Science leads to wonderful things (PIC HEAVY)

[Dave Kerbin]

I've been working on landers. My original Duna plan called for 2 landers and when I first laid out the mission in broad strokes that made sense - in theory Duna would need some big lander to safely reach the surface, while Ike was some tiny moon that only needed a little thing. However as I got an Ike lander prepared (previous shots show both the tall prototype and the more squat prototype intended to lower the center of gravity and hopefully not tip over) and put out the mission plan I got to realizing that the Ike lander actually required a good deal of delta-v, since it needed to go to Ike orbit and back to Duna orbit. In fact it was showing more delta-v then a parachute assisted Duna lander.

At the same time my biggest concern about the Duna landing was that it would require lots of fuel to do it on thrust only, but a parachute landing had some major forces being applied to the spacecraft. I eventually came up with a new mission plan. A single lander would be designed for both Duna and Ike. It would land on Duna first, so that the science pods could actually be tied down with struts during the parachute landing. After ascending and docking with the mothership it would be refueled and new science pods docked with it. The Ike landing wouldn't need to science pod struts since it wouldn't be subject to the same shock stresses.

My first design was biased by the pad tests of the Ike lander - I didn't like how those legs could sink in and sometimes tip over the ship if SAS wasn't on. So I borrowed an old concept of mine I call belly landing.

You'll see there is no landing gear, instead there are nose cones (10 impact resistance) for the ship to set down on. I used this kind of landing system in .21 for a reusable Mun lander from a Mun station. This means there is no chance of tipping over. However I ran into 2 problems, the first being that I had to put on more fuel then I really needed in order to simply get the structual dimensions I needed for placing those nose cones below the engine. The other is that while it passed basic shock tests (and the parachute calculator showed it under 8m/s on Duna) it was failing some of the harder ones (shock tests involved strapping on a big fuel tank booster rocket, sending it up to 2000m unmanned, then using the booster to send it at the ground between 100-300m/s. The booster uses sepratrons to fly off and the chutes are opened at 500m while at high speed).

To deal with the excess delta-v I designed this version which required less structural members, using a combination of parachutes to obtain the same result.

This design failed all the shock tests, they would simply tear the limbs off. At this point the science requirements for the program where also being upgraded - the lander would now need 3 science pods!

I went through 2 more major designs. One was a thrust only design that I put on the back burner due to both my concerns about such a landing and the huge weight of the lander and its refuel tank on the cargo ship. The other took on the idea of landing on Duna first with struts on the science pods one step further and attached parachutes on them with struts (since the chutes could be ejected later). This last design still failed some of the shock tests.

I went back to the drawing board and tried to refocus. My biggest problem was that the ships kept failing the shock tests, so much like my lifter I should start a design aimed at that requirement and then try to bolt on the others. The result was a ship designed to put as much of the weight into a single part, and then attaching the parachutes directly to that part. This was the result.

It passed the shock tests with flying colors. Of course it has several drawbacks. Height and landing legs mean I need to be careful about a tip over which would strand the ship. The RCS thrusters are not placed optimally (though that shouldn't be a problem for me). And the parachutes are skirting the line on drag - I will need a small amount of thrust just before landing to touch down safely on Duna. Total delta-v is very tight, it's right at the safety margin so the Ike trip will need to be flown precisely at each stage. One benefit of the design is it fixes some potential issues with the swapout of science pods - the previous up/down pod alignment posed some clearance problems when the lander was docked with the mothership, making it difficult or impossible for the RCS drone to get in there and switch them.

I've also tested the recovery system on 5km drops. A full orbital docking and return test is still needed. The recovery pod undocks from a tender where it now has about 2 hours of battery power in which to dock to a science pod and then use its 200m/s delta-v to deorbit. The parachute provides a soft landing. The ship used to carry fuel pods up to the cargo ship is also designed, though it has not flown yet. The design is intended to get close to the cargo ship, then switch the navball orientation to the fuel pods docking port - at this point the dual set of RCS thrusters and reaction wheels are used to essentially fly the ship sideways for docking of the fuel pods.

[Dave Kerbin]

Mission 17

Minmus and Duna systems proving

We need to start testing some of the Duna equipment under real conditions. I had originally envisioned the Minmus landing to involve 3 kerbals, as prep for the Duna mission, but that has been pushed back to a later mission in favor of trying out the Duna lander. I will also be testing the science pod recovery system designed for the Duna mission - rather then carry the parachutes all the way from Kerbal to the surface of Duna/Ike, the parachutes are kept in Kerbal orbit and only attached on return (this also gets around any issues with the pods becoming seperated during descent, I can send them one at a time).

The actual flight to Minmus requires a bit more delta-v then a flight to Ike, so I've attached a drop tank to the top of the lander which is probably enough. After the burn for Minmus this tank will be released and the mission will go on with the lander. Choosing who to pilot the mission was a bit tricky. I still need to expand my program by at least 1 kerbal before Duna, but I wasn't sure about giving the first Minmus landing to a rookie. I eventually figured out the perfect kerbal for the job, Bob. Bob Kerman was the very first kerbonaut chosen for the Duna flight, to act as commander of the Crew Ship thanks to his skills in navigation. However this means Bob will be the only one to remain in orbit of Duna while 2 other crew members make a trip to Duna's surface and Ike. To make it up to Bob, and since it will help ensure a rested crew is available for Duna, I've chosen him as the pilot for my first Minmus mission. This will be both the first flyby and first landing, so the large science capabilty of the Duna lander will be put to use.

The flight plan involves launch into orbit using the upper stage of the 30t lifter, which has been outfitted with a docking structure (if you look carefully it's not fuel, it's an empty fuselage) and 3 recovery pods. The lander will seperate and perform a burn toward Minmus after which the drop tank will be released. On approach to Minmus 2 science pods will be used and crew and EVA reports will be done in high and low orbit before landing, where the final science pod will be used and a surface EVA and sample collection will be performed. The lander will then ascend to Minmus orbit and return to Kerbin, where it will dock with the lifter to allow the recovery pods to pickup each science pod and return it to Kerbin. After that the 30t lifter will push itself and the lander into a deorbit vector aimed at the desert. The lander will seperate and aim for a soft landing there in the hopes of sample collection.

The night launch goes smoothly into a 132x137km orbit, the 30t lifter proves itself a very stable launch platform. The only hitch was that the attachment setup resulted in the lifter draining fuel from the lander, fuel had to be transfered before it could begin its mission.

The entire rig is placed on a inclined orbit like Minus, with the Mun getting into the photo again (this keeps happening throughout this mission)

The burn to Minmus is made and then we get rid of the fuel tank. Since the fuel tank is also on a course to Minmus I put some serious distance by tossing it instead of just releasing - turn off SAS, give the ship a little push to start it spinning, then release the tank when the ship is facing retrograde. The fuel tank is flung off behind (into a slightly lower orbit) and I use RCS to make a minor course correction (equal and opposite reactions - releasing the tank also pushed the ship slightly prograde). The 3 day trip to Minmus starts.

At Minmus we begin our science observations and settle into a 30km orbit. Bob picks out a landing site.

