<ProjectThoth>
What is the upper limit to stage recovery without significant TPS?
<ProjectThoth>
I'd assume around 2-3 km/s - ish.
<Bornholio>
it can vary wildly depending on drag to mass ratio and also on entry angle. If the angle is steep then the peak heat load is high but total load is much less. g forces are then more extreme. Also TPS can vary froma nice thin coating to altering the alloys and coating in tile. Add in whether you have some fuel mass left to soak some heat. 2200m/s is bandied about as a speed tho
<wb99999999>
I didn't even know the shuttle build was out
<wb99999999>
so how much better is it?
<lamont>
yeah it does shuttle-PEG to orbit with yaw steering now, plus its wired up to the node executor for finite burn control over node execution
<lamont>
plus it has my fixes to the transfer planner so interplanetary transfers will give you a burn that hits the planet
<wb99999999>
fantastic
<wb99999999>
do I still need to hit x to zero the throttle during the ascent?
<wb99999999>
so it doesn't instantly tries to reignite when the guidance is finished
<lamont>
i just setup an Earth-Mars transfer out of a 185x185 parking orbit, adjusted it to have a 1,005 km periapsis and the node executor hit i think a 960km periapsis
<lamont>
so thats like a 45km miss in periapsis over an entire transit to Mars
<wb99999999>
that's glorious
<lamont>
yeah
<lamont>
oh that bug, yeah turn off RSS/RO special handling in the uh… Settings Menu
<lamont>
i bet that is still busted
<wb99999999>
oh it see...
<lamont>
i just turned that on so it’ll troll me next launch i do
<lamont>
that’ll make me fix it
<wb99999999>
well I don't mind hitting x once every launch
<wb99999999>
but it's comforting to know that's not intended
<lamont>
yeah but you shouldn’t have to do that, and its vastly more precise than you are now
<wb99999999>
what? so now [boost pitch end] is an option?
<wb99999999>
oh I see
<wb99999999>
it switch to close loop once it hits the right pitch
<lamont>
yeah
<wb99999999>
amazing
<lamont>
you can play around with the buttons to modify that. you can have it end the pitch program at a certain pitch and then fly with zero AoA until it intersects the pitch prediction. or you can have it end and then fly with zero AoA until you hit a given KSP stage.
qwertyy_ has joined #RO
<wb99999999>
what does [leading angle to plane] do?
<wb99999999>
I am a bit slow on comprehending the terminology...
<lamont>
yeah people find that confusing
<lamont>
that is the exact same field that is in the “launch to plane of target” box
qwertyy__ has quit [Ping timeout: 198 seconds]
<lamont>
literally the same variable, but with a different label on it
<wb99999999>
I see
<lamont>
if you put zero in there your target plane will intersect KSC’s location exactly when you launch
ProjectThoth has joined #RO
<lamont>
(KSC will then rotate away from that plane at ~400m/s)
<wb99999999>
yes
<lamont>
right now that box only affects launching to plane of target, and is for leading the target plane
<lamont>
i’ve found that for a 90 degree inclination orbit a value of 1-2 seems to be good
<lamont>
value of 0 is best value for due east launches
<lamont>
i’ve been using 1 for launches to the plane of the moon
<wb99999999>
got it
<lamont>
right now if you launch to 90 degrees and don’t lock to a target it gets ignored, but i’m probably going to make it so you have to put a value in there for those launches as well
<wb99999999>
I think now it has been perfectly fixed but before there's a very odd glitch
<lamont>
and 1 degree of earth’s rotation is 4 mins which is about half of a typicalish 8 min ascent, and that probably isn’t a coincidence
<wb99999999>
with the atlas-centaur PEG it always got confused with my Atlas V lookalike
<wb99999999>
somehow
<wb99999999>
of course it won't happen no m ore
<lamont>
shuttle PEG should be a lot better
<wb99999999>
yes
<wb99999999>
but I am very curious on why it is so specific
<lamont>
altas-centaur PEG had all kinds of issues with higher TWR shorter burning rockets like titan 2, which shuttle PEG handles well
<wb99999999>
hmm...I guess it's because the burn out TWR of the CCB
<wb99999999>
CCB is a HUGE stage
<lamont>
yeah i think someone found a stupidly low TWR upper stage that atlas-centaur PEG had issues with
<lamont>
oh maybe that was you?
