Tempting Fate - the KSV Kraken | Pic-Heavy | Updated 3/19/15!

[Kieve]

This is a "mission" of a somewhat different sort.

My goal is to construct a space-capable aircraft carrier. Currently I am building this in sandbox mode, but when the time comes that it does not destroy itself (and the runway) as soon as I click Launch, the craft file will be transferred to a career mode, and constructed at a Minmus shipyard using EL and Karbonite.

That day may be some time in coming, however.

(for scale)

99% of my test efforts to merely get this thing to the pad structurally intact have resulted in horrible, spectacular failure and explosions. KSP physics absolutely hates this beast.

I named it the Kraken for irony, though I don't suspect I'd have any better luck if I called it something else. Still, it feels a bit like naming a ship "The Unsinkable" - not so much tempting fate as inviting it home for dinner and a roll in the sheets.

I will continue to log my progress here, in the hope that sometime this year, I come up with something functional.

- - - Updated - - -

So far, this bare-bones framework version has been the only one to survive the Launch button.

It's even capable of taking off, although it lacks the dV required for attaining orbit (let alone any kind of interplanetary shenanigans).

However, the COM and COT are horribly misaligned for VTOL use and the thing flies like a drunken brick (even with HX-scale SAS rings on each engine)

I suspect this may be my current "best" approach however - create a stable, working engine system first, then worry about the landing strip.

Edited March 20, 2015 by Kieve

[Kieve]

Tonight's update/progress (of sorts) has mostly been theorycrafting. I've continued to find B9 HX parts unreliable/unstable, even based on a reasonably solid framework. No pictures were taken of the subsequent explosions - they all start to look the same after a while. The flat structural plates have proven quite sturdy however, so I'm beginning to think my best course of action is to make use of the nodes on those (or flat attachment, where possible). Sets of downward-facing engines are a must under any circumstance - the Laythe landing will require some form of vertical thrust to set down smoothly.

Structural plates have always been part of the plan, due to their high impact tolerance (80m/s), but I'm thinking I may give the welding mod a try and fashion a solid hull to build on. Reduced part count is a perk, although not a huge issue right now. However, this means at least roughly determining the center of mass (for engine thrust) first, since part of that plan was to incorporate the "frame"-type plates as impact guards for the HX engines (at 187.5% scale, the 8m 4-way frames match up quite well with a 4-pack of HX engines). Transform gizmo can be used to fine-tune positioning as needed, but building a solid hull and welding it together, then finding engine frames are misaligned is just not something I want to go through, and I can absolutely see myself doing it.

On a related topic, I went hunting for previous topics / pics of Laythe aircraft carriers and was surprised by how little I turned up. The carriers I saw were all stock-ish builds for Kerbin, using either the grey plates or structural wings. BoatParts deserves a mention too, but I've found no indication anyone flew one of those carriers to Laythe either.

I know it must have been done, somewhere, but for the time being I have very little to reference. Anyone out there know of some Laythe missions that put a carrier in the water? I'm curious to see the results. Might give me some hints on how to continue assembling my sea-beastie here.

[Sampa]

Sorry, I saw this earlier and didn't comment simply because you blew my mind by your little project.

[Kieve]

Well... welding is a non-starter unfortunately. Attempting to weld B9's structural parts causes all the panel types to exist simultaneously in the welded result. I imagine this will be true of any part with a "part variant" feature. After tonight's tests, I'm starting to think maybe the entire project is a futile endeavor as well.

I have constructed a box!

This sorry structure is intended to test water-landing ability. It houses a 4-pack of HX HPDs with the generator, tanks, and power to supply it for a short hop to the end of KSC's runway, and into the drink.

Where it promptly sinks.

It landed at around 20m/s - not feather-delicate, but about as gentle as I can expect for something twenty times its size. Notably, none of the components except a battery survived immersion. And of course, let's not forget the sinking. That's kind of a problem too. (I really expected these structural bits to have buoyancy).

Let's try again.

This time I've given it some extra panels, thinking maybe they'd help it stay afloat.

I engage the engines, briefly, testing the "Plasma" hybrid thrust capacity - barely a quarter of the closed-cycle mode, but still enough to lift the cube.

