Design test V17 (requires mechjeb 1.8.1 and KSP 0.14.4)

[Hikaru]

This experimental design was created after I observed the differences in weights between vertical and radial decouplers, as well as observations made from the previous designs here http://kerbalspaceprogram.com/forum/index.php?topic=10747.msg163986

This is all stock save for the side mounted mechjeb, (1.8.1) and requires KSP 0.14.4

Several changes were made due to observed oddities in the previous design:

Huge amounts of drag (up to 75% during launch phase!) due to far too many boosters firing on the first stage.

Instability due to the ship\'s center of weight being fairly high at parts of the launch phase.

Weights of radial decouplers and vertical decouplers was observed to be a large unexpected factor in spacecraft weight.

Fuel inefficiency due to the liquid engines firing as soon as possible, rather than after the solid boosters had finished firing.

Some of these observations were a big surprise to me. For instance, the radial decoupler is half the weight of a vertical - so it makes more sense to have a wider ship than a taller one simply from that. Since I also gain more stability during launch, this also helps in ways I did not expect.

Mechjeb 1.8.1 added an interesting feature for its autopilot for launching to orbit which gives you statistics for how your ship\'s thrust is being used. I was shocked when I discovered 3/4 of all the thrust being put out by the first stage of my ship was being eaten by drag alone.

I was even more surprised when I discovered that simply by letting the solid boosters act alone in this design rather than having the liquid engines fire right off the pad the ship actually has slightly more fuel left in orbit.

This particular variant of the design can get to a 100km orbit via mechjeb with a tiny bit of fuel remaining in the launch phase tanks with four full tanks on the lander. The ship can then land on the mun, as well as fly back and land on kerbin with the remaining fuel.

[Hikaru]

This is a minor revision to V17; after thoroughly testing it the design unfortunately was topheavy and tended to flip over during the second booster stage. This revision prevents this problem by putting all of the boosters at the bottom of the ship.

It puts quite a bit of torque on the entire ship while firing the boosters due to weight and thrust changes, but the winglets appear to be enough to keep it on course during this difficult phase. Honestly I\'d like to come up with something less failure prone, but it appears I will need to either rethink the entire design from scratch again or come up with an idea I haven\'t yet to correct this problem.

[mGGCwnAnnwn]

The problem is its width. You can use radial decouplers without building a horizontal ship aswell if you want to save on weight. Simply stack them and leave a small gap for ejection.

[Hikaru]

The problem is its width. You can use radial decouplers without building a horizontal ship aswell if you want to save on weight. Simply stack them and leave a small gap for ejection.

Sadly the design had more problems than just that. I\'ll explain in my next post.

[Hikaru]

After doing several launches with the previous design I learned it *also* has an absurd amount of instability. Randomly, launches will fail, or work. I can\'t really explain it - but the workaround appears to be to add struts to the top of the boosters and attach them somewhere higher on the ship.

This does NOT fix its torquing at launch time, nor could I fly that beast on manual. So, I redesigned yet again...

After doing several more redesigns of the booster stages, it seems that it\'s simply easier to add an extra three liquid fuel tanks to the stack than it is to make multiple booster stages work. To give some idea, the complicated booster staging I\'ve been working on for something like 23 revisions. I\'ve never been able to make it work like I wanted - this ship design is the first one that seems to fix all of the problems I have been having without driving me crazy by being weirdly unstable.