Dave's Voyage (.23 career log, PIC HEAVY, ANI GIFs)

[Dave Kerbin]

I am close to flying phase 3. As I said before it will all sort of be one mission since I'll need to switch between them to handle burns and captures at the right time, though I'll try to break it up into 3 distinct reports. In game the whole phase will take about a year from the launch of Martian to the return of Tyson.

I was asked about the design process so I'll try and share a bit of that.

Design for a mission starts with requirements gathering. First I need to know where I am going which is usually a single planet though I have figured out multiple planet missions for non-career challenges. For career the next step is figuring out what science I should be trying to collect. For a manned mission I usually aim to collect 'everything'. For the one use experiments (goo, science jr) this tells me how many I will need to carry. With the way science works in .23 (data can be moved to the command pod) I can also determine how long into the flight an experiment needs to be carried. For the reusable experiments it is usually safe to carry just one, but time/safety restraints (like descending through an atmosphere or braking) can raise the requirements since I'll need to plan on storing the results until they can be transmitted/collected.

The next step is creating my own delta-v map. I don't get fancy with graphics, it's usually just a few columns in Excel or even just scratched down on a small notepad. To create the map I use a few sources. For landing I use the common cheat sheet with some safety factor thrown in. Unfortunately I find this map loses accuracy the farther you go out - flying to Duna or Eve is fairly safe, but the inclined planets start to get really far off. In fact the values for Dres look like they would fit exactly if Dres where on a circular, non-inclined orbit. For a better idea of burns I look at the alexmoon's Launch Window Planner, which is a handy bit of javascript that calculates raw delta-v values fairly accurately. Since I fly manually a lot of the values it gives are more of a guide. Without an in game aid like Mechjeb or protractor the easiest way to use the values is to calculate a mid-course plane change maneuver (rather then the more direct ballistic). The values calculated can be used to device a stop watch flight plan. The first burn can be checked for accuracy by comparing the resulting AP/PE, while the second burn has a more obvious intercept.

Once I have some idea of the requirements (I add a safety to the delta-v) I begin building a ship. To do this I don't start up KSP, I start up Excel along with a copy of the parts list (of which I've now made several corrections since my delta-v numbers depend on it being accurate). I do know the mass and where applicable isp and thrust of several parts by memory, and importantly I know the fuel mass to container mass for fuel tanks (8:1, and also that 8 tons of fuel = 720L of liquid fuel). The ship is then built in reverse - I start from the very last stage of the mission and work backwards from there. Because it is done in Excel using some basic forumulas and I can make cascading changes - I can add or remove components and immediately see the effect on the mission.

It's not the prettiest looking but it's not intended as a strict formal blueprint, rather I treat it more like design notes and frequently have comments written in the margin. Once I've proceeded far enough that a design seems viable I'll build it in the VAB. Not all designs can be built (this is especially an issue with micro probes) though over time it becomes more intuitive as to what parts you can pick in what combination (for example ensuring enough space is available for mounting radial parts) and I can usually picture what it should roughly look like while picking the components.

Once a design is finished I'll come back and review it after spending some time away. For the review I usually start from the beginning, figuring out the requirements and then walking through the ship design. This helps find oversights (does the ship have landing gear? Did I make a typo when recording the delta-v requirement?) and can suggest design changes. One thing that generally comes last is the design of a launcher - in fact I'll usually design, iterate and build the ship in VAB several times before I get to a launcher. With the aid of copy and paste I may test multiple designs, either partial changes (keeping the upper stages but changing lower ones) or completely new designs. For example for Martian and Tyson I went through at least 6 more formal designs and probably another 6 improvement designs that where discarded when it became obvious they wouldn't be superior to the original.

One of my requirements for Martian and Tyson was that they properly reflect a budget mission. That didn't just mean that they should be light and that only one would get a nuclear engine. It also meant I wanted them to share parts, in much the way that the Apollo spacecraft and Saturn launch system was reused for many missions (orbit, moon landing, Skylab, Apollo-Soyuz and was even on the drawing board for a Venus flyby). To that end I wanted both missions to use the same LKO launcher (which meant they should have roughly the same mass) and wanted to try and use the same lander with minor, plausible differences that suggest they still came off the same assembly line and where covered by the same testing regiment.

With my spreadsheets and iterative design I went through a number of designs and a number of mission profiles. Universally the mission to Dres was the better target for the nuclear engine. I found the difference in mass between a nuclear and non-nuclear Duna mission as little as 1 ton. The 3 main profile variations I covered where a direct ascent, orbital fuel renderous and orbital transit renderous.

The simplest mission profile was direct ascent (during the early Apollo program this was also the prefered method of going to the Moon because it was the safest, but the required launcher was outside the technology of the time). In a direct ascent a single ship sheds stages as needed to reach the planet, land and then take off and fly back. In my case it was also the safest approach, particularly for the Dres mission (Tyson) where it allowed left over delta-v to be carried over throughout the whole mission.

The next mission profile was an orbital fuel renderous. In this scenario the fuel to return home is left in orbit instead of being carried all the way to the surface and then back up again. This configuration requires a docking port on top of the ship, which in turn means that the command pods parachutes need to be attached to the side. Early on this wasn't as much of an issue (the command module used the XL chute which could be cleanly split into 2 radial chutes) but later iterations brought the weight down so that only a small chute was needed on the command pod, which meant this setup required 3 times the parachute weight in order to mount a docking port. For Duna this mission profile could cut between 2 and 3 tons compared to direct ascent, however it created problems for Dres where the balance between the cost to bring mass down was less and the delta-v required for return was more, resulting in the Duna lander being too different from what the Dres lander needed. Even if they used different sized fuel cans left in orbit it would still require two distinct lander designs and I didn't want that.

The last profile came from the thoughts about parachute placement. I considered the idea of placing the docking port on the bottom of the command pod (instead of a decoupler). One thing I should note is that ever since .23 came out I've wanted to get to the point where I could build my escape pod - you'll notice there is Kerbin Land Stage with 175dV. This dV is not included in my mission totals but is kept in reserve. It is made possible because of the small amount of monopropllent carried. Not only is the small amount more attractive then the larger tanks that where previously available, but the tank weight is essentially free. This means I just need to add 2 RCS thrusters and solar panels and I have a tiny spacecraft. In the orbital transit renderous profile the ship returns to orbit but instead of docking with a fuel tank it uses the underside docking port to discard the fuel tank, engine and ladder from the lander. It then docks with the remains of the transit stage, using it's more efficient engine and remaining fuel to return.

Despite sounding good on paper this design was actually the worst. Because the payload was so little (just short of 1 ton) the mass of the return engine became a serious factor in determining how much fuel would be required. For Duna I could at best break about even on the return requirements but for the longer Dres mission the heavy nuclear engine made the reuse of the transit portion a big loser despite having more then double the ISP (2x ISP isn't worth it when the ship is more then 3x the mass and empty fuel tanks approach the weight of the intended payload).

Ultimately I think I've come down to a final lander design. The Duna mission uses the full lander while the Dres mission uses a 95% assembled version - the 3 descent parachutes are never packed (leaving just the cones) and 3 small atmospheric instruments are never installed on the outside of the material bays. The pictures aren't perfect. The first 2 are from the prototype lander which doesn't quite get the rotation of the outer pods right (they are facing 60 degrees off). The third shot is the Dres version with a bit of the lower stage visible but with the atmospheric parts left off and the pod angles corrected. The mass of both missions in LKO is just under 15 tons.

