swjr-swis

Yes, as core part of the game, as long as core functionality of the game requires such band-aids for it to work as one could reasonably expect it to work. The docking ports mentioned by Frozen_Heart are perfect examples: in KSP, docking ports are simulated essentially like giant magnets, held together flexibly through open space by invisible force, instead of like the mechanically clamped constructs in real life. It's easy to spot this because even small movements of a craft or forces will visibly displace docked modules/payload and swing them back and forth. During the stresses of flight, or combined with still existing self-amplifying phantom forces, it's enough to destroy a craft mid-flight or shake a station in orbit to pieces. Here's a quickly put together example. Strategical use of autostruts in just two parts of the craft (the two ports in the center) stabilize the construction. But as soon as those autostruts are disabled, it starts swinging away: Craft file: https://kerbalx.com/swjr-swis/Wobbly-Strutter Longer video of the craft showing the noodly bendiness, and the spontaneous wild swinging that develops from standstill, after disabling the autostruts.

You asked for an example of something that needs strutting to keep structurally sound; I provided one. The cause of the structural deformation is irrelevant, it exists. In this case it was exagerated to make the report issue easily reproducible, but if this was not something that consistently exists in the game I would not have been able to use it for that purpose. Clipping is inevitable and quite required in many cases (place a gear without clipping if you will), many parts have invisible colliders outside of the visible surfaces... clipping is not an 'excuse' for accepting structural deformations. The deformations exist and are real, they happen because of coding, not realism, so strutting is still very needed to counteract artifacts of the code.

Sure. http://bugs.kerbalspaceprogram.com/issues/12486 Ignore the bug report itself, but download the 00a craft file. Launch, and without staging, toggle the gear. Watch how the central tube of intakes and the gear get all scrambled up due to internal collisions of the gear with the individal intakes they are attached to, just by the action of retracting the gear. This can only be prevented by either replacing the stack of intakes by a single long booster, or by strutting things up, both of which have been done in the 00b version of the craft. This sort of deformations happen all the time with KSP craft, not just with gear. Mostly they are minimal and practically invisible, but they are one of the main reasons planes and rockets inherit tiny little torque moments when launched.

I actually don't know how. So never mind... instead of support the cause, just read 'there's a feedback report for it now', which is really what I meant to say.

They still are though. They wobble less, but they still do deform for rather uncontrollable reasons (invisible colliders, for one).

Support the cause There's a feedback report for this exact thing: http://bugs.kerbalspaceprogram.com/issues/12596

Visuals for autostruts are problematic, since they reach right across the ship and attach from the parts' CoM, not the surface. Even in the best circumstances, the endpoints will always be clipped inside parts and will end up invisible anyway. I don't see how it can be visually pleasing or aesthetic in any way to see just sections of the bars sticking at random angles through the skin of parts. Think about it, it can come to no good. And I simply do not understand this aversion to options. If you personally want to limit the functionality of autostruts to the VAB/SPH in your game, you can already do so right now: just don't use it in-flight. However, the reverse will not be possible if they hard-code it to only be available in the VAB/SPH. We can argue about 1001 possible situations where for one or the other it would be legitimate to use autostruts, but why do we need to justify everything? We all play this game our own way, and some of us don't even play the same way twice. It would be nice if people would stop trying to force their limits and constraints to this amazing sandbox on others. As for autostruts being considered 'magic': I'm quite willing to concede to the realism of not being able to attach invisible struts that phase through parts... just as soon as the game realistically implements designing and building a ship's structural skeleton, skin, and internal parts separately and fit-to-specs instead of ready-made one-use-or-fit-intended-only lego pieces. Until then: autostruts, clipping, and every other trick to compensate for the lack of real world construction techniques are fair game, as far as I'm concerned.

@IronCretin, an entry for the stock 1.2 board: JunoSpeeder-1h, 757m/s survived top speed with only 9x Junos. I wanted to see how low I could go on the engine count and still get a reasonably competitive entry. Craft file: https://kerbalx.com/swjr-swis/JunoSpeeder-1h

Another entrant! Now things get interesting. But if it uses the Whiplash engine from the Supersonic tech (containing Mk2 parts), is it truly strict? Surely we'd need to restrict ourselves to Juno and Wheasly only... Any of the jets can be made into a manned missile, limiting the selection would not stop that. In fact, the Whiplash (the regular turbojet you suggest) became the 'rocket' engine of choice for a lightest craft to orbit challenge not too long ago.

I'm doing this in the 1.2.0 pre-release, default settings. I am using autostruts, I did say that, but those are tweakables on the parts, not cheats. I recorded a short video of a 102G turn, if you want to see, but you can also download the craft and try; that's why I make the craft file available, allows people to verify the results. If you prefer to consider autostruts a cheat in the context of this challenge, mention it in the rules, and feel free to move my entries to the cheats leaderboard. I have no problem with that.

I think that's good clarification. Thanks. The only thing maybe missing to remove questions/doubts would be attaching one or two craft files that could benefit from an action group for 'control from here', like a VTOL or an upside down lander subassembly in a cargo bay that when staged has the wrong docking port assuming control.

I've been able to make airplanes safely pull 99G turns at Mach 3 the past few days, making them effectively unbreakable, using nothing but a few autostruts; where in 1.1.3 and before they would've spontaneously disassembled at below Mach 2 turns even with tons of regular struts. And the cargo bay construct in my previous post here was made rigid the same way - without them it started to undulate like a giant caterpillar the second it was put in orbit. Autostruts definitely work, but they have to be placed selectively. Use the minimum possible, try to place the root part and the heaviest part on opposing ends of your construct, and alternate autostruts between heaviest and root (and where appropriate, with grandparent). Also it helps if the heaviest part is a symmetric set, because the autostruts will connect to all of the symmetric group, essentially creating a truss-like connection.

