AeroGav

Basic Mk2 Spaceplane Guide

Recommended Posts

BASIC MK2 WHIPLASH/NERV SPACEPLANE TUTORIAL


The best looking airplane parts in the game almost certainly belong to the mk2 family.     The part descriptions combined with their "hypersonic" appearance make them a natural choice for a new player, unfortunately they generate more forum help threads than anything else.

Pic - the wrong stuff !
 

Spoiler


T03zbxI.jpg

 

 

Problem 1 - Poor Performance

Mk2 parts generate at least twice as much drag as a mk1 or mk3 fuselage built to carry the same amount of fuel or passengers.   As a result they are frequently unable to break the sound barrier.  If the player tries to overcome this by spamming engines,  they often run out of fuel before making orbit.


Problem 2 - Flipping Out


The aerodynamic forces generated by draggy fuselages are not properly taken into account by the stock game's Centre of Lift indicator,  resulting in a CoL much further forward than it actually appears.     

Second, new players usually try to build something that looks like a sleek real-world airplane, with a cluster of engines at the back and a long, pointy fuselage up front.  On launch, all that fuel at the front of the ship balances the heavy engines at the back.   But when the tanks empty, CoM shifts far to the rear and the plane becomes unstable.

Problem 3 - Exploded Cockpit


In recent versions of KSP aerodynamic heating effects are much stronger at the front of a stack than they are further back.  As a result,  despite its much higher heat tolerance a pointy mk2 cockpit is much more prone to overheating than a mk1 inline one with a few parts in front of it.

Mk2 Tips - the short version

 

  1.  To minimise the drag penalty, keep the mk2 fuselage as short as possible.  Use it for Kerbals, cargo and other mission stuff, but avoid storing fuel in mk2 parts except for when you're fitting a mk2 bicoupler or mk2 to mk1 adapter anyway.
  2.  If your ship has NERV engines, fit as much wing as possible but use only fuel containing big-s wings and strakes.  You can in fact store all your liquid fuel this way.   More wings allows the craft to fly higher and at a lower angle of attack for any given airspeed, which reduces the amount of drag (and heating) on your fuselage.
  3. Use an inline mk2 cockpit

Combating the dreaded rearward CG shift as fuel burns off

This problem is such a (night)mare, I'm going to have to give its own list 

  • For your first mk2s,  stick to crew ferries or combi ships.  Passenger cabins and inline clamp-o-trons give you a bit of mass you can put up front to balance those heavy engines.
  • Try to shift engines forward if possible, especially the heavy ones.    For example, consider putting the heavy nukes on the wings or on pods either side of the main fuselage, whilst the lighter jet engine(s) can go on the attach nodes on the back of the main fuselage.
  • Try to move your fuel stowage rearward.  

Given the problem of instability, why am i telling people to move anything backward ?  Well,  if you've succeeded in balancing the weight of those heavy engines when empty, your next problem is that you've probably got more fuel tankage ahead of CG than behind it.  When you fill the tanks, your craft becomes excessively nose-heavy.  This is more likely to be an issue if your craft has a CoM well to the rear - sure, you can move the CoL aft as well to make it stable, but this places most of the fuel tanks ahead of CG.

When you  fill all the tanks it becomes a lawn dart.         So,  you need to find ways of increasing the fuel tankage at the back of the ship.        Consider putting batteries, reaction wheels and engine pre-coolers at the front of your size 1 stacks and only have fuel tanks immediately in front of the engine.     Big S wing strakes, orientated vertictically, can be used to build tail fins and they have a surprising amount of fuel capacity.   Also, you may be able to attach big S strakes to the trailing edge of the wing.   Big S strakes hold more fuel for their size than the main wing, so help to shift the fuel balance rearward.

Pic  - our original failplane, better balanced (but still melty and draggy)

Spoiler

sbhDxBM.jpg

 

Some mods that will make your life much, much easier

RCS build aid 


Shows a red dot in the SPH/VAB which indicates where your CoM will move to when the tanks are empty.  Makes it much , much easier to build planes that don't flip out on re-entry.


