The EASA [Realism Overhaul - "Constellation" and Beyond!]

[Comet Tail]

 

So, I decided to start a thread on the progress of my space program.

See, Realism Overhaul is a great deal of work, and takes a lot of time... Also the SLS is awesome.

EDIT: So let me specify something - the title says "Constellation" in quotation marks. Although I'm not launching Ares LV's, I'm basically doing the same program but with the SLS. The plan is to return to the Moon, build a base there (developing the kOS software along the way), and use those lessons to try to make a Mars base. Landing multiple objects next to eachother on Mars is going to be HARD, but it's going to be easier on the Moon without any atmosphere, so I'll make kOS software that can do that, first, then worry about atmosphere later.

Oh, yes - btw, I'm also using kOS, so the harder piloting-skill-requiring maneuvers will be completed autonomously. I'm doing this so I can get consistent performance out of my vehicles. /EDIT

So, anyways, here's my "return to flight" - EM-1, an SLS Block I sends an Orion spacecraft around the moon.

I use a mis-mash of various mods on a build of KSP that doesn't auto-update, since I don't want an update to break everything...

Chaka Monkey Exploration Systems (CMES) and the IRL SLS launch (formerly in December of this year... ) November 2018 have inspired me to try making an SLS in RO. I did some tweaking, CMES's native SLS has some collision issues with trying to attach larger boosters, and is a few tons off, so I patched together an accurate SLS, myself, but still using CMES's fantastic Delta Cryogenic Second Stage (DCSS)  / Interim Cryogenic Propulsion Stage (ICPS).

 

Anyways, the SLS is awesome. Take the Space Shuttle External Tank, stretch it longer, and stick 4 Space Shuttle Main Engines on the bottom to get the core. Add 2 extended (5 segments instead of 4) SRB's on the sides for boosters.

Now, the SLS Block I will take a DCSS [Image1 Image2] and modify it to make the ICPS. I'm not sure what modifications they are, but they're pretty minor...

 

Anyways, to get the performance of my rockets consistent, and so I don't have to fly for 5-20 minutes every time I launch something, and because it's cool and fun, I use kOS to automate the launches. I can talk a bit about how my ascent program works, but I'll save that for when I launch the SLS Block 1b, because I had some HUGE challenges with the ascent program and had to overhaul it to work with that. Suffice it to say, my program is incredibly robust in that it works for just about any rocket, but only if they have pretty high thrust/weight ratios on later stages. The SLS Block Ib (which uses a new, much bigger upper stage) has a second stage with SUPER low thrust, but more on that later...

 

For now, enjoy my run-through of EM-1!

...Oh, by the by, "EASA" stands for "Equestrian Air and Space Agency" 

 

I somewhat recommend seeing the images much bigger, but I'll also put them here:

 

 

 

I'm somewhat considering doing a Venus flyby or something, old Apollo-era style with a Skylab. Got some really cool art of it. [gallery1 gallery2]

 

EDIT: Forget that style, I'm gonna throw in, say... 3? 5? or so highlight images from each mission! That will make this all a lot more interesting:

Expand the spoiler to see EM-1's highlight album.

...I MAY have gone a bit overboard and past the 3-5. Full album is still available in the original post, above, but I have cut out quite a few and tried to make this the "Best of the best of the best".

CLICK TO SEE EM-1 - FIRST LAUNCH OF THE SLS BLOCK I, THE RETURN TO DEEP-SPACE EXPLORATION WITH A CREWED FLYBY OF THE MOON.

Spoiler

 

The SLS Block I rises on its automated ascent. One of my favorite things about using kOS is I get to appreciate the beauty of it all instead of focusing hard on flying it.

Realplumes ftw!

The gasses expand due to their pressure until they match outside pressure. Since pressure is lower higher up, they expand further and further, resulting in a spectacular display of technological power.

Booster separation. 150 second of solid power are over...

...And now 4 liquid-fueled RS-25 - Space Shuttle Main Engines - take me to orbit:

Once in a parking orbit, following a trajectory almost identical to the Space Shuttle's - I have apoapsis in about 30 minutes and the core will re-enter.

As it so happens, the Moon's in the right place that I don't even have to go from a parking orbit to perform Trans-Lunar Injection (TLI)*, and can perform it directly in just a few minutes.

*{TLI: Kicking your craft with the extra speed it needs to overcome Earth's Gravity and shoot on its way to the Moon}

Discovering I can use ALT + > to timewarp at 4x while in space is a life-saver for these low-thrust stages... We go into the night side of Earth as the burn goes for over 15 minutes.

The ICPS (Interim Cryogenic Propulsion Stage) described earlier burns out, and the Orion spacecraft finishes the burn, and shortly after dawn rises over the receding Earth.

And we arrive.

