How to make a working seaplane?

[jrolson]

How do you make a fully working seaplane?

I can get it to land and float but I can't seem to gain enough speed to take off again...

Ive tried adding more jet engines but it just causes the plane to tip over on throttle up.

[Gianni1122]

@jrolson

I use KSP SXT to make a good seaplane

P.S. If you only want to improve the one you have already made without using mods, we'll need to see screenshots of your plane to help you.

Edited January 4, 2017 by Gianni1122

[Wolfair corp.]

Firespitter have good seaplane parts too.

Edited January 4, 2017 by Wolfair corp.

[MitchS]

Never built a seaplane on KSP, but I have an aviation background. Hope it's relevant to Kerbal engineering. Post pictures of your plane if possible, please!

The most unique issue with getting a seaplane to take off is essentially drag. Water has a huge resistive force on a seaplane, and that force has to be overcome with the thrust from the engine(s). The faster you go in water, the deeper you sit in it, the more surface area you have in contact with it, and the angle of your body relative to the direction of travel through it all dramatically increase this already large force preventing you from accelerating. So, for design, I would say:

1) Reduce drag. Decrease the parts sitting in the water--a wide fuselage and two very small outriggers under the wings if necessary. Mount them higher than the main fuselage, so that when the boat is sitting stationary in the water, they sit very gently on the surface. Once you start accelerating, the lift generated well below takeoff speed should be enough to lift them both out of the water, reducing drag.

2) Reduce density (float more!). Your plane should want to float, not sink heavily in the water, barely buoyant. To do this, I would empty fuel from the tanks that comprise your fuselage where it meets the water. Store the fuel high instead. This will also help manage your center of mass a little for point 4. It's okay to not have every tank full to the brim--for effective atmospheric aircraft, I rarely take off with more than a half tank overall.

3) Increase lift generation in water. Do this by gently angling up the front of your fuselage and under-wing outriggers relative to the water's surface, if your design allows. This way, as you start moving through the water, the water will be pushing your aircraft up and back instead of just back. (Look at the Grumman Albatross for reference... see the upward-angled belly and outriggers. http://4.bp.blogspot.com/-L6Pj0tK9Seo/TgDJUGolY6I/AAAAAAAAARE/m_f--qchV8Q/s1600/GrummanAlbatross10oClock.jpg) Sure, this will cause greater drag once you're airborne, but it's necessary for water takeoffs so you're going to have to accept that.

4) Increase thrust. I know you said you added more jet engines, but the placement is important. Consider a plane with engines on the wingtips. If you shut one engine down, the airplane intuitively spins out of control in the direction of the shut-down engine, right? Well that happens because the assymetric thrust creates a yawing moment on the aircraft around its center of mass. If you have two engines pointing through the center of mass and shut one down, however, the plane doesn't spin. This doesn't just apply for placing engines on the sides of the plane--it applies everywhere, including above and below. Thrust vectors above and below the center of mass cause unwanted pitching motions. To counteract this, you want to place the engine(s) in a place where their thrust vector--the line pointing straight forward through their center--runs as closely through the center of mass as possible, or at equal distances on either side of it. Maybe this means angling them up or down a few degrees, or moving them back or forward a meter or two from your initial design. Note: Real-life propeller engines are much more practical for this mass balance than KSP jet engines, which require you to mount the heavy parts rather far back. Turn on CoM and CoT markers in the VAB and move things around until you get something that lines up.) I assume you're placing your engines high, to keep them out of the water. This is wise, but the thrust vector still needs to be balanced in relation to the center of mass.

5) Increase lift. Make your wings bigger! More wing area means taking off at a lower speed, which means less time going fast in the water before getting out of it on takeoff and slower speeds for landing, which means a better seaplane. Yes, you'll have a much lower top speed, but that's what comes with making a plane that can do unique things. Tradeoffs. Which brings me to my last and most important point...

You need to understand that, again, engineering is about trade-offs. Especially in aerospace. If you want a boat plane, don't expect it to perform perfectly as a boat or as a plane (HU-16). If you want a plane that handles well at low speeds, don't expect it to reach high speeds (Piper Cub). If you want a plane that reaches high speeds, don't expect it to handle well at low speeds (F-104). Everything comes with its opposite in another category. My favorite simplified explanation of this is in the example characteristics of Range, Payload, and Maneuverability for military aircraft. You can only pick two, or spend both points on one at the expense of the others. In reality, there are dozens of parameters to balance.

Well, I've got to get back to work. Post a picture of your plane, and let us know if you still have problems after taking this stuff into consideration in your design. Good luck!

[klesh]

1 hour ago, jrolson said:

How do you make a fully working seaplane?

 

Take a close look at Squad's premade seaplanes Gull, Mallard, and Osprey for inspiration and engineering concepts.  You can see them if you allow default crafts in the difficulty settings of a new sandbox game.

[Aegolius13]

I just made a seaplane for Laythe science collecton, with Mk 2 body,  Big S delta wings and one Panther.  I started out with pontoons but it actually worked better with just the fuselage in the water - it could go faster with lower drag. The pontoons were too small to float anyway. 

