An Opposition Mission to Mars in Realism Overhaul

[PLAD]

This report presents a minimalist piloted mission to Mars and back using the Realism Overhaul mod.

I find getting to Mars quite hard-you need a lot of delta-V, so you either need a huge rocket or nuclear rockets (or both), with all their attendant problems (H2 boiloff, launcher size, etc). Here is my analysis, my apologies to all those who've already been through this...

I know of 4 general paths for a Mars land-and-return mission (see a NASA review here):

1) Conjunction class. So called because Earth and Mars go through a conjunction during the flight. You take a (roughly) Hohmann path from Earth to Mars, then wait about 18 months at Mars to take the next Hohmann window back to Earth.

2) Sprint class. You take a considerably higher-energy path to Mars, spend a short time at it (A week or two, the longer you want to spend the higher the delta-v requirements) then take a high-energy path straight back to Earth.

3) Opposition class, Venus first. "Opposition" because Earth and Mars go through opposition during the flight to Mars. With this you take a near-Hohmann path to Venus, fly by it and get flung to Mars. After a week or two at Mars you take a near-Hohmann flight back to Earth.

4) Opposition class, Venus last. In this you take a near-Hohmann path to Mars, stay at Mars a week or two, then fly by Venus to get back to Earth.

Here are rough delta-V requirements and mission durations for a favorable example of each the 4 types. I am assuming a 10-day stay at Mars for the fast missions. Assumed start and finishes are from a 200x200km circular orbit. This does not include the landing or returning to orbit requirements at either world, which would be the same for all cases. Aerobraking to the final Earth landing is assumed. I used my Flyby Finder to get this data.

Type........Leave Earth.......Brake at Mars.......Return from Mars...... Duration

1 ............3640m/s ...........3810m/s.............2000m/s...................925 days

2 ............3990.................4410..................3980.......................416 days

3 ............3920.................4260..................2480.......................538 days

4 ............3640.................3810..................3890.......................456 days

These require a lot of fuel that has to be storable for up to a year or more, thus low-isp. But you can save a lot if you aerobrake into orbit at Mars. Let's look at the total dV requirements with and without Martian aerobraking:

Type....... Total dv,...... ......Total dV,................. Duration

...............no Mars aero ......Mars aerobrake

1.............9450m/s.............5640m/s....................925 days

2 ............12380................7970........................416

3 ............10660............... 6400.........................538

4 ............11340............... 7530.........................456

Now options 1 and 3 require only a small impulse to return from Mars and the trip can be managed with only low-isp rockets. For comparison, going from LEO to a Lunar orbit and back to re-entry takes about 5000m/s.

Here are my mission parameters:

-This is a test/plant-a-flag flight more than a science gathering mission. Think Apollo 11. I therefore picked a short Mars stay, with mission option 3 being the best deal for duration (consummables mass) versus dV (fuel mass) required.

-I limited my launcher to Saturn-5 size, so launch weights below around 2800 tons.

-After lots of testing I found my speed at Martian atmospheric interface cannot exceed 7800m/s for a managable aerobraking. I also wanted to be able to abort the Mars aerobrake and do a free-return to Earth if needed. This tightly constrained my Earth-Venus-Mars leg.

-I Settled on a crew of only 2 Kerbalnauts. I don't intend to leave anyone in Martian orbit. The carrier won't do anything while the Kerbals aren't on it.

-CG is critical for the aerobraking to work, so all motors and RCS on the carrier/lander use the same fuel, which can be pumped between 3 tanks as needed.

-I cheated with H2 boiloff (details below), and fudged the centrifuge and DRE at Earth, but other than that I used the mods as-is. I use all the required mods and most of the difficulty mods except Remote Tech and Engine Ignitor. It's RO version 8.1.1.

THE WINDOW

Here are the FFRSS readouts for the Earth-Venus-Mars leg and the Mars-Earth leg. This is the best one I found before 1980, the Mars return is fast and relatively low-dv, and the E-V-M leg gets to Mars exceptionally early. Note I limited the E-V-M graph to only show paths that arrive at Mars before day 8175. Most of them had to be rejected because the arrival speed at Mars was way too high. I will depart from Earth on June 3rd, 1972 (Y22 D160 at 0:00, UT day 7825) and arrive at Mars on May 19th, 1973 (Y23 D145, UT day 8175). Spend 10 days on Mars and fly back to Earth from May 29 1973 to Nov 4 1973.