<wb99999999>
no, it wasn't me
<wb99999999>
I am more russain style
<wb99999999>
usually I do around 1.3~1.4
<lamont>
well someone found a real rocket that atlas-cenatur PEG couldn’t do
<wb99999999>
I used to launch with 1.2 a lot but then I checked the numbers on gravity loss
<lamont>
titan 2 has two 1.4 starting TWR stages stacked on top of each other and old PEG hated that, too short to orbit
<lamont>
i’m actually testing with titan 2 with no payload which is about the worst i can think of in that direction
<wb99999999>
what about the original Atlas D
<wb99999999>
that's the worst
<lamont>
oh so the problem with that is going to be the stage-and-a-half design
<wb99999999>
and also the short burn time
<lamont>
although I bet you can do it. you’d need to use the feature to engage PEG after you stage. then …. oh wait, yaw steering may be insane…
<lamont>
i still rely on peg to converge and give me a launch azimuth on the pad
<wb99999999>
I'd suggest a box for players to type in a staging timing for half stages
<Qboid>
[#951] title: Enhancement: Declarative Programmable Autostaging | There's a bunch of rockets that we can't fly entirely correctly with PEG or autostaging because they need special things to happen which violating the "simple" Kerbal-style autostaging rules. For example:... | https://github.com/MuMech/MechJeb2/issues/951
<lamont>
it really needs support in the FuelFlowSimluation and the stage stats
<wb99999999>
oh yeah, the throttle-down ones are nightmare
<lamont>
yeah i’m not even sure how you ‘tag’ an engine to be the core engine to throttle down
<lamont>
the information is probably in stage stats though — you want to throttle down engines that are burning in the stage, and will continue to burn in the next stage
<wb99999999>
but then will it have trouble with Sea Dragon style vehicle?
<lamont>
?
<wb99999999>
basically
<wb99999999>
you have your two stages
<wb99999999>
but then the seconds stage have a bunch of verniers that ignite on launch to help steer the whole vehicle
<wb99999999>
and they continue to burn after staging, and cutoff when the second stage is done
<wb99999999>
Sea Dragon
<lamont>
i think mechjeb’s stage stats should handle that fine
<wb99999999>
alright then
<wb99999999>
I might begin to sounds ungraceful
<wb99999999>
but there's a few more extreme cases I can think of
<wb99999999>
do you want to hear about it?
<lamont>
nah i like the extreme cases
<wb99999999>
so what about...
<wb99999999>
I used to have a stage-and-half rocket
<wb99999999>
that has 3 types of propellant in it
<wb99999999>
Lox, LH2, and kerosene
<wb99999999>
so what it does was burn both sets of engines on launch, and ditch the kerolox engines halfway
<wb99999999>
XD
<lamont>
that is weird
<wb99999999>
I am serious
<lamont>
yep
<wb99999999>
I actually made that rocket
<lamont>
but that’s doable
<lamont>
that’s just the stage-and-a-half problem
<wb99999999>
okay then
<lamont>
mechjeb’s fuelflowsimulation would handle that fine
<wb99999999>
my concern was the shared Lox pool
<wb99999999>
that might cause some confusion
<lamont>
nah
<lamont>
it already needs to deal with issues like HTP and boiloff
<wb99999999>
right, HTP
<lamont>
so engines don’t necessarily drain all their propellants equally and it needs to know when to stage when any one of the inputs to the engine run out
<lamont>
plus HTP is a “non-ISP” fuel that comes off the mass but doesn’t produce dV
<wb99999999>
sounds nasty
<lamont>
hateful shit
<ProjectThoth>
LOX hates you more than HTP.