Bob begins his descent to Minmus. Outside Kerbin and the Mun slowly drop below the horizon. The radio blackout starts as he drops below 10km of the surface.

Bob lands and prepares to step off the lander.

Bob is on Minmus. This is a very different world from Kerbin.

The flag is planted and science is done. Science makes Bob hungry.

Bob begins the trip home though on getting to Kerbin orbit and lowering his PE to the desired level he runs into a problem, he is very tight on fuel. It seems like he would just barely have enough to make the correct orbit, but he needs to dock and that burn can't be predicted down to the m/s yet. For safety reasons mission control chooses to override this part of the mission - Bob will burn a small amount in order to line up his next PE pass to come within 1.7km of the docking rig. The docking rig will then perform a burn to raise its AP to 444km to match Bob, since it has plenty of fuel.

With the rig now in range the lander performs the RCS docking, passing around to the back side of the rig and bolting on. The rig now brings the orbit back down.

The mission is a little more then half way done. We still need to recover the science pods. The recovery pods are undocked one by one and attached to science pods until all 3 are ready.

Each pod seperates and performs a deorbit burn using RCS. Reentry follows and the chute is opened for a soft landing.

All 3 pods are recovered in this way, it takes a while since there are limits on time warping.

A course is plotted for the desert and the rig is used to make the burn (lander would be more efficient on fuel, but the rig has much more thrust for a shorter and more accurate burn). They seperate and descend. You can see the solar panels breaking off the rig. The chutes on the lander open and it comes in for a gentle landing, while the rig smashes into the surface.

After attempts by many other kerbals and probes, Bob has landed in the desert. And it looks like some part of the rig survived too.

Bob takes a sample and plants the flag, then goes off to examine that debris before the recovery.

About 1000 science is returned overall, allowing advances in nuclear propulsion and science instruments for measuring earthquakes. It also put us within reach of an instrument that can measure gravity, we'll want that before we leave for Duna. The Crew Ship still needs a shake down and we still need to run a 3 kerbal mission, so I'll probably plan a return to the Mun with a 3 kerbal party - a new recruit, Bill and Tomvin. None of them have landed on the Mun.

[Dave Kerbin]

Little Clean Up

That biome probe that got stuck in polar orbit has been bugging me, this is a short little mission to go and deorbit it. Our ship is some girders, a size adapter for the metal plate at the bottom and landing gear that locks the cage closed. It has RCS and a small drive unit of its own, but will be mounted to the 30t lander where it can use its big skipper engine and fuel tank for some fast orbital movements.

We launch into a low polar orbit of our own and go through about an orbit and a half before we've got a reasonable window to intercept that junk. The course change to intercept is on all 3 axis which requires a bit too keep track of - got to be careful you don't accidently plot a course through the planet which was why we didn't have a window 1.5 orbits ago. The only complication was that with all that crazy polar orbit I got mixed up over which direction the ship was going, and so while the course didn't go through the planet, it did take the ship to 52km PE on the way to the intercept. I turned this into an advantage, as the mighty skipper engine let me easily keep orbit during my dip in the upper atmosphere (the solar panels on the second stage where closed just in case) and I was able to reduce the distance of the intercept.

Aligning their courses is easy, as we close on the intercept point the relative velocity is only 800m/s - the skipper engine can handle this. Courses are aligned with 10m/s relative velocity left, enough that stage 2 can easily catch up and capture now.

Capture is a little more challenging. As it turns out the dimensions for the cage are very, very close to the probe. In fact the probe can only fit in if inserted almost perfectly straight.

After almost two and a half minutes of pushing it around I finally get a nice clean capture, it tumbles perfectly into the waiting jaws. The cage is closed and the probe is trapped.

And we turn around and deorbit the whole thing. Of physics warp doesn't simulate individual parts, just single bodies so when we return to 1x the probe is now a good 300m away but it's already falling into the atmosphere so it's ok.

[Andersenman]

I suppose it was used for Science, too, after the nudge, wasn't it?

[Dave Kerbin]

I suppose it was used for Science, too, after the nudge, wasn't it?

Unfortunately the power system had long since died, and since it used a decoupler to seperate from its original mothership (I didn't have docking ports yet) there was no way to give it a jump once I recaptured it.

[Andersenman]

Argh, I overlooked the lack of solar cells. Pity.

[Dave Kerbin]

Mission 18

Crew Ship Test / Mun Landing Themepark ride

This report is going to be missing a few parts. I had a weird hard crash while trying to return to an object from the tracking station. When I started back up I forgot to record video because I was too preoccupied on whether my save was damaged. The save was fine but I never got footage for the last section of the mission, and I found out the crash gave me a partially corrupted video for the middle part. Copying video headers with a hex editor and letting virtualdub manually rekey recovered some footage, and I eventually remembered I had a safety save (they have been used before - in Bob's mission I turned on 2x physics warp to walk faster to that piece of debris, which caused the lander to spontaneously explode!). Using the safety save I was able to replay the last portion of the mission, so after we get back into Kerbal orbit with a 70km PE you're actually seeing screenshots from a second run through. It went roughly the same as the first (since it was just a pair of tiny deorbit burns, seperations and parachute openings), though the bigger crew pod came in a bit shorter and touched down on land instead of water.

The main goal of this mission is a shakedown cruise for the crew ship that will fly to Duna. It will hopefully identify any mistakes made in design (did we forget solar panels? Is there enough battery life?) and provide data on real flight charactistics - for example this flight will include the crew ship docking with a stationary target using the reaction wheel and RCS systems. When it arrives at Duna such a docking will be required in order to get the fuel needed to come home. I also want to test a more advanced science pod recovery system. While I thought they where really neat the individual recovery drones revealed themselves to be exhausting, since the kerbal physics system required following each one down in real time. Since the Duna lander already passed low altitude shock tests with 3 science pods attached I decided I would try out a recovery system based on that shape.

We'll be flying to the Mun with a crew of 3 in the Duna Crew Ship. Before leaving orbit the ship will dock with a Mun landing rig consisting of a refuel tank/recovery system, 3 science pods and a micro lander. The lander is not as small as it could be (replacing the T400 with a T200+T100 would provide enough delta-v for a landing) but the extra fuel provides a safety margin for little cost. In the image the lander is on the bottom, with the recovery system (a T800 tank with parachutes and docking ports) above it. The crew ship has been seen before and hasn't undergone any change other then being mounted on the same 30t lifter as the Mun rig.

Bill is being given a chance to prove himself on this mission (of course it means he won't be on the Duna mission due to crew rotation). He will be taking rookies Lanfield and Dudvey along with him. I already have Bob, Lemgun and Tomvin selected for Duna, and I'm hesitant to put Geofmin in any danger, I need his brains for the design of future missions. The rookies look promising and if they survive a mission with Bill they should make good additions to the program.

After the rig is launched into orbit the crew ship is launched and things look good. Lanfield shows no issues with space flight, though I don't think that Dudvey realized he would be going into space on the same day he applied for a job at KSC.