Javascript is disabled. View full album

THE LAUNCHES

The first launch puts the carrier/lander in orbit 8 days before Earth departure (I'll call this D-8). This is followed 1 day later by the crew lifter, which is carrying 6 Kerbalnauts-Jeb and Bob who will go to Mars, and 4 engineers who will test all systems on the ship over the next 5 days. Once they give their OK, the 3rd and final launch, on D-1.5, puts 144 tons of LH/LOX (8 tons can boil off by D-0) and the departure motors in orbit. A spacewalk puts in reinforcing struts and joins the cables between the carrier and the tank. The crew lifter returns to Earth with the 4 engineers on D-1, and the ship is ready to go.

Javascript is disabled. View full album

Here was my biggest cheat. I have not mastered H2 boiloff yet, and between running out of battery power at night and the 36 hours between the tank launch and D-0 I lost too much and had to Hyperedit 3 tons back in to make the departure. (I have to finish all the orbital insertions manually and didn't have the heart to try another launch of the tank.)

EARTH-VENUS-MARS

D-0. The departure burn is 3941m/s, plus a 7m/s correction burn after leaving Earth's SOI. The tank is then extended from the carrier via the 4 cables and the assembly is spun up to 2 RPM with RCS. The CG is closer to the carrier than the tank, but I calculate the radius as 35M at the 'lowest' occupied section for about 1 lunar gravity. The Venus flyby is at D+160.4. The assembly would have to be despun for the manuevers 2 days before and 2 days after Venus flyby, then respun for the trip to Mars.

Javascript is disabled. View full album

It was suprisingly easy to get it safely extended and spun up, but the moment I hit time warp I got a 'cannot warp while under acceleration' message. If I took tension off the cable the 2 parts would fly apart and rip the cable when I hit time warp. So I didn't really spin it the whole way. Despinning was a nightmare anyway. And how would the ship keep an antenna pointed at Earth during spin? I decided next time I will eject an autonomous antenna platform with a large high-gain antenna to talk to Earth, and a small omni antenna to talk to the ship, and keep it about 1 km from the ship for the E-V-M leg. I've also found a 'maintain spin under warp' mod, I'll try that soon.

I also need to put clouds on Venus, though I fear the memory hit as I'm running near my limit with so many mods.

MARS AEROBRAKING AND LANDING

Two days out from Mars the empty tank is released before the Mars periapsis adjustment. This is the last chance to switch to a Mars flyby and direct return to Earth if the aerobraking has to be scrubbed for some reason. To prep for the aerobrake all fuel is moved to the frontmost tanks and the panels are retracted. Periapsis is 45.1 +/-0.15km, too low and the ship breaks apart, too high and it is not captured into orbit. Apoapsis after braking is around 1500-2000km. Peak deceleration is ~4.5 gees. At 1st apoapsis the periapsis is raised to 160km. Fuel is moved back to the carrier and the lander separated. Lander periapsis should be from 65-75km depending on where you want to land. The lander is rock-steady when aerobraking, that huge heat shield makes it easy. Drogue chutes release at 25km and 1500m/s, main chutes at 3km and 350m/s, heat shield then releases, gear out and motors on, release chutes and land on rockets.

Javascript is disabled. View full album

The first aerobrake is tricky, it's at the edge of possible. RCS is needed to keep it stable, and the ship moved a bit too much and lost 2 rear RCS thrusters, one solar panel array, and the outer TACLS container (that was there for balance). Turns out I forgot to remove the waste tanks from that container and it was heavier than I planned for, (yes, my plans were full of s**t) but after it ripped off the ship was more stable and survived the aerobraking. Final carrier orbit was 160x1500km, well within specs. The landing went well (the result of about 20 test runs when designing the lander) and the guys were on Mars after a 350-day journey! Note: for some reason the landing motor gimbals worked in reverse so I turned them off and used RCS for direction control.

MARS EXPLORATION

The guys lived in the Itinerant module while on the surface. They spent the first day inside getting used to moving in Martian gravity. Near noon on day 2 they took their first steps on Mars. In later days they unloaded the rover and experiments and did some science. After 10 days on the surface they returned to orbit and docked with the carrier. The only trouble was the rover dynamics, when crossing certain lines the rover would get kicked up in the air. I kept it below 8m/s and it never sunk into the terrain. Communication with Earth would be a trick, a high-gain antenna can communicate directly when Earth is above the horizon, but what do you do after Earth sets? I'd accept being out of comm for 12 hours at night and early morning, the time lag means they are never really in direct communication anyway.