A course correction burn. I think this is a failure of Kerbal Physics - but I'm not sure - but a free return trajectory (Free Return: TLI shoots you on a course where gravity swings you around the moon and right back to Earth, no course correction required, great as a failsafe against engine failure that might otherwise doom the crew) seems to be impossible for approaching within several hundred km of the Moon. Maybe that's realistic, though. Not sure. Kerbal uses a certain simplifying technique for trajectories that is like, 90% accurate, but this MIGHT BE in the 10% of things that are inaccurate with it. Not sure, though.

But anyways, to go back to Earth from a trajectory that passes within 200 km of the Lunar surface, I have to make an engine burn near the moon of about 400 m/s.

I COULD have used a free return from the get-go, but that means not getting within 30,000 km or so of the Moon, and we aren't going to travel 90% of the way here but not finish that last 10% to say "hi" to the Moon!...

Probably could've gotten closer. But future missions will get MUCH closer, don't worry...

Earth is very small in the sky. Obviously it's the same distance from here as the Moon is from Earth, but Earth has a radius of 6,371 km instead of the Moon's 1,737 km - so it's small in the sky from here, but about 3.6x larger in diameter in the sky as Luna is in ours.

Over 3 days, it looms closer:

And in the last 8 hours, it looms a lot closer:

So close you can reach down and touch it... Well, after re-entry, at least.

Note to self: the trajectory I set on was not good for Orion. I needed that deccelerating burn to make the entry survivable, and 57 km is a friendly altitude for periapsis before entry interface...

 

 

Edited April 7, 2017 by Comet Tail

[Comet Tail]

Currently have achieved orbit around the Moon. An Orion MPCV and Altair lander are currently docked and orbiting the moon. Detailed report later.

 

EDIT: apologies, it's crunch time IRL on some projects, so real life is getting in the way of this right now...

Edited April 18, 2017 by Comet Tail

[Comet Tail]

Okay, welcome back!

Still no updates on the return to the moon mission - time to break the 4th wall and say I'm currently writing the descent program! Those poor astromares have been waiting for, what, two months almost? Well, good thing time doesn't flow in their universe until I'm ready for it to.

Anyways, since my current job is writing a kOS code, let's talk about this kind of thing a little.

If you don't want to, then you can just skip on by, but otherwise, expand the spoiler below!

Spoiler

 

I use kOS, and using it to handle things like ascent and descent is really neat because it sheds a little light into what actual astronautical/launch vehicle engineers do.

Honestly, I don't actually know a lot of the real math behind what they do - but I do have a book I intend to look at... Sometime soon-ish.

But obviously I launched the SLS replica before then, and took great care to make everything as accurate as possible - this means I've done this with an accurate simulation of the actual hardware (at least as far as mass, T/W, ISP and Delta-V is concerned, and RO uses ullage simulation, too). I'm glad they put in the safety margins they do!

I've been thinking of making a challenge to put something into orbit with the low T/W ratios upper stages use in real life - it was an amazing challenge to overcome for the SLS, and let me tell ya, the ascent trajectory is absolutely nothing like an "ideal" ascent trajectory and especially unlike the ones in vanilla KSP.

Let's talk about it, and at the end I've got a link to the actual kOS script I used.

So, it's useful to think in terms of a flat Earth, with a centrifugal force C that points upward with a magnitude C = v_horizontal² / (R_earth + h) 

So the faster your horizontal speed, the greater this force. If you solve this to be equal to gravity's acceleration, g = - u_earth / (R_earth + h)² , you can derive the equation for orbital velocity. If you're on these forums, you probably know about orbits, but you may not have known you can derive the equation for orbital velocity simply by setting the centrifugal force equal to gravity!

Disclaimer: some readers may realize I'm not doing the notation exactly right - we're simply rolling with the understanding that we're working the accelerations on the vertical axis, having already divided out the masses to get the craft's acceleration, and implicitly from the context we're working along the unit vector to and from Earth's center (more precisely, the RHS of these equations is being multiplied by the vertical unit vector).

So, as we go to orbit and get closer to orbital velocity, the net force in-between these two will get weaker. Let's call this g_effective, and define it as  g_effective = - u_earth / (R_earth + h)² +  v² / (R_earth + h) = g + C

So, of course, the main goal is to reach orbital velocity, but you must escape the atmosphere as you do so to avoid drag and burning up, and overcome Earth's gravity to do that. Fortunately, Earth's gravity gets weaker as you approach orbit, as you can see in g_effective.

Heavier engines typically mean more expensive engines, more engines of course means more expense, and a worse mass ratio (mass ratio being, of course, the ratio of (total vehicle mass) to the (vehicle mass with empty propellant tanks)), and since that effective gravity gets weaker, you need less thrust on upper stages.