As is common in a lot of my designs the goo containers also serve as landing legs. The placement of the instruments on the pod is also not by chance. From previous missions I found that putting the instruments I wanted to get readings from on the pod was very convenient, since it made for a quick EVA to collect them (never leave the ladder). This convence overrode the dV gains by placing them on the lower section. Th gravity and temperature meter where placed by the hatch since those instruments would need to be checked several times in flight. The antenna could be placed on the back along with the seismic meter which only collected one reading and so didn't even need to be collected, it could land with the pod back on Kerbin. There was also an earlier design that went through about 2 iterations before I refined it into something that looked like the final lander (which went through at least 4 iterations though mostly in the transit design). The earlier design used 4 pods instead of 3 and had a heavier fuel load, resulting in a LKO mass of 19 tons. At one point the fuel renderous design for Duna weighed in at 12 tons in LKO. The earlier design also placed the small instruments on the science jrs, so they could be dropped later. You'll notice more instruments on the back and a second access ladder. The revised design simplified the number of instruments (just 3) and placed them low enough that they could read from the ground.

For the Icarus mission I think I have a final design but I want to share this earlier one. For the mission requirements I quickly eliminated goo containers (mass vs. science return was terrible) and slowly wittled down the number of science jrs to 0 as well. During requirements gathering I determined that Moho (I've never visited it yet) would have a short capture window based on other peoples posts so it needed to be a chemical rocket. However this earlier version uses an ion engine for the initial transit phase - it actually has over 1000m/s of extra delta-v, since I figured I wouldn't be able to do a direct burn due to the low thrust, so I'd need to thrust out of Kerbal's SOI, then slowly setup an encounter with Moho. It also is a bit different in that it launches upside down.

The two boosters (set for something like 30% thrust, the calculations are in the spreadsheet) are there to help lift up the fuel since the T45 can't carry it all. Once enough fuel is burned and they are exhausted they are released, followed later by the side fuel tanks. The T45 finishes carrying the probe to orbit where it detaches (the stack decoupler) and turns around to face the correct direction. At that point the 8 solar arrays (top of picture) around the ion drive spread out.

After entering Moho's SOI the ion drive and its solar array are ejected, revealing the chemical rocket. The T200 and T100 fuel tanks at the top of the probe (connected by a jr docking port so fuel will flow without ugly fuel lines) power the engine for the capture burn. Then those big tanks are released and only tiny probe (roughly the height of 3 oscar tanks) lands.

Edited January 9, 2014 by Dave Kerbin

[Dave Kerbin]

The missions are finally under way after a few days with little sleep I forced myself to at least get them running.

For time warping when I need to setup launches I've established a time keeping drone on top of the VAB.

So far I've done the first 30 days of what will probably be a 280 day 3 mission package, consisting of the launch and injection burn of Martian 1 and Icarus 1 and the correction burn for Martian 1. I would have probably done the correction burn for Icarus 1 (after that Tyson 1 needs to be launched) but I ran into a problem with Martian that needed to be fixed.

After the correction burn I found that my PE for Duna was 700km and it couldn't seem to be adjusted with a maneuver node (it would stay at that value or fly right off the Duna encounter). The reason was a rather big problem - Ike just happen to line up with my path to Duna. I recorded the information and went back to the spec sheets. According to the fuel usage I had at most 270m/s with which to navigate. One option was a change to the mission plan that would see some equipment be dropped in Ike orbit as I passed rather then sent up the hard way. While this would save some fuel the staging meant that I couldn't use it where I needed it, I would actually be lower on fuel for a critical ascent portion. It also messed up the staging of some science equipment which would have removed my ability to gather some data on Duna's low orbital conditions.

I didn't like the idea of having to wait until the last moment (after passing Ike which orbits very close to Duna) to figure out if I would have enough delta-v to setup the proper capture. I went looking for a setting that I knew would help give me some more information by showing additional projections. Ultimately it didn't help, maneuver nodes just where not refined enough to figure my way out.

So I switched to the experimental approach, using a design feature I specifically envisioned being used for very fine course correction. I aligned my ship to prograde and turned on RCS. The ship only has 2 RCS thrusters and 15 units of monopropellent in the command pod, but that is enough to both provide thrust when the command pod is used as a life boat and to make or test very small adjustments. With the 2 thrusters I can make forward or reverse changes at a tiny scale (tip: caplocks doesn't just turn on precision mode for turning, it also reduces the throttle for RCS). With this experimental approach I was able to nudge my course just enough to bypass Ike and lower my PE to 25km. At that point even the

thrusters couldn't provide the accuracy needed to set it to 10km (I could raise or lower about 20km with one tap of thrust).

Anyway, the report for Icarus should come soon. It is the middle mission and launched second but timing means it will be completed before Martian arrives at Duna and Tyson won't even have done a mid course correction. Martian will be a bit longer, the mission has a lot of tasks to complete as evidence by the mission plan: 5 packed pages for Project Martian, 1.5 pages for Icarus and 2.5 pages for Tyson.

[Andersenman]

Very nice! Seeing so much commitment is almost like making the trip myself! Also, those mission profile sketches are adorable!

[Dave Kerbin]

Well it looks like Icarus 1 might fail in its mission and crash into Moho.

I haven't narrowed down where the mistake was made - reviewing all the numbers I was tight (only about 200 m/s safety for the journey and 200 m/s safety for the expected landing for the delta-v I expected) but it doesn't seem like I got any of them entirely wrong. Fuel usage was also correct, my fuel and mass values matched calculations so I didn't forgot to account for any weight or botch the fuel usage. That leaves the angle of the injection burn though my orbital numbers made it seem like I had it right. Unfortunately I never got video of that section (I am getting more problems where it seems like I'm recording but no video is saved) so I can't review it for any errors made.

The simple matter is what should have been about a 10 m/s inclination adjustment to perfect my Moho capture actually required 715 m/s just to get the encounter (with a 24 hour difference between the planned arrival and actual). That means I probably don't have enough fuel to capture and land.

After burning another 1.07L of liquid fuel to lower my PE to 5km (maybe I should have stopped at 10km) I have the following:

Exactly 2900 m/s in the big tank, which I need to throttle off and rotate before ejecting

Another 1615 m/s in the lander itself

That's 4515 m/s in total.

My estimates are based on the assumption that it will take just under 3400 m/s to capture and another 1400 m/s to land. However some sources don't agree with those numbers. The 'more accurate' delta-v map believes that a landing with 810 m/s is possible, though I suspect that is a perfect suicide burn. Several delta-v maps also claim numbers closer to 2200 m/s to capture, but I think that may be based on if Moho's orbit was circular, reducing the relative velocity at intercept when transfering from Kerbin.

I have considered the idea of a 'direct' landing, where I simply fly straight at Moho and try to land without orbiting. That sounds crazy though and is probably not sound from a physics perspective (not enough dV). If I tried that I've have to consider what low orbit readings I might try to transmit before landing. The GRAVMAX uses almost all my battery power to transmit which would leave me in a tough spot especially for a night landing. A night landing is especially dangerous because I have no radar altimeter in a probe (it would actually be neat if that was an actual part rather then just being an invisible extra from command pods).

In talking it through like this I think my only real choice is to raise my PE to between 8km and 10km, do a capture burn which will probably leave me with 1000 m/s, do my orbital science safely (which isn't worth much compared to what I can do on the ground) and finally start looking for a very high altitude landing site during the day that I can try to perform a suicide burn landing on. My landing legs and engine provide a small crumple zone, but after that I'd be losing the batteries which would require me to use the timewarp transmission trick again if I wanted to send anything using the solar array to provide trickle charge.