Lanfield has a bit of trouble docking at first, the ship's tanks are almost full right now making it a bit heavy to turn. Some advice from the ground is given - 'just do it by the numbers' - and he gets things sorted out, pulling off a beautiful straight on dock (no wiggle on capture) complete with a 10 second long drift at the end with no additional corrections needed. Despite burning a lot of RCS at first trying to move the fuel heavy ship there is still half a tank left (and at Duna only the one docking will be needed). The first mission goal has been completed and the crew ship is ready for docking at Duna. The Mun rig is detached from the lifter stage (which had been providing power and control in case they where needed) and we get ready for a burn to the Mun. One small hitch is that there is a lot of torque on the rig when turning - the center of mass is well inside the crew ship and there is a 5 ton lander docked on the other end, putting the docking ports under a lot of stress with noticible wobble. Mission control advices that the crew minimize SAS usage while changing orientation to minimize the amount of rotational acceleration being applied.

We have LM extraction.

After deorbiting the lifter stage we burn for the Mun and then the crew (minus Bill who stays in the flight pod) can move into the habitat module by ladder for a comfortable ride.

It's a field trip to the Mun!

It wasn't planned for the Mun to be in the window, I only discovered this while reviewing the footage for a good shot of the cabin interior.

We arrive at the Mun and settle into a 24x47km orbit, with the PE over the light side (the higher AP is somewhat intentional, to help speed up the arrival of intercept windows for docking). Lanfield is the first out to transfer over to the lander.

He needs to undock and then pickup a science pod before he can deorbit and land on the Mun. A big unexplored crater is his target.

Lanfield descends, 9000m and closing. At 6000m power conservation becomes important, SAS use will need to be conserved until landing. With 40 units of juice left in the battery we set down, with some RCS helping to correct a small tip.

Lanfield on the Mun at the Northwest Crater.

Ascent and orbit is normal, with the sun appearing again for power. We dock the science pod back in its port and Lanfield returns to the crew ship.

We refuel the lander (the tank was about 30/180) from the rig tank and Dudvey transfers over. He undocks the lander and moves over to grab the next science pod.

Dudvey flies to the next and only remaining large crater near the Mun's equater.

Dudvey lands at the Mun's Farside Crater.

I have up to Dudvey returning to the ship and docking, transfering over to the crew ship and then some hypothetical maneuvers to put the ship on an inclined orbit, followed by about 10 minutes of the pause menu while I went and thought about the merits of using fuel to try and get one of the big polar craters or try my luck on one of the smaller ones.

Unfortunately this is where the lost footage of my mission is. From here Bill went down to one of the smaller craters just below the equater. From the science recovery at the end I can identify it as the Midlands Crater. He actually made the best landing of the three in terms of flat landing site and minimal horizontal velocity at touchdown. Bill's science instruments where also able to detect an earthquake on the Mun (thanks to the new seismic sensor being carried on this mission) After he returned mission control asked Bill to move the lander back to its original position, then asked Lanfield pilot the crew ship out of Mun orbit and back into Kerbin orbit with Bill still in the lander (without a probe core the recovery system would need a pilot, they just didn't tell Bill that before the mission). Once in Kerbin orbit at the AP a burn was made to lower PE to just above 70km.

From here I have footage from the redo, which differs only partially from the original. Bill's part played out almost identically since it involved a simple burn right at the beginning followed by coasting. Bill seperates from the crew ship, carrying with him the entire science rig attached to the lander. 83 fuel is left over from the 3 Mun landings (the rig carried enough fuel to refill the lander twice) which is more then enough for a deorbit burn. After time warp Bill finds himself at 69km and falling over the desert where Bob landed on the last mission.

Of course Bill has a lot of velocity to bleed off and is coming in on a shallow approach, so it looks like this for a long while as he flys past the desert and out to the ocean.

The lander body is ejected and the chutes are opened. All the elements for recovery reach the water safely. 1285 science from 3 landings.

The crew ship is brought back the same way. The solar panels are ripped off during reentry and in this redo the crew module comes a bit short, landing in the mountains instead of the water.

It seems we aren't too far from Bob's desert landing (about 80km away). Despite landing on the side of the hill instead of the nice valley everything seems to go okay, we don't get any sliding or worse tumbling.

So the mission was a success in game, though with some technical problems outside. The crew ship is ready for use and a more efficient recovery option is available.

[Dave Kerbin]

Mission 19

Gravity Survey

This is the final mission before the Duna mission. It also intercuts a bit with the Eve mission, I had to switch back and forth a few times to ensure some time critical maneuvers where done. In this mission Kerbal scientists want to get some real world data for the last two pieces of technology fitted on the Eve mission that have not received any field testing - the nuclear engines and the brand new GRAVMAX Negative Gravioli Detector that was fitted to the science pods at the last minute. So I designed a simple nuclear powered orbital probe outfitted with 16 GRAVMAX instruments to collect data about Kerbin and nearby objects and return it to KSC.

The ship is a simple design, over engineered a bit on fuel (>3000dv). Because it is light it is sent up on the Mun launcher platform. I've shortened the launchers orbiter stage and added deorbiting boosters to it, it really just needs to get the probe into a stable orbit above the atmosphere so the nuclear engine can be safely lit up. The probe core is chosen specifically because of its abundance of space for mounting GRAVMAX sensors. The launch is at night into a polar orbit. The data starts coming in even before a stable orbit is achieved, with information on the Kerbin Highlands, Mountains, Shore, Water and Ice Caps. Over the north pole we get ready to stabilize orbit and eject the last launch stage so we can fire up our nuclear engine.

We eject the last booster stage and send it back into a suborbital trajectory and then after a few system tests, including gravity scans of Kerbin's Tundra, we pop off the lead shielding from the nuclear engine.

As we come down the other side of Kerbin we record data on the gravity over Kerbin's grasslands before plotting a new course. Neither the Mun or Minmus are lined up for a straight intercept so we're going to push far out from Kerbin to just beyond its SOI where we can take some measurements of Kerbin's gravity in relation to the sun. It will take 2 days of travel.

After measurements we make a burn to get back into Kerbin's SOI, which will take a day and a half. After getting back another burn is used to help flatten us out and prepare for the next course change.

Our next target is the Mun where we enter orbit and take gravity readings over a number of craters. The number of sensors with storage left is rapidly decreasing so the Mun charting is cut short, we still want to take 1 or 2 readings from Minmus. The approach from deep space to the Mun has taken over 4 days, putting mission time at 8 days so far.

We exit the Mun's orbit into a wide Kerbin orbit where we can adjust our inclination to match Minmus and set a course for it. It will take 15 hours to reach our burn point and another 4 days to reach Minmus from the wide sweeping path we are taking. On approach we make some minor course adjustments to lower our PE to just 24km and flatten our ejection angle for when we pass back out of Minmus SOI.

13 days into the mission and we are passing very close to Bob's Minmus landing site, very convient as we will be able to match our gravity readings to his missions earlier observations.

After passing back out of Minmus SOI we begin planning our return to Kerbin. It will be 5 days before our return burn and other 4 before we reach Kerbin itself. In the mean time mission control switches over to the Eve probe for a final course correction to set its PE for aerobraking. The probe will reach Eve's SOI in 7 days. With that in mind we set our 5 day burn to take the gravity probe to just 70km over Kerbin - this way it is still in stable orbit should we be unable to return to it, but only a small push will be needed to get into a descent.

