Faghti

Rocket designs require testing. If I'm making a lander for Tylo, I am not going to wait for a launch window, do all of the mechanics of flying there, try out landing, and only then discover that my craft doesn't have enough SAS and is not sufficiently maneuverable. What to do ? I'm sure budding rocket scientists have worked out their own solutions. I rely heavily on the Hyper Edit mod, and even restore to previous save games. But I value career mode for the immersion. When it becomes habit to revert flight, or restore last save, the over-all game experience is cheapened. When that mountain peak is rushing towards you, and you are facing the wrong way, and you can't clear it, well, that was just a practise run right ? Here is what would allow us to test our rocket designs, while emphasising the importance of getting it right in career mode: The flight simulator. It looks pretty much like ordinary career mode except: We can warp our craft to anywhere we want to test it out. No Kerbals are harmed - its a simulator, right ? On exiting simulator mode, time, science, experience, funds, reputation and contracts are the same as when we entered the simulator. You don't need funds to build a craft - its a simulator. Availabile technology is limited to what you have already unlocked. If you haven't yet discovered it, you can't simulate it. One possible variation is that it is only possible to warp to places that are already visited. The key thing is, to progress in your career, you have to do that outside the simulator. You decide in advance whether your Kerbals are practising or doing it for real. Your Kerbals can practise landings in the simulator, but if they screw it up in real life, well, dead is dead. Exiting the simulator, you make a commitment. This time it is for real. Kerbals may die. Kerbin is depending on you !

If you define efficiency in terms of required delta V for a given craft then a constant altitude burn is always more efficient than a suicide burn (using the definitions of Plusk above). This thread has a particularly good video that illustrates a constant altitude burn, and some maths as well, for those that want it. Here is another thread that reaches the same conclusion: http://forum.kerbalspaceprogram.com/index.php?/topic/36960-how-to-calculate-optimal-descent-profiles/#comment-509383 Yes, the contstant altitude burn has steerage losses. Conversely, the suicide burn has gravity losses. Which are greater ? As Yasmi points out, a suicide burn has gravity losses. When burning close to vertical, most delta V is going into counteracting gravity, not reducing orbital velocity. It may not seem very much, but the difference is pronounced with a craft with low TWR. In the video in the link above by Cosmo-not (it is well worth watching !) the craft has a TWR of 1.1 at the start of the burn, increasing to 1.36 by the end. It would be very expensive to try to perform a suicide burn with such a craft, not to mention hazardous. Without doing the maths, I expect you'd have to begin at a great height in order to be able to burn long enough. One way of thinking about it is to recognise that much more energy is in orbital velocity than gravitational potential energy. Therefore we might expect that the method that most effectively reduces (or adds for take-off) orbital velocity is a good candidate. In the video you can see that as long as the burn is, most thrust is expended on reducing the orbital velocity to zero - with only enough vertical thrust to avoid crashing into the surface. But thought experiments don't answer the quesiton. The above threads refer to calculations concluding that the constant altitude burn is the most efficient. I haven't followed the maths. I did try an experiment - I found it much easier to experiment with takeoffs to orbit rather than landings. In particular, when taking off, a good compromise alternative to a constant altitude is to adjust your angle of flight so as to remain a few seconds away from apoapsis. In the limit, a constant altitude burn should be always exactly at the apoapsis. In practice, a few seconds off means that the craft is pushing the apoapsis forward and up, while remaining close to it. This allows the craft to gain height, and is easier to control than a perfect constant amplitude. If terrain means you need to climb more steeply, then aim for more seconds to apoapsis during your climb. The standard interface doesn't give you seconds to apoapsis, but mechjeb and Kerbal engineer do. I use the HUD view of Kerbal engineer to give me the read-out constantly. Here are some results from a trial that I performed on the Mun. The craft comprised 1 OCTO, 1 Z-200 battery, 2 Oscar-B fuel tanks, 1 LV1 Ant engine, 4 OX-Stat solar panels, 4 LT-05 landing struts, 2 MK1 illuminators, and a Mechjeb controller. Mass at landing site: 524 kg; TWR on surface: 2.45 I experimented with two styles of take-off: 1) As close to gravity turn as possible: turn to 20 deg at 20 m/s; Thrust prograde until apoapsis is at 10000 m. Cut thrust. burn at apoapsis to circularise. Total delta V: 651 m/s. 2) Immediately after take-off, turn over to the point where the time to apoapsis is neither increasing nor decreasing. That is approx 60 deg for this craft. At this point, time to apoapsis is a few seconds. The time to apoapsis will gradually increase - allow it to increase to five seconds - which means that the apoapsis is rising in height as the craft gains velocity; Progressively turn towards prograde in order to keep time to apoapsis approximately 5 sec. On reaching horizontal, progressively throttle down in order to remain close to the apoapsis (but always behind it). Orbit eventually becomes circular at around 5000 m. Increase to a 10000 m orbit in the usual fashion. Total delta v: 594 m/s. Not a big difference between the two - but it does refute the idea that thrusting along prograde is always the most efficient. I did try experiments on landing as well. Difficult to get a decent comparison, since landing sites for every experiment I tried ended up with landing sites at different heights - and the difference in delta V wasn't great anyway: With both approaches, most delta V is expended thrusting essentially retrograde, while travelling at high speed above the surface. Unfortunately method (2) above is more difficult to execute for landings. I don't have a readout that gives me seconds since apoapsis. And in any case, I'd have to descend steadily as well. In practice I prefer to be able to control where I land. I use a method that I think was published here: Use low thrust for most of the descent, with more vertical than retrograde thrust, to reduce descent rate and bring apoapsis down steadily during the descent. Edit: I also meant to add: Aesthetics is a big reason why we play KSP. For me, one of the most sublime parts of the game is skimming mountain peaks with a craft at sunrise over an alien landscape. Low TWR craft with low take-off and descent profiles maximise that. Why would you take off vertically from the Mun (or anywhere else) with a high TWR craft ?