RETURN TO EARTH

It took about 4200m/s to get the SSTO lander back to the carrier. I debated moving the lander RCS units to the carrier to replace the lost ones but decided that would be too unrealistic. I had 1000m/s more return fuel than needed, I could have kept the lander on during the trans-Earth burn, and even used the lander's motors if the carrier main motor had failed. But I discarded the lander and did a 2519m/s burn to get back to Earth. The lost RCS was a problem- The carrier was unbalanced at this point and needed all the RCS to keep it on course near the end of the return burn. So I had to break the return burn into 4 parts and repoint the ship when it got too far off course. This main motor also gimbaled in reverse so I couldn't use its gimbals either. Maybe a result of flipping the ship around during construction? In any case my huge fuel surplus allowed me to use a faster return to Earth than I had planned- only 159 days instead of 180 in my original plan.

Javascript is disabled. View full album

My final cheat was on Earth re-entry. I hit the atmosphere at 12.1km/s, about 1.2km/s more than a moon return. This turned out to be utterly impossible with the standard DRE settings, the ship would heat up to 2000C and explode around 82km, even with a periapsis at 80km where the deceleration was trivial. So I changed the DRE settings to allow it to survive. The Stardust mission reentered at 12.4km/s and survived with a PICA shield so I don't feel too bad about this.

MISSION SUMMARY

Primary crew total time in space: 525 days

Total launch weight: 2216+243+2775 = 5234 tonnes

Cargo landed on Mars: 3.5 tonnes (The Itinerant container and everything attached to it but not the ascent stage and fuel)

Now it's time for me to upgrade to the latest RO and figure out H2 storage and high-speed re-entries.

[immelman]

This is a fascinating and informative mission. I've read Steven Baxter's Voyage, which includes a Venus flyby, looking at your data this is primarily to reduce mission time?

This also really brings home just how hard this would be, getting parachutes to work at such high velocities or using rockets (with the weight penalties and having them fire in a supersonic jetstream).

Regarding high velocity Mars EDL, here is a very interesting paper on the subject you may want to check out:

http://www.4frontierscorp.com/dev/assets/Braun_Paper_on_Mars_EDL.pdf

It pointed me into using an adapted conical heat shield for a lifting body re-entry.

Real life has curtailed my RO Mars mission (as well as all my Kerballing), reading this has really encouraged me to get it going again!

Edited May 31, 2015 by immelman

[Mister Dilsby]

I learned a lot reading this, thanks for posting all the details. Playing stock with no life support and small planets is loads of fun, but it's definitely nowhere near this level of difficulty.

[PLAD]

Thanks all, and for the link to that article immelman, that was great. It is a tribute to RO that I was forced to deal with a lot of the issues it brings up. For instance I felt a bit guilty about not installing the engine ignitor mod, which allowed me to use non-throttling motors for the final landing by letting Mechjeb pulse them really fast. The article covers this with "Recent full-scale testing...demonstrated that pulsed-mode engine firing is robust." (Though I wonder what the maximum engine size tested is for that.) And the huge problem I had-slowing the ship down enough with the heat shield to allow the main chutes to open before the ship hit the ground- gets major coverage. For comparison my lander weighed 24,300 kg at Mars atmosphere interface, with a 78.5 m^2 heatshield for a loading of 310kg/m^2.

And yes, I flew by Venus to shorten the mission by about 400 days. I figure the first mission to Mars should have a short stay for safety's sake, and the free-return to Earth that you can get with a Venus-first flyby is nice too.

The last minute of the Mars landing was the most exciting part of the whole mission. As the article says, "In this very dynamic phase of flight, robust event sequencing and timeline margin are critically important." Though no one asked, here is a detailed breakdown of my landing:

L-139 seconds: Activate drogue chutes (to prevent the ship from wobbling as it slows down)

L-44 : Activate main chutes (any sooner and the ship is moving too fast to survive the jolt when they open, any later and the ship will hit the ground too soon and fast)

L-42 :release the heat shield (keep it on 2 seconds to buffer the main-chute-jolt, then release so it lands well clear of the landing zone)

L-41 :open landing gear (as soon as the shield is out of the way)

L-40 :Start motors (to assist with slowing down in time, in practice I turn on Mechjeb's 'land anywhere' here)

L-23 :release main chutes (they've done most of the slowing they're going to do, and any later and they will land on the ship)

(Adjust landing site by ~+/-100 meters if needed)

L-0 :Touchdown!

A lifting body would allow a lot more control on where the ship would land, but I don't know if the problem of getting slow enough to be able to open a chute big enough to make a difference would change.

[Cydonian Monk]

This is incredibly impressive. Nice mission!

[Fizwalker]

This has been a very interesting read. Kinda makes me want to fire up my copy of RO and start messing around in it again.