It then obliterated the landing strip. Apparently even bumping KSC's runway with this kind of mass causes total destruction. Also, the flat panels on the bottom got stuck in the debris.

//Reload from Launch

The HPDs get us to the end of the runway on Plasma, but drain the battery much too fast - and more annoying, it's near impossible to hit that sweet spot between "slowing down" and "going up," which results in wasted fuel, power, and time bobbing up and down somewhere around 1600 meters. Finally I just kill power and let it fall a ways, using closed-cycle mode to brake the descent.

Light as a feather, my hapless Test Cube touches down.

And still destroys everything but the panels, then sinks.

Even for KSC, this level of failure is becoming discouraging.

[Sampa]

Don't give up! Remember America's early faiures in rocketry? You can master this craft...KEEP GOING! I am giving you 2 rep...one now to reward your efforts...and the other one when you finally complete this!

[Kieve]

Don't give up! Remember America's early faiures in rocketry? You can master this craft...KEEP GOING! I am giving you 2 rep...one now to reward your efforts...and the other one when you finally complete this!

Heh, thank you. But don't worry, I'm not giving up just yet. Only wondering what else I can try to get this thing seaworthy / spaceworthy. I don't like that it obliterates the runway every time, but if that's all it does (and the craft stays intact) I can live with it, given that I plan on having it built on Minmus.

Right now, I think what I'll do is look up some craft files for stock ships / naval craft (and even aircraft carriers, where I can find 'em) and try to sort out what works, what doesn't, and most importantly why. I like the clean, flat look of the B9 structural plates, but if they sink like rocks there's no point in using them for a hull.

[JAD_Interplanetary]

I've seen closed air-intakes and empty fuel tanks used as buoys to keep thinks floating (see here for an awesome example: http://imgur.com/a/dhHfr#0) Sling them under the hull where they won't be seen and ruin the look. I look forward to seeing how you solve the challenges.

[Kieve]

I've seen closed air-intakes and empty fuel tanks used as buoys to keep thinks floating (see here for an awesome example: http://imgur.com/a/dhHfr#0) Sling them under the hull where they won't be seen and ruin the look. I look forward to seeing how you solve the challenges.

Empty fuel tanks can be scaled up, and welded without issue - those may work. Intakes are another matter, they can't be welded and at the scale I'm working at, part-count would definitely become an issue trying to coat the bottom of a 500-kiloton super-carrier. Appreciate the suggestions though!

[Kieve]

For tonight's test, structural wings!

Reminds me of a carpenter ant at work. Hi ho, off the end of the runway we go...

The wing has an impact rating of 15m/s, which is significantly less than I'd like, but the same as the HX parts - ergo, if the wing can survive, everything else hitting the water above it should as well.

However, the radial decoupler is only rated at 9m/s. Oops.

Well, ejection force doesn't really matter anyway. The wing goes skipping away into the waves.

And slowly floats off.

Keyword: floats! Buoyancy, score! /victory dance

So as a structural / hull element, at least they're viable. They can be welded. And they're versatile too - so many shapes means I can be a bit more elaborate with the outer hull.

Of course, they also have aerodynamic / lift properties, so Ker knows what that'll do to Laythe reentry profiles. I may be begging for trouble on that point. But it's not like I haven't courting disaster from day 1 already.

Speaking of welding, this calls for further testing. Let's make an actual boat.

---

I stuck one of the little RCS balls on it as well - while welding the RCS thruster to it is a bad idea, the tank is safe enough. I want a little propulsion while in the water, if only to get the craft away from my carpenter ant. In fact, I think that's what I'll name the drone!

... and my boat can withstand 40m/s impact? With just those few pieces? At this rate, I could have a hull survive orbital drop and not even need the HPDs for slowing descent, the thing would survive terminal velocity. But we're getting ahead of ourselves - let's see how a welded boat handles itself.

It's like a silly hat. I didn't think leaving it in front was a great idea, even if I could just shove it down the runway skidding on its bottom.

Surprisingly, the aerodynamics have no effect while driving along. There's a reason for this, but for now I'll just be glad my rover isn't trying to fly away.

Quicksave at the end of the runway. It's a little top-heavy and I don't want to flub this...