[Dave Kerbin]

Icarus 1

Project Icarus is not the first mission in the Journey phase but Icarus 1 was planned from the start as the first that would finish. Launched a week after Martian 1 (the first project in this phase) Icarus had a much shorter trip to reach Moho and being a one way probe cut down the mission time even more.

Unfortunately Icarus 1 is the first mission in my save-disabled career file to fail completely. While a few missions have missed some of their intended side goals (such as some missing atmospheric data on Berry 1) this is the first one that did not reach the primary goal and was in fact destroyed.

As part of budget saving measures Icarus 1 was built as another ultra light probe like Berry. The materials bay and goo containers have a very low transmission rate which when combined with a relatively high mass made them unsuitable. The resulting lander design weighed in at 0.58 tons. Learning from Berry there was no trickle charging, instead Icarus carried enough battery power to transmit any single instrument reading without recharging. With no atmosphere a powered landing with legs was in order and the solar system was upgraded to articulated solar arrays. Total delta-v was 1615 m/s, which was slightly less then ideal but seemed to be within the range of safety to land the probe from low Moho orbit.

To get the probe to Moho additional fuel would be needed. Research suggested that Moho would require 3400-3800 m/s to capture and that the capture window would be very short. For that reason a light ion drive was rejected, though one was briefly considered for the transit portion. The result was a capture+transit fuel system that attached one large tank to the top of the lander with a fuel feeding docking port, with 3 smaller tanks carried radially and discarded before the capture burn. The launch system went through some revisions until a simple launcher was eventually created. I've decided to name this the Light J-Series Booster. It's composed of a central orbital insertion engine (a thrust vectoring T45) which is carried to 11km by solid boosters which are then dropped to allow the ship to make a gravity turn.

The launch of Icarus 1 early in the morning went fine with an insertion into an 80km, zero degree incline orbit (the Mun is used to measure inclination since it is also at 0 degrees, simply selecting it as a target allows the difference to be seen).

After the orbit is precisely set the injection stage is discarded. This is important because the probe is actually carried upside down. During launch the fuel feed docking port is used as the point of control to orient the navball correctly. Once the booster is discarded the navball is reset to match the probe's correct orientation.

The launch window and angle arrive and the 1500 m/s injection burn seems to go ok, with the probe coasting out of Kerbin's SOI.

The troubles begin when I come back to the probe a few days ahead of its mid course correction burn (after having just done the same for Martian 1). The correction should be about 10 m/s and result in an encounter from previous calculations, but that is not the case. In fact it takes over 700 m/s to setup the planned encounter. With little choice the burn is made. This uses up the radial tanks and reduces the delta-v in the transit tank to just 2900 m/s. While I have several sources claiming a capture requirement of 2000-2400m/s my own claimed closer to 3500 m/s.

Combined with my lander's delta-v I have 4515 m/s in total. My plan at this point becomes to make the capture burn and hope it is closer to the lower number then the higher one, then look for a very cheap landing site - I'm told the northern hemisphere is quite mountainous which could reduce the delta-v needed to land.

However even that turns out to be wishful thinking. I arrive in Moho's SOI traveling at more then 5 km/s and find it will take about 4400 m/s to capture.

After reviewing the video data I realize that there might have been a 1 in a million chance that I could land based on the lower altitude, but at the time it didn't seem like there was any good way to do that. One thing I did discover is that being so close to the Sun my solar panels where generating a huge amount of power. The reference power for a 1x6 panel at Kerbin's orbit is just 2em, but near Moho I was getting a maximum rate of 12em. This meant I could transit with no power loss (power charged faster then it was consumed). This piece of information could be useful in the future. All that is left is trying to transmit information about Moho's low orbit - a gravity scan for which I'll need to be coasting below 80km altitude and then a temperature scan.

The temperature scan finishes transmission at MET 35:10:50:17. Seven seconds later Icarus 1 abruptly ceases all communication, ending the mission after 35 days 10 hours 50 minutes and 24 seconds.

• Project Icarus Incomplete
  
• 160 science collected
  

Moho has frequent transfer windows so another attempt can be made. The most promising window would allow the mission to be launched while Martian 1 is on its return trip and finish before Tyson reaches Dres. With my new delta-v figures it will require a heavier rocket - possibly it might need the same Heavy J-Series Booster that the 2 manned missions where launched on.

[Commander Zoom]

May I suggest that you reconsider the use of an ion engine for the upper stage? I had plenty of time (from entering Moho's SOI) to brake into a capture orbit, and thanks to the solar input that you observed, I could (and did) run the drive flat out the entire time. For a light probe engine with a total delta-V in the multiple thousands, it's really hard to beat ion. (Just don't plan on finishing your burns in single-digit minutes.)

[Andersenman]

Scott Manley recently had similar troubles encountering Moho,

. Possibly, some inspiration can be had from that.

Then again, I'm curious to see Dave's own, original progress. I really like this series.

[Dave Kerbin]

For a non-ion powered version I had two thoughts. The first is taking the T200 tank (90L) and instead of the tiny T100 tanks (45L) held on sideways with docking ports, go full side with radial decouplers. From my calculations there would be 6 attachment points. 4 of those attachment points would recieve T400 tanks (180L), these tanks would be the main fuel through the injection and transit and hopefully provide a little fuel during capture. The last pair of attachment points would recieve smaller T200 tanks but these would be equipped with an engine each, tripling my thrust (well not really, I'd be carrying 3.5 tons compared to the original 1.5, but the intent is to keep thrust up).

That design would weigh about 13 and a quarter tons, meaning it would need to be launched on the bigger Heavy J Series Launcher. Since the launcher supports 15 tons I might consider the inclusion of a Science Jr to be used in Moho orbit.

Some other thoughts I had where that the landers delta-v could be extended by a useful margin if I ditched the batteries (a battery would be carried on the transit stage instead, once we got close to Moho we would rely on high solar input for transmission) or cutting down the solar panels, and I have thought about a crumple zone or other landing protection (the M-Beam 650 I-Beam and Octagonal Strut could be useful).

However I'm not really keen on using the heavy launcher for a little probe, it seems wasteful. An additional piece of data I have from the filed Moho capture is how much time I would have - my capture window was almost exactly 60 minutes before I would slip back out of Moho's SOI. Now if I'm burning that window will extend slowly but I don't want to rely on that, but it still gives me a somewhat solid number that I can crunch against ion engines. Since the light booster was designed for 3.4 tons and the probe weighs 0.58t I would have a 2.82t budget for an advanced ion system. Since an ion system was already designed for the probe once I'm wondering if that design could be adapted quickly, keeping the cost of the 2nd Icarus probe down by reusing as much of the original design work as possible (if the same lander design is used I might be able to assemble most of it from the spares and testing articles used in the originals construction).

[Dave Kerbin]

Martian 1

As I've mentioned before Project Martian and Project Tyson are designed to share most of their mission hardware. They use nearly identical landers, sharing the same J-Series return capsule and differing only in the radial landing pods. Martian 1 uses the primary A-type pods (atmospheric) which include 3 parachutes, 2 barometers and a thermometer. Tyson uses the V-type pods (vacuum) which omit the parachutes and atmospheric instruments to reduce weight. I will likely try and reuse the J-Series lander design in phase 4 to keep design and testing costs down.

