Foxster

Saved a little over 1GB of memory using OpenGL. Game is more stable. Don't think I've had a single crash since doing this - probably because I'm not running close to the memory limit now. Only glitch is that when I alt-tab away from KSP in fullscreen and return, the Windows border stays and things are a little screwy in KSP. Pressing Alt-Enter fixes things.

Ah, yes, thinking about it yet again I see that is correct. It would work the same that way. Bamboo staging. Wasn't a great example of my idea. I'll try to think of something more illustrative.

Actually, thinking about it, I don't think what you suggest would work. That would result in the fuel being taken from the tanks from the top down i.e. just what this plumbing was getting around.

Well, yes, in this simple case. This ship was just to illustrate the point. The point being that you can set the order the fuel tanks get used in by the number of links to them - least first.

One way to use tanks in the order you want is like this... Isolate the tanks from the engine(s) with a non-crossfeed part (the I-beams here). Then stack parts that do crossfeed and connect each to a tank in the order you want the tanks emptied - bottom to top here. Engines will use tanks in the shortest numbers of links first. Here the bottom tank is one octagonal strut and one fuel duct away (two links). The next up tank is two octagonal struts and one fuel duct away (three links) and so on. And, yes, you can put a man into an 80K orbit and return him safely to Kerbin with this ship - albeit sitting on a command seat bolted to a parachute!

Here's JARS (Just Another Rocket Ship): Just add more stages to the side if you need them then work your way inwards as you launch and fly.

Yes. I think you are right. I have fairly complicated plumbing on the bigger ship and I don't think MJ is taking this into account accurately.

Thanks for all the info, guys. I tried to make both journeys as similar as I could. Used MJ for the launch with terminal velocity limiting on. Went to the same orbit. Then transferred in the same ways, again using MJ. Repeated it a couple of times to be sure. It seems pretty obvious just from looking at the craft that one is going to get further, it has tons more fuel and lots of that left by orbit. Its just that I always see DV seeming to be fairly fixed for particularly journeys, with the DV trees showing what you need to get places. So I have been struggling with why its not the whole picture after all. I think its the TWR I need to be looking at as well as DV. Its just that I was hoping for some simple thing (like I thought DV was) that will let me see that one rocket is going to get further than another. Seems its not simple though.

I have two ships. Both are lifters to get a load from the surface of kerbal to somewhere like Duna. Both have a 16 ton load on top before test launching. The first is a multi-stage asparagus rocket. The second is a simpler, lighter three-stage rocket. The first has, according to MJ, a delta V of 12,516 and the second has a delta V of 13,046. Just looking at it it feels like the first will get farther because of all the extra fuel it is hauling. And, sure enough, it does. It gets to Duna with a load of fuel to spare. It could probably make it back to Kerbal. The second though, despite it higher initial delta V, I can't even manage to get into circular orbit around Duna before its out of fuel. So there is obviously more to this than just the delta V numbers. I just dunno what. So, what am I missing here?