OH COME ON!

//reload

There we go. Cruisin' on down to the beach. Anyone got a boom-box?

Hitting the water at a nice, gentle 7m/s. Even the decoupler survives this time.

Aaaaand, there's our fail for the night. I knew things were going too smoothly.

So, did you note that small wing-piece in the center? That was what I used for the root part.

As I came to realize, the default orientation plays a key role in welding - technically I've had the boat "on its side" this whole time.

And since that "boat" is also considered an aerodynamic lifting surface, the lifting part was also on its side. Doing absolutely nothing.

Meh. I'm in the water anyway, may go for a little cruise. RCS away!

*makes motorboat noises*

Overall, I'd call this a success, but it does present an interesting choice. If I construct the hull such that the wings are considered "vertical" by the game, I can nullify any aerodynamic influence they have and let the HPDs do their work as intended. However, that presents the risk that the entire carrier will want to roll over sideways when it lands. Laying the root horizontal brings aeronautics into consideration, but leaves the Kraken far less likely to capsize when it touches down.

I suppose.

[Mr Tegu]

You might be able to change the root part with the root tool, dunno if the welding will mess that up too though.

[Kieve]

You might be able to change the root part with the root tool, dunno if the welding will mess that up too though.

Oh, I know about changing the root with the gizmo, but "welded" bits are all one, single part. They'll take whatever default orientation the "root" of the weld is, which in this case was that central wing. All it really means is, I need to be careful and pay attention to how my Kraken's hull is "oriented" at the root, when I'm ready to weld it together.

[Kieve]

Tonight's failure is my patience.

Also more hiccups in the Welding department. But mostly just patience.

Spent a considerable chunk of time putting together a small-scale mockup of a carrier deck and hull. After welding the hull... well, look at the shaded-out piece next to the mockup and see for yourself. Apparently none of the "transform gizmo" uses were kept, and almost every single wing piece was completely out of alignment. Tonight was intended to be a test of the completed concept, on a reduced scale - however, given tonight's issue, that will likely wait for tomorrow night, or the weekend, when I can be assured of more time and opportunity to wrestle with failures and glitches.

At least I had some practice making boat hulls out of wing parts.

[smjjames]

You probably should look into Better Buoyancy if you haven't already.

[Kieve]

You probably should look into Better Buoyancy if you haven't already.

I did not know there was such a thing, but have now acquired it - along with Joint Reinforcement, which might help curb or eliminate the "explode on physics" issues I'd been having with the full-scale attempts.

Thank you!

[Kieve]

Tonight's update: A new welding experiment.

When I last spoke about issues with welding B9 structural panels, I was a little imprecise. You see, my attempt had been to weld the entire deck into one solid piece - the deck being composed of five 8m-square panels at 400% scale. And the result was that all the panels welded to the root panel flickered from being several different versions of themselves simultaneously. The root panel however, remained in its initial state.

It's a fine point, and I'm not surprised to have overlooked it at first, but what it actually means is this: as long as the root of the weld is the B9 panel, it will retain its variant as a welded piece. Which means I should be able to staple fuel tanks, structural fuselage tubes, wing parts, and anything else with buoyancy to the panel, create the weld, and have a sort of compound sandwich-hull part I can use to assemble a full and proper hull for the carrier.

Only theory at the moment - I shall test it when I have the opportunity, along with some other plans.

Test 1

-Weld a "sandwich" hull piece and confirm root B9 panel remains in static state. Flickering is bad.

Test 2

-Confirm buoyancy of welded piece. Note orientation, and material stats like impact and heat tolerance.

Test 3

-Will "static-welded" B9 panels remain static if re-welded? If the panels are "locked in" by welding, a single solid hull may be possible after all. Unlikely, but worth trying. The resulting impact / heat tolerances would likely be off the charts as well. Perhaps for my next trick, a flight into Kerbol!

Test 4

-Weld a "shielded" HPD engine. This is an HX-HPD with the exhaust port protected by a B9 panel set as a frame (see earlier Cube experiment). The Kraken is no good to me if only the frame (/hull) survives the landing, and I've already ruled out parachutes due to the scale and weight of the thing. (Note: not saying it can't be done, just that it'd be hilariously inefficient to do.)