After getting things sorted on Eve we return with 2 days left before the probes 70km pass. A small burn is all that is needed to lower it into a landing vector. The probe detaches from the service stage, dropping the nuclear engine deep in the ocean north of the crater gulf while the probe safely parachutes for a recovery. Total science returned is 804.

[Dave Kerbin]

Mission 15 - Phase 2

Eve Unmanned Exploration

Phase 2 of the Eve mission has actually been played out inbetween parts of mission 19 and the prep work for mission 20, due to the time sensitivity of all 3 missions. In phase 2 the Eve probe is on approach where we revisit it briefly on day 31 of its mission. What we need to do at this point is alter its course ever so slightly so that instead of crashing straight into Eve we skim the atmosphere at about 60km to help slow down. With a tiny nudge - 1.1m/s - we adjust the approach and get a PE of 59km. We'll be back in a week when it reaches Eve.

7 days later we arrive at Eve's SOI. Eve is easily spotted due to its unique purple color but despite Gilly being close by the tiny moon proves difficult to spot. After 4 hours we are as close as we're going to get to Gilly before the landing probe and we go hunting for it again. Finally Gilly is spotted, a tiny point of sparkling light reflected as the oddly shaped rock tumbles through space.

As we approach we observe that Eve appears to have suffered extension damage from past impacts. On the right an impact crater even larger then Kerbin's great gulf crater can be seen.

Eve Explorer 1 brakes for everything it is worth. We've never done this maneuver before and need to get it right or there won't be enough fuel left to correct. At the same time we want to ensure we don't cut into the fuel that will be needed for the Gilly probe. Finally we get an orbit around Eve which quickly shrinks as the atmosphere continues to slow us down. Smaller and smaller it gets - if the AP drops below a certain point it will create a feedback loop where the probe will never leave the upper atmosphere, it will just keep tugging down the AP ahead of it until it crashes. The probe finally emerges from the atmosphere with an AP just under 4000km, a 2.5 hour orbit. Excluding the Gilly probe's tank we have 38L of fuel left.

Another 3L of fuel are burned at the AP to raise the PE above the atmosphere, stablizing the probes orbit. We take some scientific readings and then prepare the first descent probe for launch. The solid boosters should give the probe about a 100m/s kick, so a maneuver node of that magnitude is used to try and measure out the approximate launch point. Without much data on Eve's atmosphere, and frankly poor predictions even on Kerbin, it's hard to plan the exact landing zone. Our orbit means our best insertion point is aimed right at Eve's dawn terminator, with the planets current orientation favoring a water landing. As we come up on the release point that huge crater is once again visible.

And it is away with 30 seconds to the burn point. The descent probe begins autonomous function and corrects its oriention before igniting the 2 tiny solid boosters that will take it to Eve.

As we descend key readings are stored, to be transmitted once we have landed and can use the solar cells to recharge. I hope we haven't damaged the Science Jr by taking a reading up here....

The chute opens and we descend into the dawn - our backwards orbit means we go from day to dawn, and upon landing the planet will carry us right back into day. That atmosphere really is thick down here.

Not the most dignified landing but we are safely floating in one of Eve's ocean. The solar design is working as it was intended to, so that even in the sideways position at least one solar panel will have a period of several hours each day to recharge the battery. In this case we got lucky and landed it almost 90 degrees to the path of the sun, allowing us to recharge throughout the day (one panel faces the sun in the morning and one in the afternoon). Since it has no wheels or thusters, but can take readings at any time and transmit during the day, the probe is designated the first base on Eve.

At this point mission control spends some time back at Kerbin, landing the gravity probe and working on the Duna prep. When we return several hours later Eve has rotated giving us access to a new landing zone which looks to be a sizable continent. As the orbiting mothership approachs the drop point it rolls over so that the decoupler fires toward rather then away from Eve (ideally it would actually align to push the probe retrograde, but due to the timing I would prefer to keep the probe roughly aligned the way it will need to be for the deorbit burn.

The descent starts out looking good, but as we start to see through the thick atmosphere some big trouble appears - an inland sea!

As we pass over the sea the heavy drag kicks in, threatening to shorten our approach and send us into the sea. We start taking readings and just hope for the best. Fortunately things do turn out ok - at 30km the worst of the drag is over and we are passing safely over the far shore toward what looks to be land.

As before we descend into the dawn but this time we've actually shot a little farther - at 10km the batteries stop charging as we pass into the pre-dawn side of the terminator, the sun obscured by the horizon.

You can see how we are straddling the line in this panorama.

We land safely in a purple desert where the probe will patiently wait for dawn so it can transmit its findings back to Kerbin.

And this is the end of phase 2. Phase 3 of the Eve exploration will begin once the Duna mission is assembled and launched. It will consist of the ship remaining in orbit setting up a transfer to Gilly and then abandoning the interplantery drive module to land. If there is enough fuel, and if it can land upright, it will be able to make 1 or more short hops to study other areas of Gilly.

Edited November 5, 2013 by Dave Kerbin

[ramnrmeul]

Landing upright? Shouldn't matter. The torque of the probe should be more than enough to right the craft if it falls over. Also, you can use the torque to roll over Gilly's surface if you are on your side. Just a note, low orbit around Gilly is under 5 km (If I remember Scott Manley's statement correctly).

[Andersenman]

Hey Dave, it's been over a month. How's your programme coming?

Best,

A.

[Dave Kerbin]

I've been off a bit and doing some other challenges and have fallen behind in posting updates.

I did launch, assemble and send off my Duna cargo ship and Duna Crew ship and they've reached Duna and docked. I've set my game to not auto update (Stream edition) so that it won't switch over to .23 tommorow. Maybe I'll get the energy to quickly finish the Duna mission and post it before downloading .23.

After Duna I should have the tech tree complete, and with the .23 update I'll be starting a new career game with the new tech tree/science changes. I'll be taking a slightly different approach. First I've been researching how to modify the game file to enable the various hard mode options (disable respawn, disable quick save, disable revert, disable destroy from observatory). I've just found it too tempting to use various save shortcuts and I end up feeling like I've cheated myself out of some interesting gameplay when things go wrong. The other thing I'm going to do is lay down a specific program plan before I see the new tech tree layout - it will be a series of goals set out by the Kongressional oversight committee for the Kerbal Space Program. For example goal 1: Fly to edge of atmosphere (40km) and return safely; goal 2: Orbit Kerbin; goal 10: Duna landing, etc. I will then try and complete those goals in rough order, without any unbudgeted side trips to farm science (since every death is permanent I will still be looking for ways to test technology before using it, but this time expect far fewer Mun landings).

This way I'll be challenged to complete missions with lower tech (without getting into any silly tier 0 stuff that requires perfect flying) - for example tonight I was making some interesting plans for how my Mun launcher, combined with a relatively small amount of the tech tree (no nukes, skippers or mainsails) could put up a 10.5t ship capable of flying to Duna, landing a Kerbal and returning him home and also depositing a science station.