CorrectCoL

 

CorrectCoL does two things.   Firstly, it makes the blue CoL indicator in the SPH more accurate by taking into account aero forces acting on fuselage parts.  

 Second, it shows a stability graph for your airplane across the  AoA range.  When the line is above the horizontal axis, it indicates your plane wants to nose up.  When it is below, it tries to nose down.   You want the line to slope downhill left to right, so that the greater the AoA , the stronger the nose down tendency.    The point where the line crosses the x axis indicates the pitch attitude the plane will tend to adopt without any control input from the pilot.
 

A Quick Word on Part Attachment and Drag

It is worth remembering that KSP craft files are organised as a tree structure.  There is a root part, to which any number of items can be attached radially,  as well an end attachment node to which subsequent parts can be attached, and so on.     


To the game, a typical mk2 spaceplane, with an engine on the back and a mk1 sized engine nacelle attached either side of the fuselage,    is just like a rocket with a pair of boosters attached radially to the main stack. In this case we have the main stack (the mk2 fuselage)  and the engine nacelles which are like boosters.

6qNTT1x.png

(I've numbered the parts in terms of how far away they are from the root in the craft file)

Rule 1 : Every end-attached stack of parts must begin and end with something pointy.

"Pointy"  means something which has an attachment node at one end, which joins it to the stack, but not at the other.  Obviously, it also means something low drag.

Parts in this category include in order of more to less streamlined 

⦁    nose cones and fairings (Duh !)
⦁    jet engines
⦁    rocket engines (more drag than jets, due to the attach node on the back, but has been reduced in the last patch)
⦁    air intakes (somewhat draggy, inline intakes like the engine pre-cooler are better if you're min-maxing efficiency)
⦁    shielded docking port (quite a high drag part, you are better off with an inline clamp o tron if you must dock)

Rule 2 : When joining parts together with end-attachment,  the diameter of the attach nodes of the parts being joined MUST match.

Just because the game lets you join a 0.625m small nose cone to the front of a 1.25m fuel tank,  or lets you put a Poodle on the back of a mk2 fuselage, doesn't mean you should.    These mismatches create huge drag.   

If you need to attach a 0.625m part to a mk1 stack, use a low drag adapter like the FL-A10 or the NCS adapter between them.      If you want a Poodle (2.5m rocket) on the back of your mk2 spaceplane, use a mk2 to 2.5m adapter.

 

One last thing.   When attaching stuff, be aware of orientation.   End attached parts like fuel tanks, intakes, nose cones, engines etc. have lowest drag when aimed directly prograde or retrograde.  Angling parts a few degrees away from pro or retro increases drag.   

 

Just how bad are mk2 parts for drag ?

Gu5YdJf.jpg

A screenshot is worth a thousand words.   You can see that the mk2 inline cockpit has over 3 times the drag of the big S wing.     The short mk2 cargo bay has a little bit less,  since it doesn't have the sticking up bit at the top with the windows.      It's not in this picture, but the mk2 to 1.25m short adapter has about half the drag value of the short cargo bay.   And you can see that the 1.25m nuke rocket engine was very little drag at all, by comparison.
 

note - you can bring up this data any time you want by pressing ALT F12, going to the physics menu, and clicking on the Aero tab.     The big dialog box with that welter of aerodynamic data is enabled by checking the "display aero data GUI" option.    The drag info in the right click menus comes up when you check "enable aero data in action menus"

part 2 to follow (examples of common part attachment problems)

Share this post


Link to post
Share on other sites

Fail Gallery

Here is a list of the most common part attachment methods which look fine, but lead to excess drag.

Example 1

99pH9UL.png

We'll start with something obvious .  At the back of this fuel tank is a mk2 attachment node, but we've stuck an engine with a mk1 sized attachment node on it.  This mismatch creates loads of drag.   If you only need one engine on the back, use the mk2 to mk1 adapter !