<wb99999999>
blame the Germans who did it first on the A-4/V-2
<lamont>
so at the same time PEG can still handle quite a bit — i just did a manual throttle down on an Altas V HLV core by using the throttle limiter on the engine in-flight
<lamont>
PEG adjusted
<wb99999999>
yes,it is always good at contingencies
<wb99999999>
but a question out of the topic: was the HTP thing done for simplicity?
<lamont>
its more the stage and a half design where stage stats will tell you something like 8200 dV in the rocket which isn’t enough to get to orbit, because mechjeb assumes you’re dragging all the engines to orbit
<lamont>
simplicity?
<wb99999999>
I mean the engine designs
<ProjectThoth>
wb99999999: It was done because it was well-known, and readily adaptable to the problem of "flushing high volumes of propellant into a combustion chamber."
<wb99999999>
early engines have HTP gas generators, and I wonder why
<lamont>
oh got me, i’m not that good on the history of the engines
<lamont>
i just do GNC =)
<lamont>
*do
<ProjectThoth>
I mean, the steam you get from decomposing HTP is, what, 600-700 degrees F? It's hot enough to make your life miserable, but from a materials standpoint, it's really easy to deal with.
<ProjectThoth>
And the decomposition reaction was well-known to chemists in the 1940s.
<lamont>
okay i’m heading off for the night
<wb99999999>
'ave a good one
<lamont>
one tip on PEG is to make sure your stage stats are correct
<ProjectThoth>
So instead of pouring years of development work into the kind of gas-generator cycles we're familiar with, you could just rig up a tank filled with a few hundred kilos of HTP and dump the exhaust overboard.
<lamont>
i had one rocket where i had a root part on like the payload decoupler and not in the payload, so the mass of the upper stage was off, so the thrust, etc was all off, so PEG didn’t burn long enough to hit orbit
<lamont>
the huge dV in the centaur stage should have bit a hint that something was off — i just re-rooted to the payload, dV dropped to something more reasonable, PEG behaved
<wb99999999>
everybody loves the centaur
<lamont>
hydrolox baybee
Wetmelon has joined #RO
SirKeplan has quit [Ping timeout: 198 seconds]
SirKeplan has joined #RO
xShadowx has joined #RO
<ProjectThoth>
I... I think I've done it.
<ProjectThoth>
I've actually managed to recreate Saturn VB/S-ID.
BadRocketsCo has joined #RO
<BadRocketsCo>
Howdy
Senshi has joined #RO
aradapilot has quit [Ping timeout: 207 seconds]
Senshi has quit [Read error: Connection reset by peer]
<wb99999999>
what
<wb99999999>
you mean that 1.5 stage Saturn 5?
BadRocketsCo has quit [Ping timeout: 186 seconds]
<ProjectThoth>
wb99999999: yeet
Senshi has joined #RO
Senshi has quit [Read error: Connection reset by peer]
<ProjectThoth>
Which - I'm as shocked as you are - is capable of launching and landing a very stripped-down Gemini. https://i.imgur.com/mfhZ2Of.png
<ProjectThoth>
I should try doing Phoenix or ROMBUS next.
Maxsimal has joined #RO
ProjectThoth has quit [Quit: +++out of cheese error+++]
<wb99999999>
wait
<wb99999999>
now PEG knows to use RCS for final adjustments?
qwertyy__ has joined #RO
qwertyy_ has quit [Ping timeout: 186 seconds]
ferram4 has quit [Read error: -0x1: UNKNOWN ERROR CODE (0001)]
ferram4 has joined #RO
ferram4 has quit [Read error: -0x1: UNKNOWN ERROR CODE (0001)]
ferram4 has joined #RO
ferram4 has quit [Ping timeout: 200 seconds]
ferram4 has joined #RO
Senshi has quit [Read error: Connection reset by peer]
Starwaster has quit [Ping timeout: 207 seconds]
Senshi has joined #RO
<Rokker>
Bornholio: I wonder if the USAF will buy any V-280s
stratochief has quit [Remote host closed the connection]
awang has quit [Ping timeout: 198 seconds]
awang has joined #RO
<awang>
lamont: \o/
<awang>
That's amazing!