The landers, plus required transit hardware, are built to a 15 ton design parameter which allows them to be launched on identical J-Series Heavy Boosters, a man rated (it has a higher delta-v safety margin) launcher built around the same 3 way symmetry as the landers. The 3 way symmetry allows the launcher to do something interesting, it is designed to seat the payload in a lower then normal position. Instead of being seated at the very top of the launcher the payload is mounted farther down. This way the usually tall interplanetary craft (with their transit stages attached) aren't wobbly towers with a center of mass far above the booster rocket. Instead they are mounted half way down, keeping their center of mass inside the body of the launcher. This greatly simplifies the attachment of payload tie downs which can be hooked up to the body of the return capsule itself for best stability. In the second picture you can see how radial portions of the transit stage (the part with a goo container on top) easily mesh inbetween the large booster rockets.

The first manned mission to another planet will be piloted by Rayfrod, my first kerbonaut. Take off is safe and stable, the 4 skipper engines providing plenty of power and a large reaction wheel keeping things steady.

A number of my recent launcher designs have begun to take advantage of an understanding of how much delta-v is required to clear most of the atmosphere and reach the 'coasting' position before the AP burn. This design is no different and the 3 heavy boosters run out of fuel at a good time to start coasting. The center engine is left running for a split second longer to ensure the boosters safely clear the spacecraft before likely converging. They are assisted by a tiny Sepratron I each, as parts secured with struts don't get the same acceleration when decoupled.

At this point the stubby little orbital booster is used at the AP to establish an initial 80km orbit and then refine it by fine tuning the AP, PE and inclination. The reaction wheel helps it turn the payload (that's what the big batteries are for, so the reaction wheel doesn't overload the payload's power systems). I'm quite happy with how the aesthetics of the orbital booster came out, being suitably fat for a booster but not being too large in comparison to the payload. Part of that was because it has a full tank of fuel when it burns at the AP while in the past the last booster stage is usually carrying a lot of empty tank space by then.

In proper orbit solar panels are extended and Rayfrod settles in to await his injection burn in a few hours. The ideal burn would be at MET 4 hours and 2 minutes, but since the orbital position is more important then the time he actually has a single orbit window, approximately 3:45 to 4:19, where he will make his burn to Duna.

The fuel remaining in the booster is burned to provide 40 m/s for free, then it is released and the transit stage burns another 1016 m/s. Martian one then coasts out into deep space. On day 28 of the mission a tiny correction burn is made to line up the Duna encounter. The full details of that burn where in an earlier post, describing the usage of the RCS thrusters to provide the small adjustments needed to slip around Duna's moon Ike which was blocking our path.

The stop over in deep space also allowed some science to be done. A full range of readings where taken, including material experiments using one of the 3 pods mounted radially on the transit stage.

Arriving at Duna the PE for aerobraking is calculated to be 11715m. With some knowledge gained from orbital adjustments Rayfrod used a newer technique, which was to reduce the engine output to the smallest amount, 5.5% (right click on engine and use the new tweakable) and then burn at minimal throttle to produce even less thrust then RCS (which was still used to help find the right angle to fire at). This let me adjust the PE to 11719m while still over 45,000km from Duna. More science is done in high Duna orbit, using up the 2nd of the 3 science modules attached to the side of the transit stage.

Aerobraking doesn't go exactly as planned. I'm not sure if the calculations where wrong or the input figures where but following the mission plan I retracted the solar panels and then monitored AP during braking. When it reached 60km and kept falling I turned and engaged the transit engine to counter the drag, halting the AP at 45km and bringing it back up. Oddly enough there where not any obvious reentry flames while braking over the night side of Duna.

Once in orbit science is recorded on the 3rd transit stage science module. An EVA is then performed to collect from all 3 modules and return the data to the command pod. Now preparations for landing start with a maneuver mapped out to land on the day side in what looks like a valley and the science pods are ejected to cut down on mass and save fuel during the deorbit burn.

The transit engine is used for the deorbit burn, saving fuel in the lander. However at this point things go a little differently. Normally the transit stage would just be ejected and discarded. But this transit stage is a little special. The ship is turned around so that transit stage is pointed away from the direct the ship is traveling and released. A tiny push is given to send it away. Then control switches over to the transit stage which has it's own probe core. Next a preplanned orientation change is made so that the transit engine is pointing in the correct direction for a vertical burn. This is only a small burn intended to push it out of the descent path of the lander so that it doesn't 'catch up' with it when the lander deploys parachutes. Next a docking port that previously allowed fuel flow is decoupled, releasing the transit engine and an empty fuel tank. A tiny engine matching the landers (and which was previously being stored directly opposite to it in the separator) lights up and begins pushing prograde to restore orbit to a small probe that a part of the transit section.

This is the Ike lander, it was designed to take advantage of a 1.5 ton weight difference between the completed Martian and Tyson transit systems. At one point I examined making the probe reuse the transit engine, so that the transit stage would simply shed any radial loads and become a probe, but the extra weight was too much. Instead the Ike probe has it's own engine and needs about 30L of fuel left over from the transit stage to fly. With the probe back in Duna orbit the lander continues to descend. Two of the pods are used to collect science in the upper and lower atmosphere. The checklist includes highlighted diagrams to help quickly locate the correct instruments to operate at each stage, since the pilot won't have a lot of time to mess around.

The ground is approaching quickly with two piece of debris (the seperator and the transit engine) passing ahead of the lander as thrust is applied to help slow me down. The thin atmosphere wasn't providing enough drag to get the velocity down to a safe point for parachute deployment (I'd prefer to be under 150 m/s, 300 m/s as a last resort). At a little under 500m from the surface Rayfrod has hit 190m/s and the chutes are opened. There is a sharp tug and the G meter hits 15. Everyone seems to still be in one piece though. Descent speed is now 7.3 m/s. Touch down is with some thrust to lower speed to 5 m/s, nd there is a slight incline causing the lander to bounce and slides a little with the RCS kicking in to compensate. That bounce isn't entirely good, without RCS the lander might have tipped. A more rigid construction might have been better.

Rayfrod Kerman is on Duna.

Rayfrod performs some important experiments on the surface and collects data from all the science gear on the radial pods.

At this point Rayfrod can enjoy a day on Duna, though we won't be watching him. Instead we're going to leave him on the surface and go back up to orbit where the probe is. The probe is sent on a course to Ike where it will begin collecting data. The probe carries 3 GRAVMAX instruments, a pair of 2HOTs and a single DOUBLE-C and Science Jr for the surface. It also carries an antenna though that is only as a backup.

After capturing and taking low orbit readings the probe lands on Ike's surface at a gentle 2 m/s. It still tips a little though the reaction wheel stabilizes it just before the solar panel touches the ground (at the slow speed it was tipping the solar array wouldn't have been damaged, it would have actually acted as a convenient support). This is where the single Science Jr is used to collect data closest to Ike. No Goo containers where brought along because of their very poor science:mass ratio. After surface operations are complete the probe takes off and plots a return to Duna orbit - so far this is very similiar to a typical Mun mission.

Back at Duna I tried to aerobrake the probe to save on fuel. Since I had no plans for calculating it I was just doing it freehand, which required 3 conservative passes and finally a burn to circulate. Fuel seemed to be low by this point.

On the surface Rayfrod monitors the probe and when it is passing overhead he launches into what will hopefully be a renderous orbit. The plan is for the probe to do the work. Rayfrod lifts off and immediately ejects the science portion of his landing pods. As he ascends the fuel portion of the pods, which was also used for braking during descent, reaches zero and is released too.

