Flirting with disaster: Design oversight and poor planning conspiring for success (long!)

[Randazzo]

Mods used prominently in this mission:

• Modular Rocket Systems
  
• MSI Infernal Robotics
  
• Tweakscale (necessary for IR)
  

(pics at the bottom for TL/DR folks)

Having successfully begun fuel production in Duna's SOI, it was time to start making some cash again. To that end, three contracts for Jool were accepted. Exploration of Bop and Pol, and putting station in orbit of Jool. The station was simple and went off without a hitch (no pictures taken) for a very nice payoff. For Bop and Pol, I began to construct a pretty standard design with a lander and a main engine core. The lander would dock with the core for transfers, and separate for landings. This however, was boring and almost certain to succeed. The plan was scrapped in favor of a more complex design that had (entertaining) potential to fail.

The "Lewis" and "Clark" project was born! The landing module, "Lewis", from the original plan would be used, but instead of a simple in-line engine core, an entirely separate vessel would be constructed which could carry the lander in a cargo bay. As the lander would have very little clearance in the bay and to simplify docking procedures, the docking port in the bay would be placed on both an extending piston and a hinge that could swing the port to an outward facing. This concept was tested and proved rather wobbly. Following my standard design convention, anything wobbly simply requires more struts. With struts in place, the "Clark" was born. Since struts fix everything, no testing was performed.

The plan was relatively simple. The two vessels would be launched separately and rendezvous in LKO. The ship would transfer first to Jool for aerobraking into a high orbit, transfer to Pol (which is less off-inclined than Bop), then transfer to Bop from Pol. After completing the exploration contracts, the Lewis and Clark would return to Duna orbit rather than Kerbin. Both vessels would refuel at Kepler station, and Lewis would return to Kerbin on it's own. Clark would remain at Duna for use in future missions. The Delta V budget was 6700 m/s for the Clark while carrying Lewis, and 1800 m/s for Lewis to perform landings and orbital rendezvous manuevers. Lewis would be expected to have a small surplus of fuel to transfer into the tanks of Clark after both landings were completed, ultimately adding ~250-300 m/s to Clark for the Duna Encounter. It was expected to require ~1500 Delta V to make the Duna encounter and transfer to Ike orbit for Kepler rendezvous. To minimize the Delta V required from Clarks main drive to achieve a Jool encounter, the final orbital stage of the the Clark was designed to have enough fuel to take the ship into a 1000 km parking orbit over Kerbin after rendezvous.

Upon rendezvous in LKO, it was discovered that the docking array was indeed very stable. However, struts had been attached too far up the piston and were bolted to the moving cylinder instead of the stationary frame meaning it could no longer extend. Fortunately, the hinge still worked and Lewis was able to dock. The orbital stage worked perfectly, delivering the ship to a circular 1000km orbit with 70m/s to spare. It was jettisoned and Mechjeb was utilized to create a porkchop encounter. Unfortunately, using estimation instead of calculation, I had warped several days past the optimal transfer window before the launch. The transfer burn would require 1880 m/s of Delta V, as much as leaving from LKO, effectively negating any benefit of the high orbit. Another ~100 m/s was burned to make corrections and achieve an aerobraking periapsis at Jool with a desirable inclination.

Upon hitting the Jool encounter, Clarks course unfortunately did not very closely resemble the plotted trajectory. The periapsis had jumped to 150 Km and the inclination had steepened marginally. A small burn was made to correct the periapsis but nothing could be done to rectify the inclination without a significant expenditure of fuel at this point. The aerobraking maneuver achieved an orbit as expected, however it also threw the inclination even farther off from Pol. While it was not entirely necessary to match planes completely with Pol, the burn required to stabilize orbit and change inclination would consume another 1070 m/s. This was the second time insufficient planning punished me. Had I aerobraked higher and gotten a higher apoapsis, both the plane change and subsequent transfer would have been less costly. Regardless, the mission carried on. Transferring to Pol required 785 m/s, and 24 days in Jool orbit with a dangerous trajectory crossing the path of Laythe twice each orbit.

Arriving at Pol, I encountered yet another unexpectedly high cost Delta V burn. To circularize in a relatively high Pol orbit would require 980 m/s bringing the total remaining budget to 1885 m/s. Needing 1500 to get to Duna, then Ike and Kepler station, and still having to escape from Pol beforehand, the mission to Bop appeared to be in jeopardy. Scrapping the entire mission was considered at this point, but burning back into a useful Jool orbit cost only 100 m/s less than circularizing at 100 Km, so I carried on. Lewis was undocked successfully and moved to a lower orbit of 40 Km. Given the red zone nature of the mission and the premium value of fuel at this point, I decided to use Mechjeb to perform the landing in order to preserve as much fuel for transfer to Clark as possible. This however, did not go as planned. Upon activating Mechjebs landing autopilot, it immediately had a stroke and began flinging the Lewis about wildly while burning full out. I have no explanation for this, but it repeated this display after shutting it off and restarting it, so I left it alone and did the landing myself. I can't say precisely how much more fuel was wasted on this than would have been necessary, but ultimately I returned to Clark with more than necessary to perform the Bop landing, should we proceed with it. With the Lewis being lighter now, the Clark had gained about 120 m/s, putting it's budget just over the 2000 m/s mark. While certainly on the very edge of failure, I felt that with the remaining fuel being transferred from Lewis after the Bop landing and the fact that the Lewis would be significantly lighter, I would have sufficient fuel to get to Kepler and still perform the Bop landing. A transfer to Bop was performed (without plane change) at a cost of 485 m/s, immediately casting the mission back into doubt.