Keys seemed to be having a lot of wing for slow stall speed,  and keeping the fuselage section as clean as possible.   I put the rear landing gear on the wings, and gave them some dihedral to keep them above waterline.  The Panther was a good fit due to excellent TWR at rest; a Rapier would probably have given me problems. 

[tseitsei89]

1. Lots of wing for low takeoff speed.

2. As little as possible of your plane should be touching the water since water gives  you huge drag and lowers the speed you can achieve before liftoff by a lot. So use light and buoyant parts.

3. An engine that has a bigger thrust at low speeds might also help but not so sure if that is needed...

[bewing]

One trick that seems to work well for me is to have an extra set of canards below the waterline. This really helps to get your nose up out of the water so you can accelerate.

[Mister Dilsby]

4 hours ago, jrolson said:

How do you make a fully working seaplane?

I can get it to land and float but I can't seem to gain enough speed to take off again...

Ive tried adding more jet engines but it just causes the plane to tip over on throttle up.

It's tough. I spent a while developing a design that could lift off fine on Kerbin, but was too low in the water (or whatever liquid that is) on Laythe to get up to takeoff speed there. 

If none of the excellent design suggestions given here work for you, you could always try a more... Kerbal takeoff profile:

Spoiler

 

 

 

[dangerhamster]

26 minutes ago, bewing said:

One trick that seems to work well for me is to have an extra set of canards below the waterline. This really helps to get your nose up out of the water so you can accelerate.

We were moored up against a high quayside in the family boat once, little boy looking down asks his dad what the trim tabs were for. Dad answers " those are the brakes son, when they want to stop they lower them".

 

 

[MitchS]

@Kuzzter, I belly-laugh every time you post on these forums. How's that published e-book coming along? Last I heard, you're considering about to talk to Squad about getting the rights to publish KSP material, right? I've got my paypal account all set up, just waiting for the word.

[AeroGav]

I made my first seaplane last week

https://kerbalx.com/AeroGav/K133--Curlew

I'm still in early experimentation but i  suspect that water, like air, is capable of making both lift and drag.  You want more of the former and less of the latter,   so try to keep non lifting (fuselage) parts out of it, and have surfaces interact with the water at small positive aoa.   I went with 5 degrees since that works well in the air, perhaps it's good in the water too.

I think hydrofoils a good move, gets the body out of the water so you can go fast. I don't know how the game models water drag.  Can a part be partially submerged, or is it merely considered in / out of the water?

It weighs a bit over 37 tons and has one rapier, two junos for jet power.  Panthers have best static TWR in the jets, but 1 + 2 allows me to keep a nice symmetry.   Full power gets me 15.5m/s.    Deploying the flaps on the hydrofoils ( 5 degree deflection) causes speed to drop to 12.7 m/s, but makes the body rise out of the water.    At this point,  I activate the nukes.   At sea level, the pair of them are worth less than 30kn thrust.  But this apparently takes us over the tipping point and speed SLOWLY edges up. Drag seems to decrease as we get faster.  Beyond 29m/s you can cancel the nukes and the jets alone will carry it to takeoff speed.

lots more info on the kerbalx page,  a couple of videos demonstrating critical flight phases.

[HalcyonSon]

1 hour ago, bewing said:

One trick that seems to work well for me is to have an extra set of canards below the waterline. This really helps to get your nose up out of the water so you can accelerate.

I like that.  I tried several different pontoon designs: empty Mk0 LF Tanks, Mk1 Structural Fuselages, 1.25m fairings... no luck with any.  I could land and float, but they were all too draggy on take-off, or resulted in a flip on landing.  Don't know why I never thought to make a hydrofoil.  

Hmm... now I wonder if anyone has built a trans-oceanic high-speed hydrofoil ship...  I bet I could design a stock carrier for reusable micro-science drones to satisfy a bunch of Kerbin research contracts.

[Snark]

I've found seaplanes to be not too hard to build... actually, most of my planes turn out to land and take off just fine on water, even when I didn't explicitly design them to be such.  In other words, I build seaplanes by accident. 

But that may be an idiosyncrasy of my design habits with airplanes.  I'm not primarily an "airplane guy", I'm not particularly adept at flying them, so I tend to overengineer them to have lots of wings and fairly modest takeoff speeds.

One factor that can make a big difference:  is your plane a "tail-depressor" or a "nose-lifter" for takeoff?  Both work equally well on land, but the latter tends to have an advantage in the water.

Here's what I mean by that:

Forget seaplanes for a moment, and think about a regular plane that goes down the runway and needs to pitch upwards in order to lift off.  Pitching upwards on takeoff basically requires one of the following:

• Push the tail down, or
• Lift the nose up

(of course, one can always do both)

Most "conventional" airplane designs seem to rely on pushing the tail down.  That is, they have the rearmost landing gear not very far aft of the CoM, and they have a tail assembly that sticks out pretty far behind the rearmost landing gear, and the tail assembly has elevators on it.  So when it wants to take off, the elevators on the tail push the tail down, which relies on the rear landing gear as an immobile fulcrum and thereby points the nose upwards.