<awang>
Haven't been able to play KSP since mchenry9, unfortunately
<awang>
Also, I could have sworn that MJ switched directly from pitch program to PEG instead of pitch program -> unguided gravity turn -> PEG when ending the pitch program early
awang has quit [Ping timeout: 200 seconds]
Senshi1 has joined #RO
Senshi has quit [Ping timeout: 200 seconds]
Senshi1 has quit [Read error: Connection reset by peer]
Senshi has joined #RO
Senshi has quit [Read error: Connection reset by peer]
schnobs has joined #RO
Senshi has joined #RO
Senshi has quit [Read error: Connection reset by peer]
aradapilot has joined #RO
Senshi has joined #RO
aradapil_ has joined #RO
aradapilot has quit [Ping timeout: 194 seconds]
ferram4_ has joined #RO
ferram4 has quit [Read error: -0x1: UNKNOWN ERROR CODE (0001)]
BasharMilesTeg_ has joined #RO
BasharMilesTeg has quit [Ping timeout: 207 seconds]
wb99999999 has quit [Ping timeout: 180 seconds]
<lamont>
awang: there’s a button you get below that to wait until the a KSP stage that lets you do a gravity turn to PEG
awang has joined #RO
<awang>
lamont: Yeah, I know the "start PEG at stage ____" will use a gravity turn, but might it cause a pretty big change in pitch if the gravity turn isn't just right?
<awang>
Er, when PEG finally engages
Maxsimal has quit [Ping timeout: 180 seconds]
Maxsimal has joined #RO
Starwaster has joined #RO
Starwaster has quit [Ping timeout: 194 seconds]
Starwaster has joined #RO
<lamont>
awang: yep
<lamont>
that’s the problem with not having any kind of proper optimizer for launch azimuth and gravity turn
<awang>
Ah, yes, lack of trajectory optimization strikes again
<awang>
Even with that missing, would it be possible to have the early end of the pitch program go into a gravity turn instead of going into PEG right away?
<lamont>
oh yeah i thought about that
<lamont>
so give it a kick, gravity turn, then do the degree countdown to matching the PEG prediction
<awang>
Something like that
<lamont>
yeah i should fix that
<awang>
The set turn rate makes me nervous towards the later parts of in-atmo flight
<lamont>
the AoA limiter can help
Probus has joined #RO
<lamont>
i think we could do something pretty similar to that Ares I launch optimizer fairly easily
<lamont>
mechanizing initial rise, pitch over, follow AoA and feeding it into an integrator is pretty simple
<lamont>
but then would need to find a way to extract the guts of PEG into a routine that would be able to run a simulation
<lamont>
we’ve already got some optimization algorithms in alglib that get used for the transfer planner
awang has quit [Ping timeout: 194 seconds]
<schnobs>
Been a long time since I dabbled in launch scripts, but IIRC you gained quite a bit of freedom by 30km, and could do (almost) as you liked by 60km.
<schnobs>
So I just hardcoded an initial pitch maneuver that should be "about right" and left it to later-stage guidance to sort things out in the upper atmosphere (and beyond).
awang has joined #RO
<awang>
lamont: Abstract of the paper makes it sound pretty simple
<awang>
I lack the knowledge to agree/disagree
<awang>
Also, have you read "Rapid Optimal Multiburn Ascent Planning and Guidance" and/or "Closed-Loop Endoatmospheric Ascent Guidance"?
<lamont>
schnobs: that’s what we’ve got
<lamont>
yes
<lamont>
both ping lu’s?
<awang>
Yep
<lamont>
yeah
<awang>
How impossible to implement would those be?
<lamont>
those are fairly advanced
<schnobs>
lamont: what's the problem, then?
<schnobs>
Please don't tell me that you want it to be prograde-only all the way.
<awang>
lamont: Probably not feasible in the near future, then?
<schnobs>
that's not very hard to do but requires a (relatively) OP upper stage.
<awang>
Also, when exactly does PEG perform the roll maneuver?