With the return ship now in Duna orbit the probe makes an inclination burn to match, but this depletes all of its remaining fuel. A review of the fuel situation on the capsule indicates that there is a good safety margin - there is 135L of liquid fuel and plotting a course home shows it would take 1979 m/s. With a safety margin that would be 100 L of fuel leaving a good amount extra that can be used to get to the probe, instead of the probe coming to the ship. If it had been impossible to reach the probe (or if the probe had failed somewhere) it was equipped with an antenna for transmission and a priority list for which data to send first in the event that the solar panels where damaged. Since they had originally been on a close intercept but the fuel situation delayed it both ships have to spend hours orbiting Duna at 10x warp until a cheap intercept was possible. Once the craft where close together the relative speed was zeroed out and Rayford went on an EVA to collect the data from all the probes instruments. At this stage it didn't matter if he accidently broke any of the solar panels so no effort was made to retract them. He also repacked the ships parachute before returning to the cabin.

The burn to return to Kerbin is uneventful, the capsule accelerates away from the probe and toward home.

Arriving in Kerbin's space there was something familiar, another moon getting in my way. I couldn't avoid this one completely so I passed over it, my speed preventing my path from being adjusted too much.

Landing was polar but straight forward. Rayfrod is now also the first kerbal to visit Kerbin's tundra. Total science for the mission was 4495.

• Project Martian completed
  
• Manned landing on Duna, unmanned but returned probe on Ike
  
• 4495 science gained
  

[team.leit]

Well done, I wish I could be as good as you are at planning missions, to bad I am not I wish you the best of luck

[Dave Kerbin]

Tyson 1

The last and longest mission in phase 3 (not counting Icarus 2) is the manned mission to Dres. It uses the same launcher as Martian and a V-Type variant of the same lander (same capsule design, radial pods leave out the atmospheric components). The only piece of new hardware is the nuclear powered transit stage. This transit stage is heavier then the one carried by Martian and has no science components or probe, it needs all the mass to provide the delta-v to reach Dres safely.

This mission will be a major test of the feasibility of a mid course correction (Icarus didn't go so great but I think there where several factors there). When we burn from Kerbin we will not have an encounter shown for Dres, only an orbit around the Sun. At the correction point we will either need to make Dres with our budgeted delta-v or abort and come home. The launch is at night though this isn't really an issue, the heavy booster proved itself simple and easy to use. Orbit is the same 80km orbit used for Martian and a similar launch window is setup.

The first burn for is for 1461 m/s and again we borrow some delta-v (72 m/s) from the orbital booster. This puts us on a path to Dres, but without the inclination. Without a correct in 51 days we'll pass far below Dres's orbit (brown). Because Tyson 1 was launched back on day 98 you can see Martian 1 on it's way to Duna (it will arrive and land in 28 days and then return safely before Tyson 1 even reaches Dres) and the doomed Icarus 1 probe a week prior to impact

The correction burn works out nicely, 497 m/s to adjust inclination for an encounter with Dres 100 days later. Martian 1 is also on it's way home with the Ike probe abandoned in orbit. The radial transit tanks are finally depleted here a little more then half way through the burn. Dropping them shaves 0.45t from our mass and improves efficiency - despite carrying less fuel the lighter center stage provides more delta-v. The final PE is lowered to 174km but the alignment makes lower it below that difficult (with the nuclear engine still attached it's tedious to change angles and find the correct orientation).

Arrival at Dres also goes as planned. A few hours before entering the SOI the expected PE is lowered to 30km. The capture burn is simple, costing a little under 1500 m/s. Science is done at high orbit and low orbit, using 2 of the landing pods. Originally the low orbit science was planned to occur during landing, but the capture came in a little lower then expected, putting the ship low enough for data to be gathered. The plan is the same as Duna, the ship will land and perform more science and then all the science can be collected from the pods before discarding them on take off. Descent begins with the last of transit stages fuel before being handed over to the main engine, feeding off the fuel in the landing pods.

I think this was the point where I was sure things where not going right. I still have 70 m/s to cancel out.

Speed is 22.2 m/s, rapidly changing to -5.1 m/s.

The second landing is more gentle at 3.2 m/s.

Torque from the pod combined with a little seesaw motion is enough to tilt the ship upward and unblock the hatch. One Science Jr. is responding and it's the unused one so it might as well be exposed on Dres surface.

Gregfield carefully gets out to take a better look at the damage.

Ok, so things aren't good. All 3 Goo containers where smashed. They have an impact rating of 12 m/s (same as landing legs) and absorbed most of the impact. Two of my Science Jr. units also acted as a crumple zone. The bad side is that the two destroyed containers where the 2 I used in orbit and hadn't recovered data from. The good side is that they protected the engine which is intact and functioning. I also still have all my fuel and the fuel lines are intact - if I had lost a fuel line it would mean imbalanced fuel flow. I'd have to lock off the good tanks, burn from the center tank and then manually transfer fuel once there was enough space to put it in the one good center tank.

The situation seems salvageable. If Gregfield uses torque to lean while carefully throttling up he should be able to get off the ground without scraping the fuel tanks along the surface. Before then he might as well do some science and collect from the Jr. after all the trouble he's been through to get it.

Take off is successful. Things should be coming back on track. 15% of the pod fuel is left after reaching orbit.

The trip home is estimated at 2971 m/s which the capsule alone could make with some margin to spare. After the injection burn and exit from Dres's SOI the Kerbin PE is lowered to 20 km. It takes another 100 days to reach home and land off the coast. Video seems to be busted for the final splash down and recovery which makes it more work to account for science. Working backwards the numbers suggest the mission added 1446 science, which sounds about right as I remember it being 1400ish it also matches the expected yield when the destroyed equipment is subtracted. This is enough to complete the rest of the tech tree.

• Project Tyson completed
  
• Reached the surface of Dres and returned home
  
• Collected 1446* science
  

[Dave Kerbin]

Sidenote: Ion powered Icarus probe

Icarus 2 is not going to be ion powered, I've done ion probes outside of career and I just don't have the free time to do the burns. But that doesn't mean I can't design an ion powered version.

The first step is the weight, it must still fit on the same launcher so the final mass needs to be 3.4t or less. A second requirement is delta-v, which I have some revised estimates for. And last it will need to be able to execute a capture burn in a reasonable amount of time. To solve all those requirements it's off to Excel.

Based on data from Icarus 1 I have a realistic value for how much average power I will get from my 1x6 solar panels. Along with other data I created this little scratch sheet. At the top left are my delta-v estimates while to the right are my findings with various engine/fuel combinations. Below that is the craft stages in the manner that I normally do them in Excel. The blue highlighted numbers are the ones I was experimenting with, I wanted to see how many ion engines would be optimal. The lower right has some calculations used to determine the current builds burn time for the Moho capture.

As you can see it seems like 6 ion engines meet the estimated burn time requirements. I can actually add a little fuel (2 more stack tanks) to the transit stage to bring it closer to 3.4t, providing me with some extra delta-v. The placement issues I've got marked are for configurations where I can't see the placement of engines and solar panels in my head - it would seem like the panels would overlap or clip each other though there might be a creative configuration.

The probe itself is reconfigured. The mass is lowered by using static panels, we won't need to worry about alignment during transfer since the transfer/capture stage will be full of solar arrays. The batteries are also redesigned, same mass and capacity but using a stack battery to replace 2 radial batteries. This makes the weight balance a bit harder (I need the antenna and radial battery properly aligned to balance each other) but lets me introduce a big crumple zone. The highest impact rating part on the probe is the probe core so it is on the bottom of the survival section (the legs, engine and fuel tanks can all serve as crumple zones). Above that is everything needed to collect and transit data.