 

Example 2

 

zunaUsu.png

This sequence shows a method to create a custom cockpit.   Looks really sleek no?     However, the offset trick makes no difference to the game engine, aerodynamically it sees the first picture.   So we've got a mismatch between the rear face ofthe mk2 to mk1 adapter (mk2 sized attachment node) and the front face of the cockpit (mk1 size node).  At the back of the cockpit is a mk1 sized node, joining a mk2 fuel tank, with a mk2 sized node.

 

Example 3

On first glance, there doesn't appear to be a problem with this setup -

uVSRc7Z.png

Spotted the problem yet?

Kq7Cey5.png

The eagle-eyed among you will have noticed that if the nose cone was attached to the front of that FT-800 tank, it would have lit up green when it's parent part (the fuel tank) was highlighted, just like the Terrier engine did.        The problem is that both the Type B nose cone and the FT800 tank are capable of radially attaching to other objects,  and when chopping and changing a vessel it's quite easy for this to happen unnoticed.

yP7zvj1.png

Why is that a problem?  Well , instead of one pair of "booster" stacks either side of our main fuselage, with a pointy nose cone at the front and a somewhat pointy nozzle on the back,  we now have two pairs of booster stacks,  one of which lacks a nose cone and is presenting a flat face to the air stream, and the other (the one with only a cone) has nothing covering its back end, again just a cone.

Example  4

300px-Mk3EngineMountNodes.png

Welcome to the mk3 engine mount, beloved of those poor souls trying to create a Space Shuttle replica.    As such, it is almost always seen with 3  mk1 sized (1.25m) engines hanging off the back, but what a lot of people forget about the single large 2.5m node in the middle.       Unused attachment nodes create huge drag - plug them.   If you don't want to put a 2.5m engine on here,  you should fit a 2.5m nose cone to cover it up.

Share this post


Link to post
Share on other sites

Part 3 - Building a Basic Whiplash NERV Mk2 Shuttle

The basics 

Going to orbit means accelerating to orbital velocity where freefall (the centrifugal force of hurtling around the planet) cancels out the force of gravity.

The closer you get to that speed, the more of your weight freefall supports, but until the effect is strong enough to support 100% of your weight, you require lift from the wings to hold the plane up.      Having more wing area means you can fly higher or at a lower angle of attack, for any given airspeed, both of which reduce drag.   

RgEkzDg.png

Spoiler

 

What's the optimum angle of attack ?

In supersonic flight, it's between 4 to 8 degrees above prograde.

 

Achieving low drag in rocket mode is going to be doubly important in this design because we're using the relatively weak NERV engines.    One RAPIER produces three times as much thrust as these monsters, and at 3 tons each, you simply can't get the sort of TWR on NERVs as you can on RAPIERs,  at least, not without adding so much engine mass as to undo the benefits of their vastly better ISP.

Rule of Thumb - Engines

You need about one nuke per 15 tons of mass.    One airbreather per 30 tons.

In this design we're keeping it small and basic, so we're shooting for a 30 ton gross weight trimotor - one jet, two nukes.

For our jet,  I normally recommend the RAPIER.   It has more thrust above mach 2,  and airbreathing top speed is 300-400 m/s higher.   This means that instead of the rocket stage needing 1100 delta V,  it only needs 600-700 or so.   

However,  the bottom end is feeble and in a draggy mk2, this can make it hard to get through the sound barrier.   In a beginner ship, that matters.    On larger, more complex vessels, I like to go with 1 RAPIER and 1 Panther per 50 tons.     The Panther's strong bottom end compliments the RAPIER's cammy nature, and this combo is lighter than two RAPIERs.  However, for this project, it's about K.I.S.S.

So, we're going with a single whiplash.