Edit: And here is that hypothetical Duna ship (Without launcher. It weighs just over 10t so in theory my Mun launcher could lift it). It could be layed out a little different to make it more stable during launch (change or remove those arms, or add struts near the docking port).

Tech tree wise you don't need much - I think only the lander legs are a bit out, otherwise if you have RCS, docking ports and the small lander can you can build it.

It also does fast burns - I think the longest burn you need to make is about 120 seconds.

After you get into orbit the plan is that you burn all 3 engines (the one on the very bottom and the 2 side ones) to setup the transfer to Duna. That will use up most of the fuel in the 2 side tanks. At Duna you aerobrake and then use a little more of that fuel to circulize at about 60km. From there the bottom portion undocks with the Kerbal inside (make sure the fuel tank is full, transfer from the side tanks if needed). The top portion under control of the Stayputnick stays in orbit while the little lander does a parachute descent (it has 4 chutes for a 2.55t ship, it should easily make a landing using only a drop of fuel to lower PE into the atmosphere and let drag do the rest. During descent a crew report can be filed (there is an antenna balanced on the side opposite the ladder). On the surface the Kerbal makes an EVA report and collects samples and sends another crew report. He then blasts off into orbit.

Once in orbit the probe section should setup an encounter for docking (we want to use up any fuel left in the side tanks) and uses its RCS system to dock with the lander (the lander is the target in this case since it has no RCS system). Once docked all the fuel from that tank just above the docking port (the one with solar panels and chutes) is transfered to the lander. The lander then detaches and with the full tank of fuel does a burn to send itself back to Kerbin (there should actually be just enough to actually eject from Duna and perform a powered capture at Kerbin, though we'll just do a straight 'impact' capture at Kerbin by flying right at the planet).

With the Kerbal and his samples safely on their way back to Kerbin the probe section ejects the now spent fuel tanks and uses RCS to put itself into a declining orbit (if there is any fuel left it can use that and then eject them) and closes up the solar panels for safe keeping. The probe section descends on parachute like the lander and ideally lands on the two goo containers. However it is designed for a soft roll over and with 2 antenna it can operate from any position with 2 solar panels extended. At this point it has access to any science instruments you've unlocked in the tech tree (I've attached all of them, though realistically you probably don't have the GRAVMAX).

So all the Delta V needed to go to Duna, land and return using just a little 350 isp engine.

Edited December 17, 2013 by Dave Kerbin

[Dave Kerbin]

Ok, some off topic stuff I've been making in the past month just because. These are two related ships, essentially a Soyuz/Progress pair in that one is a cheap dependable crew shuttle for 3 and the other is a cargo variation.

The main goal of this was to create a ship that was relatively efficient, safe and most importantly aesthetically pleasing (at least to me) in form and function, with some nods to the real world requirements that don't exist yet in KSP.

This is the crew shuttle, designed to ferry 3 Kerbals safely to just about anywhere in Kerbin orbit short of the Mun and/or return them to the surface. It is a simple 3 stage design with no asparagus staging. Solid booster stage, liquid fuel suborbital injection stage and final orbiter stage. Point #1 for aesthetics is that it's a relatively short, uncomplicated looking ship. No 30 story maze of boosters. You'll also notice the surface is clean of any sufficient drag inducing or breakable parts (the RCS ports are exposed, but then they where exposed on the Apollo/Saturn launch vehicle too).

It can stand up with or without launch clamps. Launch is as easy as turning on SAS and lighting off the boosters. In the event of a failure the short design means the boosters easily clear the body without converging and hitting the crew pod or any other part of the ship. I actually tested that under a number of fuel and flight situations (ranging just off the pad to high speed in a spin). After the solid boosters are done the liquid stage lights up at full power for a lazy gravity turn. Like a lot of Kerbal space ships and indeed ships in the real world (like the space shuttle) it coasts a bit after getting suborbital in order to burn at the AP for maximum effect. With this ship the liquid booster stage runs out at just the right time for this coast, allowing it to be ejected and fall back to Kerbin's ocean on its suborbital path.

The crew ship has a safe 1.5-2 minutes to plan an orbit burn which will usuallly start about 25 seconds from the AP (50 second burn), but what do they burn with?

Well hidden in the hollow space of that seperator are the engines - 6 of them in total which provides enough TWR to lift off on Kerbin. In fact the TWR is good enough that it can actually function as an abort if the staging is activated too soon and the booster is unhooked before shutdown, lifting the crew pod away from the out of control booster. The orbital burn itself can be made with just 2 paired engines in the event that one or more failed or where damaged during take off. Action group 1 is also pressed, which deploys the solar array, high gain antenna and unshields the docking port.

After orbit is achieved the crew shuttle has plenty of fuel left, from 500-800 delta-v in total. It also has RCS and an automated guidance system capable of operating for over 5 hours between solar rechages, allowing an empty crew shuttle to be sent up to retrieve a crew. The crew shuttle can easily dock with a station or link up with a refueler or other platform (for example a mission package could be sent up on a seperate launch and then attached to the docking port in orbit). For docking action group 2 allows the solar panels and high gain antenna to be retracted for safety.

For the return we need to worry about reentry heat. Action group 1 packs everything away again before we make a deorbit burn (this can be done with liquid fuel or RCS, the RCS tank carries about 120m/s). Once we are on a suborbital path back to Kerbin the service module is ejected to burn up on reentry (well in KSP it will crash about 1km away from the soft landing) at which point more inner workings are revealed. The service module contains batteries and RCS tanks in the unused space in the decoupler ring. There is still 60 minutes of power left in the pod. The rest of the reentry is done with SAS off, allowing the pods heat shield to align correctly and take the force.

After we slow down to terminal velocity the pyros are fired, releasing the protective reentry cover and deploying both the regular and backup chute (the automated control system is also lost at this point, the ship having nothing left to control). Normal landing is soft, though a high altitude landing with only one chute would be rough. Testing showed such a worst case landing wouldn't be too far off a Soyuz hard landing, though according to NASA research from the 60s would be enough to cause minor injuries at least some of the time (block II capsules used on all the Apollo missions had crash couches designed to cushion the crew to these survivable speeds in the event of a solid landing as might happen during a launch abort).

Once we set down, either on water or land, action group 3 can be used to deploy tracking beacons to aid in speedy recovery of a live crew.

[Dave Kerbin]

And here is the Progress counter part, the fuel shuttle.

Without a live crew we can go a little more toward efficiency over safety. To get more fuel to orbit 2 more solid boosters are used which increases the stress during launch.

It also has only half the engines for the orbital module, so burn time is double (start burning at 50s to AP for 100 seconds). This also means a launch escape is impossible (of course there is no crew to save) and any engine failure will result in the fuel shuttle falling back and crashing into the ocean.

During the shorter coast period there is also more to get done. In addition to setting up the burn time the craft also needs to turn 180 degrees (this is easy, the craft has a powerful reaction wheel that lets it turn very quickly) before using action group 1 to deploy the solar panels, the high gain antenna and discard the launch shroud away from the direction of the orbital burn.