The ion drive section is in two parts. The capture stage comprises most of the system while the transit stage is just some extra fuel linked to the center which can be released to reduce weight before the capture burn.

The solar panels are setup in pairs - you'll notice the each panel has a neighbor on the next ion engine that runs parallel to it. When fanned out it should make 6 pairs of parallel solar panels.

[team.leit]

Wow, how many mission was it in total? I don't think I have the patience to do something like that

[Dave Kerbin]

Tyson was the longest mission in days for my kerbal pilot Gregfield (251 days), but most of that was spent drifting in deep space. Looking at the video files it looks like about 2 hours of actual play time split over several sessions and between other missions.

In terms of real time Martian 1 was more, it had a lot of complexity to it (so many readings to take at the right time and be collected in the right place) and I had to wait at 10x warp for my docking to line up. Video suggests a little more then 3 and half hours.

The earlier missions didn't overlap with each other in game time so they where all played in one sitting. Approximate times are (because I'm not going to figure out how much of the video is a VAB tour and science review)

Reach: 12 minutes

Leap: 33 minutes

Gander: 55 minutes (a lot of checking out the window to see if I was over a new biome)

Cheese: 53 minutes

Mint: 1 hour 25 minutes

Berry: About 1 hour 10 minutes, missing the end of the video due to a crash

I've flown Icarus 2, I just need to get it ready for reporting. The first video file (launch and injection burn) seems to be corrupt and I want to try to recover it.

With Icarus 2 flown, it completed shortly after Tyson 1 landed, I am now at year 1 day 359 (or roughly Christmas) and have completed Phase 3. I already have some ideas for Phase 4 (the Icarus 2 systems factored into that). It won't have the same kind of mission planning and fancy science as the previous phases and in fact I'll be trying to run on existing hardware to safe money for the last phase, 'Future'. Phase 4 will mostly be about interplanetary travel and learning about getting around Jool's moons. I'll be landing on 3 moons (never been to Pol or Bop at all, only done unmanned mission to Gilly in my .22 career). I'm not sure how exciting it will be for observers as it will mostly involve me learning to estimate more complicated delta-v requirements.

The last phase, 'Future', won't be big on science either (I have finished the tech tree after all) but it will be my biggest engineering challenge ever. I'll be doing a 2 kerbal mission to Moho (that should require some huge delta-v), a mission to Laythe (never been there) that will require me to bring and setup a base of some sort, and finally a manned return mission from Eve which is probably the biggest challenge in stock KSP. And it will all need to be done with 5 nuclear engines (2 are being saved from phase 4)

[Dave Kerbin]

Icarus 2

This is a follow up to the failed Icarus 1 mission which crashed into Moho when it arrived without enough fuel for the massive capture burn.

I didn't want to use ion power for time reasons (see ion design above) and that meant a chemical rocket which would immediately go over the 3.4t allowance of the light booster. I didn't really want to invest in the R&D for a new booster for this mission, over budget as it already was, so the heavy booster was picked. While that left plenty of weight allowance to design something new I decided to first see if there where any existing designs that could be pulled up and used to avoid new design and testing costs.

One design that fit the bill without any modification was the center booster/transit design used in Cheese and Mint.

Seen here without the small tail portion used during lift off.

There where several downsides to this design however. First was that it was very tall which could pose stability issues during launch. It was also based on older technology (before fuel lines) and so it had to carry multiple LV909 engines used in succession, driving up the material cost. Finally there where a few unknowns with stripping off all the radial hardware.

So instead of going back I went forward a little. For phase 4 I am going to be heavily reusing hardware from the current phase 3. One piece of reuse that was already in the final stages of consideration was a cheaper conventional engine variant of the nuclear transit stage used in Tyson 1. The fuel duct system, radial clamps, payload tie-downs and top mounting point optimized for a Rockomax 48-7S would all be the same, but a much cheaper LV-909 would be substituted for the expensive LV-N atomic engine.

Since the probe used the 48-7S it would fit perfectly, the optimized mounting point even allowing the probe to launch facing right side up. The unmanned Icarus 2 will be used as a practical test of the modified transit stage, allowing some of the ground tests to be cut.

With the heavy booster the launch is the same as Martian and Tyson. The same injection burn is done as Icarus 1 only using the bigger transit stage. I'm lacking illustrations but I changed my mid course correction which is done once I reach the 9 o'clock position around the sun (assuming our orbit is anti-clockwise when viewed from above and Kerbin is at 12 o'clock). I found I had two similar corrections I could make that both cost about 1800-1900 m/s. The first was very close to the correction I made with Icarus 1 which would take me to a direct meeting with Moho at 3 o'clock after I swung around the sun. This encounter would be on the original ~40 day schedule and let me land on Moho while Tyson 1 was still in transit back to Kerbin.

However I found a second possible correction that seemed better. This correction would better align my inclination with Moho, so that when I eventually encountered it I wouldn't be coming in at a sharp angle that needed thrust to cancel out. This alternative course would require me to take an extra trip around the sun - I would continue without correction toward the 3 o'clock position where I would miss Moho but instead make a course change that lined up their planes and setup an encounter at the same time on the next orbit. The longer trip would take 86 days in total but bring my capture delta-v down to 3600 m/s, though I needed to spend more delta-v then I should have to lower my PE first. If I had set that up earlier I would have saved some fuel.

The Kerbin to Moho injection burn, the orbital correction burn and the PE correction burn are all done using the new transit stage (I need to give it a proper name if I reuse it). The final capture is done using the original probe's tank.

Sadly the only shot I have of the correction, taken shortly after it was made.

We burn into a nice orbit around Moho.

I want somewhere flat to land. Orbital observations don't really help me figure out what areas are bad to land in. If Moho is bumpy then the elevations must be much lower then planets like Duna - many small hills instead of large mountains. Without any other informed options I opt to land in the middle of a crater where most inclined should have been wiped out. There is one good candidate on my light side orbital phase.

With some fuel left in the big tank I should be ok for landing. After using it up to help kill my orbital velocity I release it and then thrust up a bit so that the tank starts moving ahead of the probe. I am trying to use it to judge the altitude of the landing zone - when I see it crash I'll know how far away the ground is. When it is showing 5.2km and my own altitude is 6.1km it disappears (it is too far away now to see an explosion).

Knowing that the ground below is at roughly 900m altitude I burn accordingly. I also look for my shadow but I can't see it until I'm about 300m from the ground. However I have plenty of fuel so the last 250m or so are made at under 15 m/s. The landing is a bit off balance, leading to another instance of reaction wheel torque holding the ship on 2 legs with a solar panel dangling inches from the surface. Altitude at landing zone is 922m.

Once on the ground the Icarus mission can finally be completed by taking readings. The huge amount of solar energy being recieved means the batteries are not required, I can transmit continuously with solar power. Even with a small set of instruments we can learn a lot about Moho.

• Project Icarus Completed
  
• Landed on Moho and took temperature and other readings
  
• Learned more about Moho transfers
  
• 210 science gained
  

[Trann]

Only found this thread recently; took me a few days to catch up.

Great story telling. Great project management.

Thanks for another way to play KSP.

[Dave Kerbin]

Ice 1

As I've mentioned phase 4 "Inquisitive Minds" is probably not going to be as interesting (particularly the unmanned missions). I'm going to be mostly reusing hardware from phase 3 so research and design costs will be close to zero. There will also be no nuclear engines, I really need to save up the LV-Ns I have budgeted for the big engineering challenges in phase 5. With the tech tree already unlocked the knowledge I'll be gaining will mostly be in advanced transfers as I tackled a return mission from Gilly, a trip out to Eeloo and a lot of movement around Jool's moons. Just as Icarus provided some key data that will be used in one phase 5 project these missions should help in planning the requirements of the other 2.