Serving Mass

Altar_Servers.jpg

First things first,  our engine pack is going to weigh 7.8 tons (1.8t Whiplash,  2 x 3 ton nuke).     The front of the ship needs something to balance this out when the fuel has gone, so we're adding a 2 ton passenger cab behind the inline cockpit.  OK ,  I guess this is going to be a crew ship then.   Since it's a crew shuttle, it makes sense to fit an inline clamp-o-tron too.   Reaction wheels, batteries etc?    It would be lower drag to simply put size 1 batteries and reaction wheels in the engine nacelle stacks, but if we make our mk2 fuselage too short it can spoil the looks, which is after all part of the reason for building a mk2.   

Overall that's just over 4.5 tons of dry mass at the front of our ship.   The engines are still heavier, but it's not Jabba the hut  playing with a kid on a seesaw bad.    The CoM will be nearer the back than the front of the plane, but not unmanageably so.

At the front, it's important to use a nosecone with at least 2400 temperature rating.   Again , if we had no cargo bay , we could have just used a procedural fairing for the nose cone (2600k heat tolerance) and it would give us somewhere to put irregular stuff like RTGs.    But, you'll see I find another use for that cargo bay later.

 

26xPGvg.png

Alrighty, now let's fit our engine.

V3gMtFM.png

I'm leaving a bit in this oxidizer tank because it can run some Vernor engines to power our RCS system - and using the benefit of foreknowledge - this tank is fairly close to where the finished craft's CoM will end up.  Why didn't we just put shock cones on the front of the engine nacelles containing the nukes?  Well, for a start the pre-cooler is lower drag.  Also,  this method creates a bit of space to tuck some nukes in where they won't get in the way of the main wing.

Going nuclear

sTXTpds.png

Type B nose cones can attach radially - we glue them onto our rear adapter to give us something to hang the NERVs off.

ylaHIMZ.png

Unfortunately, the curvature of the mk2 to mk1 adapter has caused the nose cones to attach at an angle.   We'll loose a bit of thrust if the NERVs aren't pointing straight back,  and remember what i said about parts having increased drag if not aligned prograde/retrograde.   To remove stray angles, select the rotation tool [3], and make sure you're in Angle Snap mode [C] rather than fine rotation.  Select the nose cone, then press [F] to toggle Absolute mode rather than "local".  Now when you rotate the part it should snap to dead-ahead.

Finally , use the offset tool on the Type B nose cones to tuck things in a bit, but not so far inward that the nozzle of the nuke engine clips into our Whiplash and not so far forward that the front of the nose cone pokes into our cargo bay.   And that folks, is our fuselage basically done.

gI4lxNk.png

 

edit - more to follow - wings !

Edited by AeroGav

Share this post


Link to post
Share on other sites

Part 4 - Bingo,  Wings

hGagCyR.png

The cargo bay makes a nice straight thing surface to put your main wing on

T2YbJbP.png

But of course, that's nowhere near enough fatness.    You can attach stakes to the back for more area, then strakes to the strakes.  More than two though, it starts to look silly. So, I move on to the leading edge.    Unfortunately with a short fuselage like this,  I have to stop at one leading edge strake or we end up with a wing overhanging the front of the fuselage.

uXibfhq.png  

But that's not the end of it ! We can attach a strake to the cockpit or crew cabin as a leading edge extension to smooth the transition to the main wing.   Strakes are awesome.  What looks better than one strake? Why two strakes of course !    OK that's about it for horizontal surfaces.   Just don't forget to fill all of these with fuel.

hrBMLEF.png

Just because we're done with wings, doesn't mean we can't have more stakes.   Rotate 90 degrees, they also make good tail fins for yaw stability.  Tail fins that hold fuel. 

Bear in mind they usually attach facing slightly off prograde , because the top side of the wing isn't flat.  Use the rotation tool, in angle snap (C) and absolute (F) modes to straighten them.

g3l4cRd.png

I like to cant these outward 15 degrees, since i think it might provide some of the benefits of roll stability.  Also it makes you look more like a 5th gen fighter.  But it's easier to attach rudders when they're still stood up straight.