And in the case of the SLS and its EUS upper stage, the thrust/weight on the EUS is rather low. The first few times I tried flying it, it just dropped back to Earth like a brick, and when I didn't, I used way too much fuel on launch. So let's talk about this a bit -

So, obviously, you all know that as a stage's propellant is expended, it gets lighter, and the engines continue putting out the same thrust, so it accelerates faster. Here's the actual curve for the Saturn-V (with SIVb). There's all kinds of little details that change it slightly, but those are off-topic (we may approach them in later posts if they become relevant), so for now just appreciate all the intricate beauty of a modern marvel (and I'd be happy to answer questions!). But as you can see, as each stage burns out, the acceleration of the vehicle kicks up. This little "kick" is vital! By getting a big burst of acceleration, during that time, it's very easy to overcome gravity and give the second stage a nice, strong kick in vertical velocity.

If my acceleration is 1g, and I'm facing straight up, I'm using all of my acceleration just to hold still. If it's 2g, I'm using half of it, and the other half to climb. If I'm going at 5g, only a tiny portion is being lost to gravity. Thus it makes most sense to fight gravity the most directly when your acceleration is greatest, thus at the end of a stage. But at the same time, an excess of vertical speed is a huge inefficiency, since that is energy lost to gravity and didn't go towards orbital velocity.

And in other terms, if you fire sideways to your velocity vector, you're not adding velocity, you're just turning the vector, and any time you have any angle in-between the direction your engines are accelerating you and the direction of your velocity, there's some component of that vector that's sideways to your velocity, thus is being wasted. Yet at the same time, you have to overcome gravity.

So, with these things in mind, since I don't know the math yet for an optimized trajectory, I decided to just play with shapes (values here are pitch*10 so that I could put  it on the same graph as data points for the vertical speed (m/s)). I simply used some simple parabolas so that the pitching movements would be smooth, and correlate roughly with what the earlier launch guidance program had done.

I simply made a guess, then flew the thing about 3-6 times and tweaked the values with each flight to get what you see now. Finally, around 720 seconds when the last stage is near orbital speed, an "interactive" guidance program takes over, which tries to null vertical speed right before making orbit.

And that's the jist of how the guidance program works for the SLS I built.

It's rough around the edges, and only works with a vehicle of a near-identical build, but perhaps some readers may find it interesting. Here's a link.

If you recreate a vehicle with identical T/W ratios, DeltaV's and ISP's across all the stages, (note the masses don't matter - just the accelerations, deltaVs, and ISPs.) then the ascent program will actually work - though you may want to remove all the "STAGE." commands, since they're tailored to this exact vehicle. (first stage command is entered by the user to ignite the engines. The program then does all the staging. The stage commands then, in order; 1. detach launch clamps and ignite SRB's. 2. Jettison SRB's and fire SRB sep motors. 3. Jettison fairings and launch escape tower. 4. Seperate EUS from the SLS core and ignite the ullage motors. 5. Ignite the 4 RL-10 B-2 engines of the EUS stage.)

Though I have a much more robust program, if I'm going to share a program here, it should probably be this one. You have to do all the staging, but it handles most vehicles pretty well if they've got about a 1-1.4 T/W on ignition of the second stage. Though this one throttles back in an attempt to keep the acceleration under 3.5 g's, which may be an issue with how it nullifies vertical speed. But here's a version that doesn't do that, though I'm less certain how reliable this one is.

Every one of these programs have an additional little feature - if you select the moon, or some other object that's up in the sky before starting the program (it may be glitchy if they're on the opposite side of the planet, though! - ie, if they're more than 90 degrees across the globe from you), then during the ascent it will try to align the inclination with that object's orbit.

 

ADDENDUM:

Okay, so, just putting up the Kerbal Engineer readouts really wasn't quite good enough, so here's more explicit information on the vehicle I'm using. Specifically, this is the Orion MPCV Lunar Mission launch configuration - for the one "in flight" right now in my save, to actually take men ponies to perform a lunar orbit rendevouz with an Altair lander waiting for them in Lunar Orbit, land on the moon, perform their surface mission, rendevouz again with the Orion, and return to Earth.

SLS Block Ib Orion MPCV Lunar Rendezvous Mission Configuration

Object	5 - Segment RSRM (each)	Core Stage	EUS	Payload
Wet Mass	748.2 t	998.27 t	154.13 t	41.3 t
Dry Mass	97.35 t	94.07 t*	15.84 t*	33.6 t
Thrust (Vac/SL, kN)	16,295 / 14,389	11,599.5 / 9,295 (total)	440.4 / - (total)	-
ISP (Vac/SL, seconds)	265 / 234	453 / 363	465.5 / -	-
Notes	On SRB sep (t+ 142s) about .2-.5t of propellant remain.	*includes interstage and separation motors.	*includes 1,920 L of MMH and NTO for RCS and ullage	Wet/dry mass refers to before/after SM fairing and LES jettison at T+ 154s

 

*all masses in metric tonnes.

With this, I was able to perform the last part of orbital ascent, TLI, and most of LOI with the EUS. Note that the RSRM's thrust curves mean they burn a bit longer than their thrust, mass, and ISP would indicate (since their thrust drops for part of their burn for max-Q, and again as they burn out).

Edited May 30, 2017 by Comet Tail