The first mission in this phase, Project Ice, really brings home a comment I made a while back when I said that I seriously under estimated the speed with which I would complete the tech tree. When I was coming up with these missions I was trying to put this one far enough along that I would have the Double-C Seismic Accelerometer unlocked (I was even concerned that Project Berry would come too early for me to have a probe core). Instead I'm starting this phase with the entire tech tree unlocked and I've had the accelerometer since I went to the Mun.

The R&D reuse on this mission is hopefully going to restore some honor to the Icarus 1 design. It was light and low cost to build and packed 5000 m/s for transit with a 1600 m/s versatile lander. The fact that I encountered serious delta-v inflation on the way to Moho (in particular I learned that there is a huge delta-v difference in the exact approach used to its orbit) shouldn't be allowed to tarnish the design. So to get to Eeloo I will be reusing the design. Not a new ship based on the old design, but an exact copy. Only the name plate and the data programmed into the probe core differentiate Ice 1 from Icarus 1. I should have a large safety margin for transit (1000 m/s) and the radial tanks should take me through the injection burn almost exactly.

One month into the new year Ice 1 is launched into the night. I'm going to see what results I get performing the injection on the same day instead of calculating the exact orbit for the optimal burn. The mission plan itself is just a sheet with my injection angle (I think I described before how I just hold that up to the screen to get a rough idea of where the node should go) and the dates and delta-v for injection and correction.

The injection burn will be for 1980 m/s at about 112 degrees from prograde which puts it squarely on the dark side of Kerbin for the entire burn.

I'm only running one mission at a time so as soon as the ship is out of Kerbin's SOI I plan the course correction 100 days in the future. It turns out to be very easy to get an encounter for just 164 m/s.

After 330 days the probe finally arrives at Eeloo, the farthest planet (combined with the fast forward 1 month to reach the launch window that puts the mission at almost 1 year in total).

The capture burn is another 1623 m/s plus a bit more to smooth out a 40km x 50km orbit. Orbital science is done using battery power, out here the solar panels are operating at 50% so the probe needs to take breaks to recharge. For landing the surface seems pretty uniform. The brown lines might be interesting though I'm not sure if a precise landing could be made on them or if it would be safe - those could be jagged rocks for all we know. The only other notable feature is a crater so a landing is planned to set down there.

Fuel for the deorbit and orientation to the landing zone are done using fuel from the big tank. After that the tank is carefully released and the probe thrusts away. With the altitude at 21km I don't want to be dealing with releasing the top mounted fuel tank closer to the ground where it could pose a danger of colliding with a fragile part of the probe. There is plenty of fuel left for landing, though the ground is very uniform which makes it hard to judge how close the probe is to the surface. No scatter (rocks) can be seen.

I take the landing safe and slow. 1.2 km from the surface I can finally see the probe's shadow, visible as a small dot above and a little to the right of the landers top leg.

After a gentle landing the probe can complete it's mission. Some interesting activity is detected which could provide the answers kerbal scientists where looking for.

This was a pretty straight forward mission to fly (49 minutes real time, a lot spent at maximum time warp), the mid course correction being the critical section for success.

• Project Ice completed
  
• Landed on Eeloo and transmitted seismic readings
  
• 555 science gained
  

[Dave Kerbin]

Sidenote: Block II J-Capsule

The manned missions for phase 4 are reusing the well tested and well liked capsule design from Martian 1 and Tyson 1. The landing gear design is getting a bit of work though, with a block II design that includes 6 landing legs. I usually use 4 legs but here that would interfere with the pod mounts unless the legs where moved closer together (which would make it easy for the pod to tip left or right). 3 legs would require 1 leg under the ladder which would require ladder modifications. 6 legs works out fine and also makes up for the fact that I'm using the probe lander legs instead of the normal sized ones. Impact on delta-v in minimal, the pod still gets over 3000 m/s.

The second addition is the easy development of new R-Type pods (range) which simply drop the complex science components of the previous pods and extend the fuel tank. These tanks add another 2000 m/s and with the new landing leg design the capsule can land with or without the pods.

Block I capsule used in Martian 1 and Tyson 1

Block II capsule

Block II capsule with R-Type pods

The whole thing goes on the modified transit stage that was tested with Icarus 2. This transit stage adds another 1900 m/s (900 m/s from the radial tanks, 1000 m/s from the core). Finally it is mounted on the original heavy booster. With the shortened transit stage the ship sits even lower, putting the upper stage just above the booster. All the original clamps on the transit stage and booster can still be used to tie down the capsule.

[Dave Kerbin]

Dust 1

Project Dust is concerned with collecting soil samples from potential captured asteroids: Gilly, Bop and Pol. Because they are in opposite directions transfer wise I decided to split it into two missions, one inward to Eve and one outward to Jool. Both will use the block II capsule design though I had some other thoughts along the way.

For Jool I considered the idea of carrying project stone along with me. I even worked out a weight balance system so that under one pod it would carry an ion powered probe to explore Jool's moons and the other 2 pods would have equally weighted science and fuel packages. I also considered a similar layout for Eve - Ice 1 would have been carried to orbit on the same heavy booster, saving a launch. However in both cases I ran into concerns over how much fuel would actually be needed. I had some estimates but they contained a lot of guesses for how much transfers would cost. When I examined worse case scenarios the manned missions always came up short on fuel which was unacceptable.

I also had a seperate idea for cheaply reaching Gilly. The core block II capsule (without the fuel pods or transfer stage) weighs just under 3.4 tons meaning it could be launched on the same light booster as the Icarus 1/Ice 1 probes. That capsule has a lot of delta-v, just over 3000 m/s, and I was trying to figure out on paper if gravity assists and aerobraking could make it possible to reach Gilly and come back. However I found a lot of unknowns regarding a cheap transfer to Gilly and coming back from Eve was a real problem - at the optimum time it is relatively cheap but it gets very expensive very quickly. I had no intention of orbiting Eve for half a year waiting for a transfer window.

Ultimately I came down to using the full transit stage even though it might be overkill for the Gilly mission. I was using the heavy booster anyway so I might as well not waste the weight allowance. I also decided on going to Gilly first, mainly because the transfer window was first. The launch on the heavy booster should be pretty straight forward by now. Rayford will be piloting this mission and I'll turn things over to Gregfield for Jool.

I'm going with a mid course correction transfer again. 1050 m/s out and with the small inclination difference between Kerbin and Eve that puts it close to an encounter already. Like always the orbital booster is fired to give just over 100 m/s of free delta-v. That works out nicely when combined with the delta-v in the transfer stages radial pods, since they will run out just as the injection burn is completed. The seperation of the booster happens in an interesting place - below from left to right you can see markers at the Reach 1 landing site, the VAB and finally the launchpad.

In deep space a 28 m/s correction is done and we arrive at Eve 36 days after leaving Kerbin. Aerobraking is high at 68 km - I just want to slow down enough to enter orbit around Eve with the AP out near Gilly. This is the first time I've ever brought a Kerbal eyeball out to Eve so I make sure to take a good look.

Now there where some issues with my orbit. I was not aligned with Gilly and was actually well inside Gilly's orbit. But I also also orbiting Eve retrograde which was an issue.

After 4 orbits I have an opportunity to setup an intercept. But remember that Gilly is coming counter-clockwise and I will be meeting it coming clockwise, essentially a head on collision.