It's also possible to achieve runway pitch-up by lifting the nose rather than depressing the tail.  This requires having canards (or similar) on the front of the plane.  The result works out to be pretty much the same, but it has some design implications.  Notably, it's not relying on the rear gear as being an immobile fulcrum-- it's exerting an upwards force on the rear gear, not a downwards force.

This becomes significant for seaplanes.  Tail-depressors tend to have difficulty lifting off the water.  They need to rotate up, but the way they try to do that is by pushing down on the tail.  That works fine on terra firma where, 1. you don't have to worry about pushing the rear gear down under the ground, and 2. your tail is held high off the ground to start with, so you don't have to worry about it smacking into the ground on takeoff.  But in the water, tail-depressors don't have either of those, and exerting the downward force on the tail causes the plane to sink into the water which brings up drag and kills your takeoff ability.

Whereas nose-lifters do better in the water than tail-depressors, because the mechanism they use for rotating up involves an upward force rather than a downward one, which tends to make the plane ride higher in the water, thus reducing drag.

I suspect that part of the reason I've had such (accidental) luck with seaplanes is that I tend to design my planes as being nose-lifters rather than tail-depressors, so they have kind of a "head start" on seaworthiness even when they haven't been explicitly designed as seaplanes.

[fourfa]

^^^ good info

I didn't see anyone mention flaps yet?  One way to generate more low-speed lift, without excessive high-speed drag, is flaps.  

Take an ordinary control surface (the bigger the better, for water takeoff).  For purposes of testing, let's say that you deselect all the control axes in the right-click context menu (so uncheck pitch, roll, and yaw), so now it doesn't respond at all to control input.  Now use the "deploy" function in its right-click menu.  For example if you have a relatively standard central main wing with trailing-edge control surfaces, the deploy check-box should cause the surface to move downward, and will now generate tons of lift at low speeds.  If for whatever reason it moves up, click the "reverse direction" box.  Presto, you should have a much easier time clearing the fuselage from the water.  Once you're flying though, that lift vector also adds a ton of undesirable drag.  So un-deploy, and it's just a standard wing segment again.

The deploy/undeploy can also be bound to keys in the action groups, if you've unlocked those.  If you haven't, just pin the context menu in the flight screen and click away as needed.

[HalcyonSon]

On 1/4/2017 at 3:45 PM, fourfa said:

^^^ good info

I didn't see anyone mention flaps yet?  One way to generate more low-speed lift, without excessive high-speed drag, is flaps.  

Take an ordinary control surface (the bigger the better, for water takeoff).  For purposes of testing, let's say that you deselect all the control axes in the right-click context menu (so uncheck pitch, roll, and yaw), so now it doesn't respond at all to control input.  Now use the "deploy" function in its right-click menu.  For example if you have a relatively standard central main wing with trailing-edge control surfaces, the deploy check-box should cause the surface to move downward, and will now generate tons of lift at low speeds.  If for whatever reason it moves up, click the "reverse direction" box.  Presto, you should have a much easier time clearing the fuselage from the water.  Once you're flying though, that lift vector also adds a ton of undesirable drag.  So un-deploy, and it's just a standard wing segment again.

The deploy/undeploy can also be bound to keys in the action groups, if you've unlocked those.  If you haven't, just pin the context menu in the flight screen and click away as needed.

Good info from both of you!  I couldn't figure out how to get flaps to work correctly.  Trying to trim them with [ALT] [ S ] was awfully slow and fouled the elevators.  I didn't even realize that "deploy" worked outside of the SPH.  Water landings have never been a problem, but then I'm stuck moving at less than 20 m/s - about 10 m/s below dead minimum takeoff speed.

This should reduce my need for high-banking S-turns on tight landings.  Just last night, I had to set down in a valley in the mountains. I hit more than 8g several times.  Had to get my speed down from 150+ m/s to 30 m/s so I wouldn't bounce off a roller or slam into the wall.  My designs typically stall around 15 m/s, but it's difficult to scrub speed.  I think flaps will make a big difference.

edit: Apparently [ S ]  is strikethrough..   But how do you remove it?  Got it!

Edited January 6, 2017 by HalcyonSon

[GoSlash27]

Honestly... I've never made an airplane that hasn't worked as a seaplane when I tried it. Nothing special about them that I'm aware of or any effort to make them work on water. It seems to me that KSP is extremely forgiving for seaplanes.

Best,
-Slashy

[regex]

Seaplanes are pretty easy to build in KSP. Having the weight be towards the rear or having some positive AOA built into the wing helps immensely. You might also want some empty tanks to shift fuel around to, helps fine-tune take-offs. Also, while traditional seaplanes need to keep their engines out of the water there is no such consideration in KSP.

Two I've built:

Spoiler

 

[jrolson]

Thanks guys for all the help, the SXT mod was very useful!