Kerbal Engineer estimating wrong dV?

[Gordon Fecyk]

I have this monstrosity I've constructed a couple of times now. First by four Kerbal-X rocket launches, and later by three cargo SSTO trips.

Since this first iteration, I've removed the Mk1 inline cockpit and added three more Liquid Fuel tanks. I also removed a little bit of mass here and there, used smaller parts where I could, and lightened the load as much as I could while still being reasonably comfortable to a six-kerbal crew. Aside from KER, I have DMagic's EVA Struts in use.

If I just cheat this monster into orbit, I get slightly more than 3500 m/s according to KER's estimate. Enough for Duna and back. If I launch it as intended, the first module containing the Mk1 crew cabin and engines will have in excess of 7200 m/s. But then I start piling on modules, and the estimated delta-v doesn't seem to change once it's fully assembled. 

I have one of these out at Duna in a career play-through that has slightly less than half of its fuel, and has landed the rover package on Duna's surface. Yet KER thinks I still have 3700 m/s.

First off, the Iktomi II, in Duna orbit, has 257 parts, 56.57 tonnes, and 1866 / 4800 Liquid Fuel. This is the ship that already dropped off its rover kit. Supposedly I have over 3700 m/s. Assuming one unit LF = 5 kg, that's 9.330 t fuel, making dry mass 47.34 t and dV = 1397 m/s. Enough to get home, barely, if I'm careful with Ike and Mun assists, or I could dump a lot of hardware first.

Second, the Defrahnz, in Kerbin orbit, is prepared for departure to Eve. This has 306 parts, 75.92 tonnes, and is fully fueled at 4800 / 4800 Liquid Fuel. Supposedly I have just over 7400 m/s, but in reality with 24 t fuel I have only 2980 m/s. At least this craft could reach Gilly and land on RCS alone, and refuel.

Is KER not taking into account the added mass of modules as I assemble this thing? Have I missed a setting somewhere?

[Tyko]

I've had this problem before. With my experience it was due to my Root part not being in the upper stage. The root was somewhere in the middle and this threw off KER's logic.

Edited December 7, 2017 by Tyko

[Gordon Fecyk]

On ‎12‎/‎7‎/‎2017 at 5:52 PM, Tyko said:

With my experience it was due to my Root part not being in the upper stage.

That's interesting...

There was another case where I had to Klaw a Class D asteroid with what ended up being a pretty small craft. By the time it reached the rock I'm pretty sure I had a km/s or more, but after catching the rock I had only 125 m/s. That's the behaviour I would have expected. Fortunately, I only needed a few m/s to steer the rock away from Kerbin.

I wonder, then, if I rearranged my staging to put my active engines at Stage Zero, as the engines were part of the first module of this craft and would have contained the root part. I'll check that when I get home today.

[Update] Well that didn't work. No matter what stage the original root part was on, it had the same incorrect dV amount.

[Another update] KER seems to deal with rocket-launched components better than space plane-launched ones. The same version of KER on the same version of KSP (1.3.0) is correctly calculating the Iktomi's final dV as I add components to the craft. Maybe it has something to do with the original craft being launched from the launch pad and being the last component left; this was the case when I sent a pusher craft to that Class D rock.

Edited December 14, 2017 by Gordon Fecyk