TL;DR: make your plane over-controlled, and yank back on the stick at exactly the right combination of speed and altitude. You need to tune your control surfaces way stronger than would be preferable or comfortable for regular flight. You want to try to slam your plane to present the bottom of its main wings into the airflow as quickly as possible, making it into one giant deployed airbrake. So try give them high authority and a large arm relative to the CoM so you get lots of torque. Normally that's a very very bad idea as a) you would black out and it would possibly kill you, b) it tends to result in spontaneous craft disassembly (which would probably kill you), and c) aircraft maintenance crew will hate your guts even if your plane survives (and they may plot to kill you). But in KSP and with a little help from 1.2.0 autostruts, we can disregard these things. My process, with a RAPIER craft (other jets have different optimal altitudes): Get up to 20km, and speed up to Mach 3+ first. Don't waste too much time optimizing your top speed, just up to whenever you stop accelerating. Then start diving at a 45 degree pitch-down while keeping as much speed possible (follow prograde, or at least stay close to it) You start picking up a bit more speed as you dive, but at one point you'll notice the lower thicker atmosphere starts slowing you down again. Keep going, this is not your layer yet. At around 8km, pull up hard, as if going into a loop; keep holding until the G meter drops. You will have lost some speed, but the thicker air makes up for it by allowing you to make a tighter turn. The optimum altitude varies depending on the craft, and sometimes around 5-6km results in more Gs while others between 8-10km. So try several times at different altitudes, it can vary 20-30 Gs easily. You want to still be flying as fast as possible when you initiate the turn, obviously, but since it's the control surface effectivity that will most directly affect how much Gs you're gonna pull (how quickly can you present your plane's belly into the airflow), the air density where you turn is almost as important, maybe more. So you need to find the optimum balance of air thin enough that it doesn't slow you down too much due to drag, but still thick enough to let your control surfaces slam your plane around. Happy G hunting, and pack a probe core for when your kerbal blacks out.

A bit premature to think no direct response in this thread means they have made a decision on this, especially during a very hectic pre-release phase. A feedback report in the bug tracker would be more likely to be seen.

I do, they can be useful both as pure airbrakes and as control surfaces, but it's often limited to bigger craft due to their size, and I tend to build planes too small to fit them in a visually acceptable way. (hint hint Squad, would be nice to have a small set too...)

Ok, time to add an entry to the single-engine board. Name: SpeedFighter6 (still horrible with names) Top speed: 1519.3m/s (1519.3 points) Highest G-force endured: 99.4G (99.4 points) Parts: 51 (no points deducted for staying under the limit) Cheats used: none (5 points) Total score: 1618.7 points. Link to album: http://imgur.com/a/ZriO9 Link to craft file: https://kerbalx.com/swjr-swis/SpeedFighter6 Pics (more in album):

You have an Mk2 cargo bay in between them. Cargo bays are very finicky at the time of deciding what they 'occlude' from airflow, and parts right outside of them can suddenly, for no apparent reason, be decided to be occluded. This is most often noticed by the dreaded 'Cannot deploy while stowed' message of things like landing gear or panels nearby a cargo bay, but it also affects things that provide lift or cause drag, in which case you get unpredictable behaviour of your craft as sometimes the lift or drag is there and sometimes, without apparent reason and no real change, it's not.

The heat gages/bars are actually from two PB-NUK RTGs that are clipped in one of the tanks behind the kerbals, a bit hard to make out from that angle. Just look at the thermal details in the menus: they never deviated more than a few degrees from the starting values. 2000/2000 for the RC-001S probe core, which was the first to blow up, has higher thermal tolerance than a kerbal (800/800). The kerbals should've poofed before that. They didn't because they remained at around 300K the whole time. Just to compare I went and redid the flight, but with the command seats out on top of the plane. With clear results: So heat and drag are not treated equal. Drag still happens in the cargo bay, heat doesn't.

If you're ok with some functional clipping, you can add two of the passive large radiators panels, and offset them into the cockpit. When I do this for the purpose of thermal protection, I place them in mirror on the sides, and clip them in and halfway back. The next parts in the fuselage will hide the exposed half of the radiators. At the expense of a tiny bit of extra drag, mass, and some EC usage, you can make the Mk1 cockpit survive extreme heat from reentry (or Mach 3+ speeds in the lower atmosphere).

I have noticed very high drag values when kerbals are inside bays/fairings and should be shielded, in both 1.1.3 and 1.2.0pre, but I'm getting conflicting results about heat. Posted some results in this album. I'll add the pic with the visible drag arrows in your bug report, for confirmation.

It looks interesting, but I get a feeling this is more of a showcase than a challenge.

Look at the drill again. By the end of it, the astronaut in the back (sorry, I can't remember his name right now) is actually in a slow spin now, having started at full standstill.

The drill at 5:45. They are close to the floor, but if you look closely you'll see they are not actually touching it. They can't bem because the wild leg movements would bounce them off in the direction of the ceiling.

They can, to a limited degree. It would be very hard to use that to impart a significant force on another object though, like enough to flip a rover. It's an awful lot of effort just to rotate themselves 90 degrees...

This one maybe (check at 5:45): @Red Iron Crown, this the one you meant?