Test #4 is particularly important to me, in the sense that if successful, it should also raise the engine's impact tolerance and make it less likely to disintegrate on splashdown. Trying to land a normal VTOL aircraft is tricky enough, trying to land an unbalanced super-carrier is just begging for trouble.

But then, I've been tempting fate since the beginning, so why stop now? Jeb, I have the perfect assignment for you...

[Kieve]

TGIF. Time to run some of those tests!

I've begun with a 4x4m panel, scaled up to 400%, and welded rows of structural fuselage to it, along with some scaled-up wing parts.

That impact tolerance. Yikes!

. . .

What's the terminal velocity for Laythe, I wonder?

Well, let's not get ahead of ourselves. Gotta see if it's sink or swim first.

The Goliath Ant is just a super-scaled version of the Carpenter. With lots of torque apparently.

Stop showing off, Goliath.

Seems I have some issues with scaling to fix yet.

May as well correct that when it gets back to the garage. Levitating wheels serve no one.

The panel is ejected into the water...

...where it promptly sinks. Discouraging, but we are not defeated yet!

Goliath attempts to shove the panel deeper in, hoping for some buoyancy to take effect.

Okay yeah, this is probably a lost cause.

Failure Count: +1

Back to the garage.

---

Panel #2 is strictly wing parts. I know those float, thanks to my earlier tests, so here's hoping more of them will help the panel float too.

It's interesting to note that position adjustments made with the transform gizmo held - the black wing lines are perfectly lined up with the panel edge, on both sides.

Maybe it was just the rotational adjustments that welding doesn't save?

Let's hope this one fares better.

Failure Count: +2

While I had it back in the SPH, I tried the compund-welding idea of Test #3.

Failure Count: +3

Wasn't expecting any different on this one, though it'd have been nice to have something go right this evening.

The attached panel still tries to exist in every variant at once - apparently Schrodinger's theories about observation are slightly off.

---

Panel #3 will be our final test for tonight. The fuel tanks have all been emptied, and I stuck wing panels on it again, for aesthetic reasons.

It's also worth noting that they haven't provided any lifting effect at all, in any of these tests - even when panel #2 was classified as "Aerodynamic" instead of "Structural."

Goliath shoves the third panel into the water.

Every time I eject these, there's a fun little explosion from the decoupler hitting the ground. Doesn't do any damage to the panel or the Ant, but it makes me jump just the same.

... Well crap.

Failure: 4

Kieve: 0

Total strike-out for tonight. Apparently while I can preserve a panel's characteristics by making it the root, that also means preserving its complete lack of buoyancy. No matter what I glue to it, the thing is dead weight in the water.

I may still go back and complete Test #4 to see if I can reinforce the HPDs for durability's sake, but sandwich hulls are a complete wash (har har).

As a side-note, I initially tried to upload these as an embedded imgur album, and only ever got a URL link out of it. What's the trick to embedding, enlighten me?

Edited March 14, 2015 by Kieve

[Kieve]

The further adventures of the Goliath Ant...

After having some time to reflect on my poor hull choices, I decided to run some tests on the structural fuselage piece.

Initially, I just had a Carpenter Ant carry one down to the shore and launch it to see if it'd float - which it does. Went skipping across the water like a flat stone, from the decoupler ejection force.

So let's try scaling things up a bit.

7m is the biggest TweakScale will go, but that's plenty big enough for anything I'd use these for.

Regular 1.25m size floats just fine, let's see how 7m works.

Well it's technically floating, but I had to clip below the waves to see it properly.

The standard-sized piece had no issue sitting nicely on top, so maybe my problem isn't with welding B9 plates, but with TweakScale - I'm beginning to suspect buoyancy does not scale up with the weight / size.

But making assumptions is rarely a good idea and many of my theories have failed already, so let's run another test.

I probably could've used the Carpenter for this, but Goliath is actually slightly faster and can get further into the water thanks to the big tires.

Oh yeah. Scaling definitely plays a part in this damned buoyancy puzzle - the 9-piece weld bounces off the top of the water for a bit before settling down.