Voila, our finished airframe :

1cEgtCr.png

Note, the last screenshot draws attention to how useless the SPH delta V reading is.   It's got that by averaging the ISP of our Whiplash with the Sea Level ISP of our nukes.     What it does tell us, is that we're just under 30 tons right now.

Edited by AeroGav

Share this post


Link to post
Share on other sites

Part 5 - Trimmings

qmXIhpT.png

Since we've got a clamp o-tron, we should probably pay lip service to docking.   The problem is, RCS ports are draggy, especially the 4 thruster blocks people love to spam.

A nice stack of reaction wheels takes care of swivelling our nose around, without creating any drag.   You can see that I've also stuck a pair of RTGs on the side of our Reaction wheel stack.   Unlike solar panels, they work in the shade,  and you won't break them by forgetting to retract your solar array before re-entry.

But what about translating sideways, left, right, up , down ?    I've attached Vernor engines with 4-way radial symmetry to our rearmost reaction wheel.  This RW is very close to our CoM , so translating won't create torque that the reaction wheels can't handle.    For backing up, we have the Vernor engine i'm currently pointing at.

To translate forwards, without using the main engines, we've this one..

gzgLqPL.png

Of course, these RCS thrusters won't work if the cargo bay is closed.    But, when you're in space, you can just leave that bay open.    When you're in the atmosphere, the bay will be shut, and these Vernors won't create any drag.

Since we don't have any thrusters that use monoprop, don't forget to empty the mono tanks of the Clamp-o-tron and Cockpit.

Action Groups

This is a simple airplane so we only need one.   Create an action group to toggle the nukes on and off.   That's it.

 

Adjust your dress

It's time to do some very short test flights.   What you need to do, is angle the front strakes (the ones that attach to the cockpit/passenger cabin, not the main wings) upwards slightly with the fine rotation tool.  What you need to do is find the right angle that makes the plane settle into a very slight nose up attitude - only 1 or 2 degrees above prograde - even after you release the controls, with SAS off.      If you want to know your airplane's angle of attack precisely, ALT & F12 to bring up the cheats menu, go to Physics, Aero and tick the "Aero Data GUI" checkbox, it's all in there. 

What is the purpose of this step ?

Firstly, it means if our pilot has a fatal heart attack or is otherwise distracted, the airplane doesn't immediately nosedive.

Second, the elevons at the back of the plane , raise the nose by pushing the tail down.   By dialling in this slight uptrim, we're eliminating the need for them to push down, generating drag and subtracting from our total lift, in normal  cruising/gliding/accelerating to orbit flight.

Third, it makes our airplane a bit better balanced for part 6.  

Note that this adjustment screws with the blue CoL indicator in the SPH, so it's worth leaving this step till last.

 

Tweaking Controls

Right click on the control surfaces to disable features we don't want.   Our Rudders should only control yaw, and not try to roll or pitch the airplane.

The Inboard elevons on the trailing edge of the wing should control pitch, and nothing else.    The outboards should do roll.

Also, use the "limit authority" slider to bring the inboards down to 60% and outboards to 40%.    Why would I want to do that?  Well, when you limit authority, you are limiting the maximum deflection angle.  Control surfaces , like wings,  gain lift as their deflection angle relative to the airflow increases, but most of the gains are had at small angles, and there's not much extra to be had past 15 degrees.  At 30 degrees they stall completely.

The drag produced by the surface , on the other hand, builds up exponentially as the deflection angle increases.   For our elevators, limiting to 60 or 80 is ok.    For ailerons, I prefer a  limit of 30 or 40.     Firstly, because when going to orbit, you're not making sharp turns,  you're just making tiny corrections to keep the wings level and to correct single degree heading deviations as you try maintain a course of 90 degrees due East. 

Second,  when you roll to the right,  the aileron needs to generate more lift on the left wing to pick it up, so it's deflection angle increases.    If it deflects to a large angle, it will create a lot of drag out near the left wingtip, that's not being generated on the opposite side,  so the plane ends up yawing to the left - in the opposite direction to which you're trying to turn it.    Limiting deflection angle cuts aileron drag, minimising this undesirable behaviour.

