 21t and then goes GSO?). If various launchers can use a cryo upper stage to do better getting mass to higher energy orbits (closer to BEO) then that's great. It makes the argument more complicated without challenging the basic thrust; dock assemble specific mission vehicles out of modular specialist chunks; if the mission changes re-arrange the chunk collection chunks lifted by commodity commercial launchers which will over time lift heavier due to the need to put more functionality into limited GSO slots It's going to be impractical to plug together too many components. You can only launch so many chunks in a given time. Baikonur can launch 2 Proton a month, but if you can utilise many launch systems you could do more. I haven't looked at what missions would be possible using this approach. I guess that's the first thing to do - see if there is anywhere interesting to go that's practical using this approach, well you'd need multiple potential destinations. Candidates: Flybys, Moon, Moons of Mars, NEOs, ...

This the space review is sobering reading - SLS launch costs will be huge if one assumes a low launch cadence and sharing the dev cost across the actual flights - billions per flight for 130 ton (even the most optimistic estimate seems not that much cheaper per kg than Proton - at least it would tank many many launches to pay back the dev cost out of the savings). Also the political direction NASA gets keeps changing, you cannot really design highly mission optimised hardware if the mission keeps changing. 20 ton to LEO seems to be the current commercial launcher 'sweet spot' since it's whats needed for commercial GSO satellites, there are a half dozen or so systems that can throw this. Why not spend the heavy launcher dev $$ on developing BEO flight hardware that comes in 20 ton chunks which could be combined in different ways for different missions? e.g. small 3 ton crew launch and high speed reentry vehicle, 5-10-20 ton HABs, 3-6 ton propulsion & manoeuvre modules, 20 ton propulsion modules. Start off with docking assembly and later augment with 'robot arms' to move modules into position and 'couple' them (mechanical, electrical, control, propellant). Probably LVs will be driven larger as limited GSO real estate pushes satellite mass up, the old smaller modules are still useful but throwing bigger propellant modules up will reduce cost and expand possible mission profiles.

@Aazard @theJesuit I'll take a look at the RP-0 tech tree and IFI Life Support.

AGX & kOS The realistic craft often need an 'ullage' burn to settle the propellant. I found some references for the Briz seems they use the verniers for 15 seconds, then light the main engine, leave some over lap, and finally 'tune' the burn with the verniers again. That means each burn has four regimes for computing delta-v - yuk. Anyway that's how they do it. I installed Action Groups Extended AGX - so I could set up a button panel with labeled actions like: toggle verniers, main, RCS (when it's part of a stack with a Soyuz I'll likely switch the 4 blocks off cause the Soyuz ones will have more leverage), etc. It works very nicely, though you have to be careful not to make 'overlapping toggle groups' that confuse it (or you). The panel works fine but it's still painful to do the burns, so ... Kerbal Operating System kOS - which lets you write programs for your craft. It's easy to install, and you can 'come at it slowly sideways' if you are not a strong programmer. I played around learning about; tags ( https://ksp-kos.github.io/KOS/general/nametag.html ) and getting the various engines by tag, some control structures, and did some test burns. The Briz verniers can pulse for 0.05 secs and KSP physics + kOS seems to have 0.02 sec tics - but the engines take more than a tic to turn off so it ends up pretty close to realistic 'delta v resolution'. I had a look at a couple of 'maneuver burn' scripts: https://www.reddit.com/r/Kos/comments/2nfmno/burning_for_set_amount_of_deltav/?st=ivg62ezv&sh=d9132254 https://web.archive.org/web/20160310064841/http://ksp.baldev.de/kos/mtkv3/exenode.txt but they all seemed to focus on getting the right delta-v - my burns will be like 5 - 12 minutes so trying to match the impulsive maneuver delta v doesn't sound right I decided I'd shoot for hitting my desired Ap and thus orbit period. I set the script up to do the ullage and overlap then burn main until the Ap is 4 * the last Ap change below the desired Ap, kill main (it doesn't shut off immediately), start the verniers (about 1/10 th main thrust) and burn until Ap is 3/2 the last ApChange below the desired, and kill verniers. At 200 km it will get the Ap within a couple of meters, and within a couple kilometers out past the moon. it's doing the ullage burn with verniers, debug is printing per kOS tick: deltaV m/s, deltaT seconds, Ap, delta Ap error (here the error is getting 5m smaller per second) the script is below in case it's useful

It was working for me fine in version 1.0 under ksp 1.1.3 things like: worked fine - it would be a shame if that functionality went away. I did have to put some WAIT 0 after messing with the nodes - but I forget why (it's somewhere in this thread).