So while previous experiments concluded that B9 panels have zero flotation ability, perhaps there's still the possibility of adding buoyancy after all. For now though, I want to try a different test.

---

The HPD and frame-panel, welded together. A small fuel tank, battery, SAS ring, and a box-probe sit on top, for control and fuel purposes.

Welding the HPD did some funny things to it, most notably with the engine modes. For those who haven't used HPDs, they're like a RAPIER engine, but with ion dual-mode instead of intake air.

Primary (closed cycle) mode has more thrust, but chugs LFO like a frat boy at a keg party. Secondary mode uses substantially less fuel, but sucks down electrical power instead.

The welded part however? It's locked into "primary" mode, which eats only Oxidizer and Electrical, leaving liquid fuel untouched. Strange.

Good thrust though.

Welding the panel to the bottom however, turns out to be a terrible idea. Firing up the engine destroys the panel - and, engine being rooted to it, the engine as well.

Kaboom.

So let's try reinforcing it a bit differently.

Girders and grey panels ring the top, well out of the exhaust path. It's hardly a perfect solution, but it's made the engine nigh-indestructible, which is definitely a perk.

Test-firing the engine reveals another funny glitch - no exhaust plume / effect at all. Invisible thrust!

This is also when I realized the part about eating Oxidizer and battery only - and it gulps oxidizer like crazy, so that's not really an acceptable trade-off.

Looks like welding the engines in any way is a no-go. While I love how sturdy it made them, I'd prefer not to sacrifice basic functionality, and having the fuel-efficient Plasma mode available for interplanetary burn is a must.

[Kieve]

Just noting that, while I haven't given up on this endeavor at all, I'm done postponing my entire .90 career restart for the sake of one sandbox project. Updates will continue for as long as it takes to get my Kraken operational.

[Kieve]

Testing time once again, but we're going full-scale. I want to see what it'll take to get the Kraken floating...

Structural fuselages welded together - 160 meters long, in all (8meter panels, scaled up 400%, x 5 panels). When I originally welded them, they were end-to-end, but the welding did something wonky to their alignment and so they're side-by-side now. I'm thinking this is a sign I should probably do welding in the VAB instead, apparently SPH alignments confuse it.

Oh well, let's get it out to the runway and take it for a swim.

Bonus points if you spot the problem already, because I sure didn't.

So we cruise on down to the end of the runway, quicksave, and prepare to head into the water. The deck becomes stuck on the end of the runway once the ground tips down, and refuses to go any further.

I scoot the camera down under the thing to decouple some wheels, hoping to lower its center a bit for traction, and...

Oh. That would explain it.

I drove down the runway on PhysWarp 4, and assumed it must've happened some point along the way.

//Reload from launch

Great. My center panel shrunk. And... the wheels are broken already? What the hell?

So I reload back to the hangar, thinking maybe more wheels will help. Now each panel has a rover strut across the middle as well as endpoints.

This actually goes worse - twice as many wheels break on physics launch (and this is with Ferram4's "Joint Reinforcement" mod as well, which includes an 'easing' of physics load!)

Physics Easing

Slowly dials up external forces (gravity, centrifugal, coriolis) when on the surface of a planet, reducing the initial stress during loading

All parts and joints are strengthened heavily during physics loading (coming off of rails) to prevent Kraken attacks on ships

Apparently the Kraken really has it in for my Kraken. But no way in hell am I giving up - I haven't even gotten to the actual test yet!

Let's see if landing legs perform any better.

In a word: No.

In multiple words? Still no.

While they do prevent the wheels from breaking instantly, two of the legs break on Physics start and refuse to retract. To compound that, once you pull the legs in, the wheels hit the ground... and break.

Clearly we need something with a higher tolerance for abuse. But what...? We still need motor function to get it down to the bay, and I am not trying to fly this brick yet.

Solution? B9 heavy-duty landing gear!

80m/s impact tolerance, steering and motor capability, and as a bonus they don't actually consume electricity. Kinda cheaty, but this is for Science.

Unfortunately, I completely forget that the Stayputnik probe core I'm using for control does need power. Oops.

Rebuilt with plenty of batteries and RTGs, because why not.

Quicksave at the end of the runway, to avoid being smote by avoidable disasters. (Smited? Smitten? Get off me, Kraken, no kisses for you!)