 

What of the canard ?   I'd normally say, set it to control pitch and nothing else.

Spoiler

 

I'd also normally set the "limit authority" slider so that it's not going past its own stall angle , which just creates more drag.

Remember, the angle of attack of the canard surface is  = AoA of the airplane  + deflection angle of the control surface

Let's just say that flight testing reveals our airplane has enough pitch authority to pull 20 degrees AoA max.

We're using a standard canard, with a max deflection of 15 degrees.      Given that all aero surfaces stall at 30 degrees in KSP,  we don't need that canard going past 10 degrees on this airplane or it's just going to act as an airbrake.   So we'd limit authority to 66% (10 degrees being two thirds of the 15 degree normal max deflection).

 

However, on this airplane I'm going to say disable Roll, Pitch, and Yaw.   I have other plans for this device, which we'll cover in the next part  "Precision flight school"

 

Edited by AeroGav

Share this post


Link to post
Share on other sites

Part Six - Precision Flight School

 

Flying a spaceplane to orbit requires you to adjust pitch angle in 1 degree increments or less.   At high speed, exceeding the desired AoA by a small amount creates huge drag.  At Mach 3+,  a 5 degree climb angle can send you over the max effective altitude of your jet engine in a matter of seconds.   A  5 degree dive will soon have you in a part of the atmosphere that's like treacle with huge drag, heat and fuel consumption.

The problem is, keyboard control is "all or nothing".     KSP's patchy joystick support can make it impossible to calibrate analog controllers with sufficient sensitivity.

What other options are there ?

Technique 1  - SAS  Stability Assist lock

iufb2ro.png

Simply pressing T activates SAS in it's default Stability Assist mode.  In this mode, SAS makes control corrections to try and hold the nose attitude same as it was last time you touched the controls.   At high speeds,  due to the curvature of the planet, this results in a slowly rising AoA.    When it gets too high, pressing any of the control keys, resets SAS's target attitude and drops the nose a couple of degrees.        For the same reason, this technique is less useful in situations where frequent corrections are being made , since every time you correct a roll or a yaw, the nose drops, which you might not want.

Technique 2 -  Pitch Trim

Pitch trim cannot be used while SAS is active.  However, if your plane is stable enough to fly without it, and you're not in a situation requiring rapid changes in pitch, it can give very precise control.     

Hold down the ALT key and press S to increase nose-up trim.   ALT + W to decrease it, and ALT X to remove all trim.

The effect of trim is like pulling back very gently on an analog stick and holding the controller there at the same angle throughout the flight, minus the wrist cramp.

 

Technique 3 -  Real Time Authority Limiter Adjustment

 

eWmmzHA.jpg

Have a look at the above pic.       We've got SAS enabled, and it's set to Prograde hold mode

SAS will make inputs to the flight control to keep the plane on Prograde.  Unfortunately, prograde is zero aoa, which means zero lift, and we're falling rapidly towards the sea.

Now if we click "Deploy" on our front canard, the lift at the front end generated by this surface, would normally cause the plane to settle at an AoA of about 10 degrees.

However, because SAS is controlling the rear elevons, it starts generating extra lift on those trying to push the nose back onto the prograde.

qizzO0v.jpg

This results in an AoA of 3.9 degrees, which generates sufficient lift for level flight whilst still keeping drag down.

You can actually control the climb rate very precisely by tweaking the "Authority Limiter" slider on the canard as you spot changes in the flight path.   You don't actually need to make much in the way of normal control inputs (the very occasional heading correction, if that) because the Prograde Hold is taking care of things for you ! 

Note that if you're maxing out the Canard's authority but still not getting as much AoA as you want, you can start limiting the authority on the elevon instead, which is fighting it.