The fuel ship is designed to either dock with an existing space station with its RCS thrusters to bring up supplies, or operate as its own fuel station. This fuel station role is supported by the pair of docking lights to assist other space craft (they come on automatically with action group 1). The fuel shuttle is not designed to survive reentry. It is intended to bring fuel and RCS into orbit and then be disposed of in the atmosphere. On an 80km orbit it has brought up just over 500L of liquid fuel and 160 units of mono propellent.

It can also be used in a number of other roles. For example it can be used as a tug to rescue or assist other ships in orbit (as long as they have standard sized docking ports). It can also be used as a range extender for the crew shuttle - instead of simply docking and refueling, the crew shuttle can dock with the fuel shuttle to form a single ship with almost double the fuel the crew shuttle could carry. Minus its fuel cargo the fuel shuttle adds very little weight in this role so the delta-v gain is large - if the crew shuttle had landing gear it could land and return from the Mun using the fuel shuttle as a range extender.

[Dave Kerbin]

Mission 15 - Phase 3

Eve Unmanned Exploration: Gilly

This is the final leg of the Eve mission which was started way back when I had only visited the Mun and had only about a dozen items unlocked on the tech tree. With both probes deposited on Eve's surface the mothership will launch toward Gilly before discarding the interplanetary drive and fuel tank, transforming into a lander for the tiny moon.

Getting an encounter with Gilly was a little challenging. I could get a close approach, but actually getting close enough to enter Gilly's tiny SOI took a lot of tweaking. It didn't help that I was on a deeply inclined orbit and didn't want to spend the fuel to match Gilly. Even with the encounter setup some challenges are already apparent - the probe will have less then 10 minutes in Gilly's SOI to break and enter orbit.

The 1000m/s burn to Gilly is started with the interplanetary motor so that we can burn off every drop of fuel we've managed to save over the mission (about 35L of liquid fuel and a matching quantity of oxydizer).

Including an additional 3m/s course correct 2/3rds of the way to Gilly we burn 35L + 22L of liquid fuel, leaving us with 68 of 90 in the tank.

After some coasting we arrive in Gilly's SOI - we are already very close to what was once a tiny dot.

I wasn't sure what to expect, and with a short encounter barely affected by Gilly's gravity I ended up doing two burns. The first burn got me into a longer safer encounter under the effects of Gilly's gravity and the second burn actually brought me into orbit. While I wait for the second burn the probe takes some readings on the conditions at the edge of Gilly's SOI and transmits them back to Kerbin. Finally a 20km x 5km orbit is achieved with 50L of liquid fuel left.

As we reach the low point (6km from 'sea' level) in our orbit the irregular surface of Gilly becomes very apparent. Some readings are taken but it seems like we should start considering a landing soon - that black spot in the lower right corner of the picture is the probe's own shadow, visible on the surface of Gilly as we fly over the surface at a relaxed but still dangerous 23m/s.

Landing proves interesting. With the very low gravity my normal attempts to break orbit seem to result in me actually gaining altitude (though showing an eventual suborbital path back to the surface). I had to burn toward the sky a few times to get things right. Another issue which I only later understood was that my navball stuck on orbit instead of switching to surface (perhaps because of the altitude, I landed on a peak about 5.7km high). Because of this the navball showed me with almost no horizontal velocity, yet the ground was clearly moving sideways at a good pace. I ultimately braked my sideways velocity by sight and landed after a few false starts where I bounced. On touchdown I must have slid at least 250m at a slow speed.

A bounce and then sliding across the surface of Gilly.

With the probe 'landed' but still sliding across the surface at about 2m/s I start doing science anyway.

My attempt to get the probe in the context of Gilly

And that's it for the Eve mission. Next is the Duna/Ike manned mission that has been assembled at the same time the Eve micro probes have been making their exploration.

[Andersenman]

navball stuck on orbit

Ha, happened to me, too, and I'm sure to many others!

I like your functional designs and your very sensible use of part clipping. Speaking of which: If you replace and (0.025 t + 0.175 t = 0.2 t) with the wider (.05 t) you save 150 kg mass per side, provided you're willing to wait with the Duna landing until the Tech Tree bestows it. (Actually, I would even use the slim decouplers and go only with a single in the first place. – Yes, I'm that avaricious!)

I did get my .23 yesterday night; unfortunately, playtime was again not included in the deal, hehe. Hopefully I'll be getting some crafts up over the holidays!

goal 1: Fly to edge of atmosphere (40km) and return safely; goal 2: Orbit Kerbin; goal 10: Duna landing

Atmosphic flight, Duna landing – Sounds like a plan to me. ... Oh, you weren't counting in binary? I was wondering already what this bit in the middle was about!

Edited December 18, 2013 by Andersenman

[Dave Kerbin]

I'm finished the Duna/Ike mission with a massive amount of science returned (I think about 5000, which was way more then the 1500 or so I needed to complete the tech tree) and the safe return of my Kerbals. I'm going to start getting it all up, I've got 34GB of footage to go through (about 250mb looks corrupt but I think I got everything I needed).

[Dave Kerbin]

Mission 20 - Duna

Overview and Assembly

The Duna mission has been in development for a long time with various systems tested in real world missions. It will send 3 kerbals to land on Duna and Ike and bring back science samples and instruments from everything they can touch. The mission itself will involve 2 interplanetary spaceships along with a lander carried along. One is a crew ship with accommodations for a long term flight and powered by a conventional engine. The other ship is nuclear powered so it will only be under thrust when Kerbals are far away. This cargo/support ship will carry fuel for the crew ships return voyage, science instruments and the lander to Duna and back.

The crew ship has already been tested in full in Mission 18. It uses a convential engine to carry 3 kerbals and has a command pod, where the crew go for launch and splashdown, and a habitat module for them to live in during the extended flight.

The core of the cargo ship is a quite skeletal. A central T800 tank has girders to support 3 surrounding tanks. Those tanks carry nuclear engines and the fuel to run them. The top of those tanks contain mounting points for science pods, which have already undergone testing in previous missions. These outer tanks also house the solar arrays and RCS thrusters away from everything else. The central span has docking points for additional fuel tanks - 3 radially, 1 on top and 1 on the bottom. Other support systems are also present on the central span including monopropellent storage, batteries, lighting, attitude control and flight computer. The core of the cargo ship weighs about 25 tons so it can be lifted with the standard 30t lifter created for the Duna program and used in several previous missions. The cargo ship is deposited with full tanks in a 200km parking orbit to await additional attachments. After the lifter has detached itself it deorbits.

Next to launch is a fuel ship. I had created another design for this ship, but with the performance of the 30t lifter (and its ability to ferry cargo right up to a renderous instead of just to orbit) I created a larger version of the fuel ship to carry 3 modules instead of 2. It still falls under the 30t weight limit because it ditches the previous renderous module and instead relies on the lifter for everything but the last 100m or so, which is carried out by a micro tug carried on top.

The micro tug is essentially a docking port attached to a round RCS tank with 4 thrusters and guided by a tiny probe core and battery with a single solar panel on the back for trickle charging. It was initially a bit challenging to use the tug (the RCS ports are as far from the center of mass as they could be) but after wasting monopropellent on the first 2 tanks I got the hang of it, piloting it more like an RCS-less docking. I also found that small translations could be done accurately by shutting off SAS, making the translation and then using torque to correct my pitch.