It's entertaining watching the rubber smoke trail up the line as each wheel hits a terrain bump or dip.

But it's night, and we're not getting into the water in the dark. Who knows what's lurking out there.

There we go. Shiny daylight. Let's go for a swim.

This crawler may be ungainly, but it slides into the water like a fish. And with much glee, I see we have FLOATING!

*happydance*

...

Wait, those panels aren't even touching the water.

Time to decouple.

...SONOVA-!!!

Well alright, that was only one pair of pontoons. And clearly the crawler was keeping it afloat so at least we know for a fact that they will stay buoyant under some circumstance!

Let's add another row and try again.

Version 2 runs into some trouble getting down the runway. While I enabled steering on the wheels, they all steer in the same direction, so it just kinda slides left or right and doesn't shift direction at all.

Back to the hangar to fix that.

Honestly, I could've adjusted them from the runway, but for one stupid problem - the scaled-up I-beams interfered with clicking the wheels, so I couldn't actually get at them to change it.

Besides, deleting the forward half of the crawler frame and cloning fixed wheels proved faster anyway.

I made the mistake of using "revert to hangar" rather than "recover" so it was night once again. I threw the brakes on, quicksaved (just in case) and time-warped until dawn.

Autobots, roll out!

Trying to control this thing on a downslope is a bit like guiding a landslide.

But we make it to the water safely...

I had to fuss with this one a bit more to get it in the drink - tidal motion kept washing it back up on the beach.

Eventually I got it into the water enough, I could test.

Moment of truth...

... May sperm whales feast upon your rubbery hide, foul tentacle beast.

I returned to the hangar with the intent to coat the bottom of the panels with pontoon prototypes - and the game crashed.

Very well, Kraken, if that's the way you want to play it...

[Kieve]

I've done some further research on KSP's wonky flotation physics, and I'm once again questioning the value of welding. The source of my doubt this time is part count - specifically, does one 5-part weld provide more buoyancy than 5 parts?

As always, further tests are needed, but I'm finding increasingly curious that the crawler frame (nothing but I-beams and landing gear) kept that landing deck firmly above the water, while the welded pontoons did nothing. I have a suspicion that higher part-count on floaty bits is actually going to end up being a requirement.

The other doubt I have is the Kraken - the KSP physics one. I've seen time and again how physics start goes bonkers over these supermassive creations I've been working with, to the point where I am genuinely concerned I will never be able to get the KSV Kraken back into loading distance once it touches down in the water. I have no clue what horrible physics glitches await me, but at this point I'm expecting nothing less than total disintegration, or just straight-up vanishing of parts.

[GreenWolf]

*furiously scribbles away on a notepad*

This is a very interesting project you have here. Boats aren't normally something you see that often in KSP, much less interplanetary aircraft carriers. I'm particularly interested in your ideas and experiments on bouyancy, since I've been toying with the idea of building a floating base on Laythe.

[Kieve]

*furiously scribbles away on a notepad*

This is a very interesting project you have here. Boats aren't normally something you see that often in KSP, much less interplanetary aircraft carriers. I'm particularly interested in your ideas and experiments on bouyancy, since I've been toying with the idea of building a floating base on Laythe.

Floating base wouldn't be too tough, I think - I-beams or structural fuselage parts for pontoons would serve you well, and either RCS on your base modules (slow) or a jet-engine tug (slightly faster) to push them around and get them docked. Test in Booster Bay to make sure your docking ports are aligned height-wise and you're golden.

The hiccups here all come from working on such a massive scale. Weight does play a factor in the buoyancy puzzle, and the Crawler weighed in at 320,000t, give or take a couple thousand. About a third of that was the deck - right now I'm estimating the completed Kraken to top 500,000t.

[Sampa]

Man, after seeing all the trouble you are having (shudders) I think I'll stick to going to space!

[Kieve]

Man, after seeing all the trouble you are having (shudders) I think I'll stick to going to space!

The greater the obstacle, the more glory in overcoming it.

~Moliere

The KSV Kraken has proven to be a huge pain in the tail, but seeing it finally break atmo and set down on Laythe will be worth it in the end.