Entering Gilly's SOI I throw away the transit stage even though it has 45L of fuel left, I want as much TWR as possible as I start trying to slow down. I eventually pass by Gilly but my exit point keeps getting extended the slower I go. Eventually I finally fall out the other end of Gilly's SOI but by that time I'm doing something interesting with my orbit. After burning most of the 3 fuel pods I am now reversing course and catching up with Gilly.

Now I'm burning to catch up with Gilly from behind.

I am now in orbit around Gilly moving at city speed.

It really doesn't take much to deorbit (15 m/s) so I still have fuel left in the pods. It takes ages to descend so I actually use RCS to push myself down a little fast. Landing is simply the opposite, I use RCS to slow down.

It actually takes a while to get the ship nice and settled. Eventually SAS has to be turned off because even that torque is causing the ship to shake, there just isn't enough gravity induced friction to cancel it out. The ladder is not needed, movement on Gilly is by jetpack only.

Rayford gets the samples we where after but not without a bit of difficulty in the low gravity.

Escaping Gilly and getting back into Eve's orbit isn't really an issue. Pointing the rocket up and thrusting for a few seconds is all that is needed. I got a nice surprise with the return cost, only 1169 m/s. Until now I had been holding a lot of fuel to guard against the possiblity of a 2400+ m/s cost if Eve had traveled enough out of alignment with Kerbin.

There is one final experiment I wanted to conduct though it is not entirely risk free. I want to check if the fuel pods can be ejected with the landing gear extended and the engine on. I have some small concerns that the fuel pods could catch against the end of the landing legs. At this stage I don't need the legs anymore so the risk would be a chain reaction that resulted in the fuel pods coming back at the capsule itself. The experiment goes fine, the decouplers produce enough force to easily clear the fuel pods away from the ship.

The rest of the return is uneventful as I have a literal ton of fuel (1.05t) left for any course changes. I pass by Gilly (but outside it's small SOI) again on my burn away from Eve. Rayford takes one last look.

A total of 1889 science appears to have been collected.

• First of three objectives for project Dust completed
  
• Landed a kerbal on Gilly and returned with samples
  
• Performed a backwards capture
  
• 1889 science gained
  

[Dave Kerbin]

Dust 2

Using the same mission hardware as Dust 1 my other kerbonaut Gregfield is going to make my first career manned visit to the Jool system (I went there once trying to land on Tylo with 10 parts, was 500dV short) and my first visit to Bop and Pol. You've already seen how this gets into space, setting up a transfer to Jool isn't very hard because Jool has a massive sphere of influence that ensures anything that comes remotely close will have an encounter. Cost of the injection is 1966 m/s which uses up the transit stage, the rest of the mission will be flown with the capsule and fuel pods. After leaving Kerbin's an adjustment is combined with a plane change to help bring me closer to Jool's inclination and lower my PE. Unfortunately I did make a mistake I've made in the past and instead of setting my PE to 134 km I set it to 134,000 km. So once I enter Jool's SOI it cost another 192 m/s to lower my PE to 120.6 km for aerobraking. With Jool's massive SOI it will still take 9 days to reach the atmosphere.

My aerobraking goal was to get my AP around Pol's orbit. I'm not too far off at 195,000 km but I do have a big inclination.

With my inclination close to Bop's I decided to go there first. It helped that I could reach it with a course change right after aerobraking allowing me to get there in less then 1 orbit of Jool. Another 487 m/s for that burn.

There's something off about Bop.

1000 m/s to brake. Like Gilly Bop is not a normal round moon but an irregular captured asteroid. The braking used up the fuel pods so they are released. I land on a tall mountain with an incline, but the low gravity combined with 6 landing legs seems to let me land without tipping over.

After setting the flag I go on a little trip to investigate a piece of debris I can see about 1 km away. Unfortunately it turns out the reason it survived impact was because it actually glitched into the terrain.

I launch back into space and leave Bop's orbit. From there reaching Pol takes some more time. I correct my inclination but I'm now in a slightly lower orbit then Pol and not quite in the right place to reach it. It takes almost 2 orbits around Jool (at 6 days each) to get into position. Fortunately it only costs 118 m/s which is good because I'm starting to run low on fuel. I've figured I need to budget at least 1500 m/s to get home and I've used the ship's blueprints to scribble down how much dV I'll have for each 10L increment of remaining fuel. I draw a big line between 80L and 70L marking the cut off after which I may not have enough to get home.

With my orbit now much closer in shape to Pol's it only costs 128 m/s to capture. Deorbit is another 100 m/s followed by more fuel to land leaving me with 94.5L. The surface of Pol looks dangerous, I have to be careful where I land.

Gregfield makes some important observations for the scientists back on Kerbin.

After taking and leaving Pol for Jool's orbit again l have 82L of fuel. The trip home will take 1427 m/s leaving me with a small margin left. I reach Kerbin with just over 8L in the tank and half a supply of monopropellent left (or about 300 m/s in total after spending about 6700 m/s on the mission)

• Project Dust completed
  
• Collected samples from Bop and Pol
  
• 4154 science gained
  

[Dave Kerbin]

Stone 1

This will be the last mission in this career for a while. I intend to come back at some point and tackle the engineering challenges of phase 5, but at the moment those (part from the LV-N restriction) are basically show off sandbox missions and that's not really what I've grown to like. Until .24 comes out (at which time I will do a new career with money and reputation, in hardmode if it has been added to the game) I do have other ideas for careers, like playing with mods for the first time (kerbal life support, FAR, deadly reentry and possibly any others that are focused on adding challenges) but I also have an idea for a shorter career I will probably pickup and play.

The goal of Project Stone is an orbital gravity survey of all of Jool's moons. Stone 1 uses the same hardware as Icarus 1 and Ice 1. While it may have failed to reach Moho it is a capable spacecraft. I did make some minor modifications, I removed the landing legs (at the end of the mission I regretted that) and replaced the double-c with a barometer. Since Bop and Pol have both been explored and had their gravity mapped out this will just be a mission to Vall, Laythe and Tylo. Since the second mission of phase 5 involves a major operation on Laythe I intend to visit there last, I can easily aerobrake into a stable orbit so that the probe can later be used destructively to gain some data on how Laythe's atmosphere affects descent profiles (an accurate landing will be required for the manned mission given the small amount of land).

Reaching Jool on the 3 radial tanks the aerocapture is set to try and put us close to Tylo's orbit.

Adjustments start 17 hours after capture as we pass our orbits AP and make a plane change for 194 m/s to match Tylo. This plane change also lifts our PE out of Jool's atmosphere for free.

As we swing around for a second orbit we setup another burn, this time for 158 m/s to bring us on a low pass by Tylo to collect data. Since we also want to visit Vall we might as well do a gravity assist that sends us there for free. I don't even know why I need the big fuel tank.

We pass by Tylo and take some readings.

Then we just coast over to Vall.

15 minutes outside of Vall's SOI we need to make another burn for 537 m/s to setup an encounter with Laythe in 6 hours. This survey has gone very fast, we'll be at Laythe within 3 days of the initial Jool aerobraking.

Aerocapture at Laythe occurs as Jool is setting, while our approach to circulate comes as the sun is rising.

Stone 1 will remain in orbit until it is needed by Project Garden.

• Project Stone completed
  
• Made orbital gravity measurements of Tylo, Vall and Laythe.
  
• 560 science gained
  

[Rath]

but oscar bs can run real rocket engines so ants must be real rockets or they don't burn it they just chuck it out the back

EDIT: double post

Edited February 22, 2015 by Rath
Bloody double post