The first run put up 3 tanks to go on the radial attachment points. I brought the tug back to the lifter for disposal before realizing the lifter didn't have a port for it to dock to. I disposed of the tug using RCS to deorbit and use thrust to deorbit the lifter.

The second run brought up another 3 tanks and it's at this point I should talk about the fuel "plan". The fuel plan was drawn up without much data about aero-capture, so I built in a lot of delta-v for 'capture' scenarios. I also built in a minimum 10% safety on all the requirements. The plan for fuel looked something like this:

- Cargo ship burns for Duna, using up the fuel in the 3 T800 tanks on the engines with a safety margin that calls for a little fuel from the other tanks too.

- A little over 1.5 drop tanks (1 drop tank is equal to 2 T800 tanks) are depleted to refuel those T800 tanks and the 1 empty drop tank is released in deep space (weight reduced by further 0.6t in advance of capture).

- Capture at Duna which burns slightly less fuel but enough to use up what remained so that we finish depleting the second drop tank and deplete a third. We've burned 3 of the 5 drop tanks so far if we keep moving fuel into the T800 tanks

- The crew ship flies to Duna, using most of its fuel. It docks with the cargo ship and depletes the last 2 drop tanks to refuel itself for the return journey.

- The lander goes to Duna, using up its fuel supply. The center T800 tank is depleted to refuel the lander for the Ike landing, leaving only the fuel in the three engine T800 tanks

- The cargo ship discards all the drop tanks and returns using the fuel in the three T800 tanks

Since I was bringing up 3 tanks anyway I decided to simply add the extra tank to the cargo ship, it could be safely attached to the tank on the bottom. It would obviously add weight but the fuel would counter the weight addition at the cost of slightly longer burns and the need to transfer some fuel during the first burn.

With some more experience I moved two drop tanks at once on the first go (they get stacked on the bottom port so there isn't a reason to seperate them). With what I learned the move goes smoothly. The last tank is attached to the top of the cargo ship and the micro tug is refueled from some tanks on the lifter before returning and docking with the cargo ship. At one point the plan called for using the lander as a tug to move around science modules and tanks but that idea was dropped.

The only thing left now is the lander. The lander was tested back in mission 17. To bring it to the cargo ship it will be launched a little bit different. It is attached upside down with a microtug linked to a shroud on the engine. This way the lander can be docked without a pilot (the lander can only be flown manually, to save power and weight it has no computer control).

The landers delivery micro tug tops up all the monopropllent tanks on the cargo ship and its own tug before removing the shroud. Now all that is needed to leave is to release the aerodynamic caps from the science modules and burn for Duna when the launch window arrives.

[Dave Kerbin]

Mission 20 - Duna

Transit

Both the cargo ship and crew ship will be leaving during the ideal transfer window near day 60. The cargo ship is scheduled to leave several days ahead - this way there is lots of time to capture both ships and prevents a situation where the cargo ship might come into brake in an area near the orbiting crew ship. This way we can wait a day or two after the cargo ship has used its nuclear engines and left before the crew ship goes into orbit and launches to Duna.

This is the course for the cargo ship, though the crew ship followed a nearly identical course just a few days delayed. The cargo ship is very stable for turning even with all the parts strapped on. Only the lander gives me some wobble when I adjust my angle. I abandon an empty fuel tank after the first part of the burn and because it has docking ports KSP has decided it is a probe (there is nothing for me to select to reclassify it) and continues to track it as it flies into solar orbit.

Both ships follow a standard procedure of making minor corrections after exiting Kerbal's SOI. About half way to Duna both ships make course changes to bring their expected PE under 20km. That will be refined once they enter Duna's SOI. On the map you can see Kerbin's orbit on the far left followed by the empty fuel tank (probe icon), the crew ship (pod icon), the cargo ship (lander icon) and finally Duna. Remember that we have the cargo ship selected so we only see encounters it will have. The crew ship has an encounter where its line crosses Duna's orbit, about 4 days after the cargo ship does. Once in Duna's SOI the PE for the cargo ship is dropped to just over 9km for aerobraking, we will try to get to a 60km orbit. We get an initial AP of 80km, with 43m/s required to raise our PE up to 60km.

The cargo ship settles into a 60km orbit and is lined up to match Ike's minor inclination. This will assist with the Ike landing where fuel for return will be tight.

The crew ship follows the same approach. After gaining orbit it about 6 hours for a docking window to line up but beyond the wait there is no trouble.

Edited December 19, 2013 by Dave Kerbin

[Andersenman]

Quite the ride there during capture!

[Dave Kerbin]

Quite the ride there during capture!

With the solar panels retracted for safe aerobraking I always turn off SAS to conserve the battery, hence a lot of tumbling as drag pulls the ship around looking for the sweet spot.

[Dave Kerbin]

With the crew ship docked we can review the situation. We've saved a lot of fuel by aerobraking and through efficient ejection, circulization and docking. Almost 3 drop tanks worth in fact (the crew ship only used about half its fuel supply arriving, meaning it only takes 1 drop tank to refuel it instead of 2).

After some basic orbital readings the first landing can commence. Lemgun, my crack pilot, transfers over to the lander which will descend to Duna. For this leg of the trip the science pods on the lander are still secured with struts so that they can better hold up to the strain when the parachutes open. There is a minor technical issue when the docking port won't release. This apparent bug in the cargo ships computer system will repeat several times during the mission and ultimately requires that a specific item be undocked before the intended part will be released. In this case undocking the crew ship temporarily allows the landers docking port to function.

From a tank of 360L of liquid fuel we burn 10L to put us on a long deorbit path that will take us around to the day side of the planet. The lazy descent is so we can spend as much time in the atmosphere as possible, both to collect scientific readings and to aerobrake as much as possible before opening the parachutes.

The parachutes are opened only about 800m from the ground while velocity is still 350m/s. Everything stays together thanks to the design work focused on parachute shock. The chutes fully deploy and we slow to 15m/s, now about 200m from the surface.

The final landing is powered and we touch down at a safe 6.5m/s. We've found some reasonably flat ground and the lander stays stable which was my biggest concern with this landing.

Some science readings are taken on the surface and then Lemgun mounts the ladder and prepares to set foot on Duna.

Import jobs that only a Kerbal can do are performed.

After waiting 30 minutes for the cargo ship to line up where we want it for an intercept the lander lifts off. We take some readings while in the upper atmosphere but this distraction proves to have some negative effects. We've overshot our AP.

We bring up our PE to match the cargo ship but our new orbit now means the cargo ship will overshoot the lander and we'll have to wait for it to 'catch up' by going around the planet. It takes 8 hours for them to come into alignment again and allow the lander to dock. We manage to return with 42L of liquid fuel left - a little better then planned though the Duna landing was already good for delta-v, it's Ike where things will be tight.

The final stage of the mission involves Lemgun returning to the crew ship while the lander is refueled and the micro tug goes about the chore of moving all the science pods over to the cargo ship and replacing them with fresh ones. This takes about 22 minutes of play time.