Arriving at Bop, the circularization cost 370 m/s which was slightly more than expected. Clarks Delta V budget was down to 1150 m/s. It no longer had enough to escape from Bop and then make the journey to Duna. Failure seemed likely, but we still had the landing and final fuel transfer to perform. Given Mechjebs propensity to bite me in the ass, I performed this manually from the start. With no AI plotting to murder my entire crew, the landing went smoothly, SCIENCE! was gathered, and Lewis returned to Clarks bay. Lewis had just under 600 m/s of Delta V with it's remaining fuel, but the amount of fuel it actually had was far less significant when transferred into the more massive Clark. Even with the extra fuel and the lighter Lewis, the total budget had only increased to 1316 m/s, and this gain was quickly lost when burning back into a very, very high Jool orbit. It took very little to circularize this orbit at 1.5 million Km, but at this point it didn't really matter anyway. The Clark had barely over 1000 m/s left, and could not return to either Duna or Kerbin. It would have been possible to achieve a Duna encounter for ~950 m/s from this altitude, but it would have been extremely risky. It would leave less than 100 m/s for obtaining a low enough periapsis to aerobrake over Duna, which is LOW. The Clark also had no RCS thrusters of it's own, only carrying RCS fuel to supply the Lewis lander after docking procedures.

Now essentially stranded, there were three options considered:

1. The Jool station sent earlier still had almost half a Jumbo of fuel and a Skipper engine. While fuel could not be transferred between these vessels as they did not have matching docking ports, the possibility of rendezvous and transfer of the crew to Jool station for returning to Kerbin was considered. However, the station only had a budget of ~2,000 m/s (which was intentional) and was at a 250km circular orbit over Jool. It would have very likely taken all or most of this fuel simply to rendezvous with the Clark, leaving the crew in the same situation.
   
2. Sending a fuel drone from Kerbin or Duna. Dead weight fuel is costly to transport, and especially if sent to Duna to be filled at Kepler, would have required a large expenditure of both time and money, leaving the Clark to orbit for possibly years in the vicinity of two moons. The risk of getting an undesired slingshot in this time was deemed too great.
   
3. Transfer all remaining fuel to the Lewis and abandon the Clark or transmit all science module (thermometer, gravity detector, seismic sensor) data and undock the Lewis and leave it behind to reduce the Clarks weight. Boarding the Lewis seemed to be the best option, as though the Lewis was packing only a single LV-909 engine, it was capable of generating more Delta V from the same amount of fuel when compared to an unencumbered Clark at a ratio of almost 4:1. It's small tank gave it a maximum budget of 1,800 m/s which was enough to return directly to Kerbin.
   

Choosing to abandon the Clark, the crew transferred into the Lewis lander and filled all tanks, including monopropellent, from Clarks remaining stores. This left the Clark with a little fuel, and after undocking, it was set on a full burn into space for dramatic effect. It was actually rather undramatic, lasting only a few seconds. While no longer stranded, the Lewis was still in a relatively close vicinity to both Bop and Pol at different periods of the orbit and was not in the clear yet. It would be almost 200 days before the Kerbin transfer window, and the burn would cost 1430 m/s. This left sufficient fuel for corrections, and with RCS as a backup. The landing would not be at a controlled location, but there was no longer a question of being able to return.

No moon encounters happened during the waiting period, though Bill did convince Jeb that the seat cushioning was made out of snacks. The burn was executed, and the course was corrected midway to achieve a 70km periapsis. This was again lowered to 40 at the three-quarter mark. The encounter did not alter the orbit significantly, and a very slight burn was done to push the periapsis to 32km. This was enough to create an orbit, but we had a few seconds of fuel left and to reduce the number of aerobraking nodes, the rest was used in a retro-burn to bring the AP just under the Mun's influence. Since there was nothing better to do, RCS was used to get as close to 0' inclination as possible for no particular reason. Soon enough, the Lewis was dropping in over a very dark Ocean and made a (not really, 6.7 m/s) soft water landing. Water landings were a touch risky, as the Lewis was designed to land on it's gear on solid ground to save on parachute weight. Fortunately it's empty tanks rendered it just light enough not to blow up the engine or the tank, which was nice since all the SCIENCE! bits were attached to the tank. The Lewis was recovered for some pretty darn phenomenal gains in both science and experience, and another harrowing mission came to an end.

Pictures!

Javascript is disabled. View full album

Edited February 14, 2015 by Randazzo