<awang>
Er, PEG and/or MJ
<lamont>
PEG doesn’t do roll, that’s in the ascent guidance module, i think its at 50m or something
<awang>
Ah
<awang>
Hard-coded?
<awang>
schnobs: Just saw your messages. Were you talking about the stock system with the 30/60km thresholds?
<lamont>
schnobs: the problem is if you’re trying to hit say 24N latitude in Principia to hit the plane of the moon from a 28.603N lat at the cape and matching up the gravity turn with the upper stage yaw steering in a somewhat optimal way that keeps AoA reasonably low
<awang>
(your messages weren't in my scrollback, so I had to go look online)
<schnobs>
look online? Didn't know that was an option...
<awang>
Someone linked it on kspacademia a day or two ago
<awang>
I wish I knew about it earlier
<schnobs>
awang: nah, the 30/60km lines were RSS. For a high-powered first stage like Titan, 30km is a bit low to point anywhere but prograde... but it's a good rule of thumb.
<awang>
schnobs: Wow, those are a lot lower than I expected
<schnobs>
awang, re: "hardcoded": basically, "try to make a smooth-looking turn that puts you at x degrees pitch when you reach y velocity".
<awang>
I've had to use a pretty slow pitch program in PEG to reach over 90km for my upper stage ignition
<awang>
Otherwise, it'll flip
<lamont>
awang: https://arc.aiaa.org/doi/abs/10.2514/6.2012-4843 is another Ping Lu paper — that algorithm is broadly similar to PEG (analytic predictor-corrector) but has more flexible targetting rules and is more similar to a calculus-of-variation implementation
<schnobs>
Coupling it to airspeed proved to be pretty versatile, it was suitable for a rather wide range of SLTs.
<lamont>
awang: that was probably buggy PEG thrust integrals flipping your rocket at staging, i think i fixed all that and staging is now much smoother
<awang>
lamont: Wait, really? I thought it was aero forces
<awang>
Yaw was pushed all the way in the right direction, but the rocket just kept turning the wrong way
<lamont>
yeah so did i until i debugged some numbers and was horrified
<awang>
And that was at like 80km
<awang>
Oh
<schnobs>
odd, that.
<awang>
lamont: Time for me to download another paper which I can't understand :P
<lamont>
you’ll also see the new code has a “quiet period” of 3 seconds after staging when it doesn’t update PEG at all
<awang>
My "read later" folder is getting a bit too full of those
<awang>
Tries to not have any spurious inputs screw things up?
<awang>
Oh, btw
<awang>
Should I zero out the staging delays in the autostage menus?
<awang>
Or are the defaults fine?
<lamont>
defaults should be fine
<schnobs>
awang: stock atmosphere @30km seems to be a lot thicker than the real thing.
<awang>
schnobs: Really? That seems... counterintuitive
<awang>
Thought Squad would basically have stuffed the actual Earth atmosphere into 70km
<awang>
Apparently not?
<schnobs>
Or maybe it's just the diference in trajectories: when aiming for 180+km, you're not going all that fast @30km. Still busy going up. In stock, you're already desperately building horizntal velocity.
<schnobs>
awang: nah, the "earth-like" principle holds to pretty high altitudes. Don't forget that people want to build SR-71 replicas and expect them to work at their nominal altitude.
<lamont>
yeah i think they stretch out the lower atmosphere a bit so planes can fly higher, otherwise it would be way too thin
<awang>
Ah, makes sense
<schnobs>
for some kinds of "sense", at any rate.
<awang>
Gameplay sense, at least?
<schnobs>
I guess they really didn't know how much small kerbin would afect about everything else.
<awang>
lamont: Doesn't seem I can access the full text of that paper
<awang>
My university doesn't have access, I think?
<awang>
Not a big deal, though
<lamont>
heh, there’s ways...
<awang>
Chances of me understanding what's going on are pretty small
<awang>
Still have a bit to learn about academia-style research, it looks like
<awang>
Haven't had to do much circumventing of access restrictions so far