[Dave Kerbin]

With the lander prepared Tomvin can EVA over for his trip to Ike. The original design calculations for the lander show that Ike is very close on delta-v, the mission must be flown correctly or Tomvin might not be able to return. One of the original backup plans in the event that the lander could get into orbit around Duna or Ike but not get back to the cargo ship was for the crew shuttle to undock and go rescue it. Because of the large amount of extra fuel I have I am using a different method to ensure the success of the mission. The lander will carry with it one of the fuel tanks which will be filled to the 25% mark - this will increase the landers total fuel supply by 50%.

Setting a course for Ike is quite simple, the cargo ship was already on the same inclination so it's no different then a visit to the Mun.

Arriving at Ike a capture is calculated and Tomvin calls in his observations from high orbit.

After orbiting Ike at 15km the descent burn in begun to land on the daytime side in a nice big light area. After a short burn the drop tank is depleted and it is cast off. The drop tank will impact Ike farther along.

From higher up the landing site looks good, lots of flat terrain. We set down gently.

Getting out Tomvin forgets the ladder. He claims he was distracted by a UFO.

Since the ladder can't be deployed without someone in the pod Tomvin just uses his jetpack to get back in. With all the fuel in the lander the return trip is very easy. Tomvin arrives back at the cargo ship with almost 90L of liquid fuel instead of the 1 to 5 from the original fuel calculations.

[Dave Kerbin]

Next we need to get ready to depart. Everyone returns to the command module of the crew ship. The science modules are stripped from the lander and stacked on the cargo ship. The crew ship is refueled and fuel is arranged into the fewest tanks on the cargo ship.

We also have to decided what to do with all the fuel. The crew ship will take a full drop tank with it. A small amount of fuel will be used to help deorbit the lander and any other debris. That leaves one full tank and one tank about 25% full, in addition to all of the cargo ships structural tanks being full.

After the lander has been stripped the bolts holding on the lander can are disconnected, letting the body drift free. The tug comes around to move it out of the way and then retrieve the can and move it to a more appropriate location for weight balance.

A small amount of fuel is loaded into the remaining exposed radial drop tank (which we will need to remove before the cargo ship can leave) and the tug carries it over to the lander body and joins it to one of the science module ports.

From here, using the tugs computer, the landers engine is used to carefully deorbit the somewhat unbalanced collection of discarded equipment.

The crew ship leaves next. There where some issues with the docking port not responding, the crew ship itself had to be undocked, moved back a bit, and then the drop tank could be undocked and the crew ship moved back in to capture it. A course is plotted for Kerbin and the crew ship takes off. The drop tank proves to be a bit of a bother attached to the front of the ship - no adjustments can be made without the tank rocking hard on the docking port. As soon as the fuel in the tank is used up it is happily discarded. We use about 1400m/s and have close to 2600m/s left. With so much extra fuel the plan is now to capture at Kerbin to allow for an accurate landing instead of the original plan of a direct landing (fly straight at the planet).

The crew ship gets about a day past Duna's SOI and then the cargo ship follows.

After about 70 days both ships approach Kerbin. They will be arriving very close together so this will have to be managed carefully. Being closer the crew ship is the first to spot Kerbin (tip: you can right click on windows in IVA view and then use the mouse wheel to zoom in farther then you can from the external camera; from the external camera Kerbin is just a dot).

As we enter Kerbin's SOI a problem emerges. In my confidence over the fuel situation I never made any course adjustments beyond the encounter and so it was up at 40,000km. That created a major problem, since even a basic capture would now be 3500m/s, 1000m/s more then what the ship carried fuel for. So I took the alternative option where for 2000m/s I could send the crew ship on a 2 hour express trip back to Kerbin.

Unfortunately it seems I have a corrupted segment here, maybe I screwed up the recording and that's why I have a 250mb incomplete file. The crew ship itself was going at about 3.5km/s when it hit the atmosphere and it was the fastest landing I've ever done - like having time warp in the atmosphere where it took only seconds to reach parachute altitude. Everything seems to work ok and the crew landed safely, thanks to the drag physics of KSP being able to bring just about anything to terminal velocity before it impacts the surface.

The cargo ship was half a day back outside Kerbin's SOI and carrying 5500m/s of fuel, so I not only had some time to setup a partial course change (bringing the encounter to just below the Mun) but lots of delta-v to make the required capture. Without parachutes a direct approach was out of the question. After capture I made an inclination change to level it and then brought the orbit down to 150km with about 50L of liquid fuel left in the tanks. What started out as a huge amount of extra fuel turned into almost not enough (though as longas the cargo ship could make a basic capture I could have brought up a ship from Kerbin to refuel it).

Now the science pods and lander can with samples needs to be returned.

[Dave Kerbin]

I've used a copy of different return methods for science module. The original method for returning science pods was tested in mission 17 and involved an individual return drone for each science module. That approach proved to be very slow since each one had to be followed through the atmosphere. A second approach was tried in mission 18 where the refuel tank for the reusable lander included ports and parachutes, allowing it to return 3 science pods and a lander can at once.

Since that approach worked I expanded it to handle the 9 science pods coming back from Duna plus a lander can. I created 3 similar return modules based around the small fuselage (rather then a fuel tank). Each section contains 3 docking ports for science modules and 3 parachutes. The unit on top contains a top docking port for the lander can and during launch stores a tug. The bottom module includes a battery and probe core, and below that is a fuel tank, LV909 engine and some other support equipment. The whole structure has decouplers and staging so that starting at the top each section can decouple and deploy its parachutes. The whole structure is carried on top of the 30t lifter (it's not that heavy, but I wanted the lifters well tested rendezvous equipment) and there are girders to stabilize it during launch. The tug is very similar to the micro tug, but with a bigger battery and solar panels and with RCS placed closer to the expected center of mass for better handling.

The whole thing is brought up to an intercept point with the cargo ship.

The tug undocks and first grabs the lander can to go on the top of the recovery ship. After that it grabs the science modules 3 at a time from their stacks and brings them over to the recovery ship where it is easier to place them one by one (after placing one it is a simple translation to line up the placement of the next).

It takes about half an hour to move everything over to the recovery ship. The cargo ship and the tug will both deorbit to get burned up in the atmosphere. The recovery ship powers up its engine and a specific landing site is picked out for the best recovery. Drag from the atmosphere ensures we should land somewhere in those nice grasslands west of the mountains.

As we descend below 60km the drive section is ejected, we don't need the risk of any debris when we open the parachutes. The probe core and a large battery remain to control the recovery systems.

After braking our landing zone looks to be a nice valley.

The parachute sequence is started and all 3 sections seperate and deploy in sequence. After dropping another 2km the chutes fully deploy.

The module with the lander can rolls over after landing on its head but seems to come away fine. The other 2 touch down at 5.6m/s.

Including the science already earned from crew reports and the return of the crew ship we get close to 5000 science, more then enough to finish the entire tech tree with 3700 science to spare. The middle section of the recovery ship had neither a command pod or a probe core so it was listed under debris in the tracking station.