 

Part 6b - Our Flight Profile 

In parts 4 and 5 I took you through the thought processes involved in creating a spaceplane.   I've decided to call this ship "The Flying Lamb",  and you can download "The Lamb" here.  

https://www.dropbox.com/s/cqgnxgf81kkyxi4/Flying Lamb.craft?dl=0

However, the flight profile should be applicable to many other spaceplanes with low/moderate TWR.

 

Initial Climb

After takeoff, we should try to climb into thinner air before attempting to go supersonic.

How high?   This depends on the wing area of the aircraft.   At some point, you won't be able to generate enough lift  without either breaking the sound barrier or increasing the AoA above 5 degrees, where drag builds alarmingly.

What AoA ?  THe main thing is, stay below 5 degrees.    At low subsonic speeds, optimal AoA is about 2 degrees but as you near mach 1 it starts to increase.

How fast ?     Try to stay below 240 m/s until you're ready to cross the sound barrier.

How steep ?   This depends on the TWR of your airplane.   Control your AoA first and keep an eye open for speed - when it starts creeping over 240 m/s it's time to go supersonic.   The Lamb climbs quite shallow, only 10 degrees at times.

 

Between 0.75 and 1.3 Mach is the transonic region.   Drag is unusually high here.   After mach 1.3,  drag decreases again - not as low as it was when you were subsonic, but by this point the ramjet effect will have greatly increased the power of your engine, so good times ahead.

 

Time to go Chuck Norris Yeager and bust that sound barrier

Start levelling off the climb as 240 m/s nears.    When velocity exceeds 260,  lower your AoA a bit, maybe 2 point something, and let the plane fall into a shallow dive.  Use the action group you prepared earlier to activate the nukes.    The Lamb doesn't really need them at this point, but it's good practice.

At 1.3 Mach (about 440 m/s) turn the nukes off and gradually raise the AoA back to 4 degrees or so, so we begin climbing again.     

 

The Speed Run

 

Your goal here is to reach the highest possible speed in level flight on your airbreathing engines.

For the Whiplash, 15-17km is  a good altitude.       Above 17km, engine power falls away more rapidly than drag.    The Whiplash peaks at mach 3 (900m/s) then thrust falls quite rapidly.    By 1100 m/s it will be down to half the power it was making at 900, so keep your expectations realistic.

Spoiler

 

(for the RAPIER,  20-22km is a good speedrun altitude.  Peak power is at 1150 m/s and you'll still have 80% of that at 1300m.s.   Then the decline is more rapid, but most of my craft can hit at least 1400) 

 

 

 

During the speedrun, just try to stay in the optimal altitude band when your're nearing your top speed.  I found the "Tweaking Authority limiter" method the easiest way to achieve this.

When should you call it quits?    If you want to be scientific about it, watch the Aero Data GUI and compare total thrust with total drag.     When total drag becomes more than two thirds of total thrust,  I light the rockets.

 

Run for Orbit

Start the nukes.   At this point, you might have a very low AoA as you were trying to stop the plane from cilmbing above 17km.     Remember, in supersonic fight, 4-8 degrees AoA is best.    Too low, and you won't make much lift and be stuck in the thicker part of the atmosphere longer than necessary, creating drag.   Pull the nose too high however and your excessive AoA creates a bunch of drag.      Initially i might keep the AoA to 3 or so while i squeeze the last drops from the jet engine, but as it fades into irrelevance it's best to put aerodynamic efficiency first.

As you get higher the reaction wheels start to overpower the aerodynamic controls and we find our AoA dropping.   This may be a good time to switch to Stability Assist mode instead of Prograde hold.  In this mode,  AoA slowly rises due to the curvature of the earth, so periodically tap the down key to stop it getting above 8.

At 36km the Navball switches to Orbit mode.   Be aware that if you're still in Prograde hold mode   , this causes the nose to drop a few degrees.  I normally click it back into Surface mode till i'm over 70km .  

As your velocity exceeds 2000 m/s, think about switching to map screen view so you can see your AP, that way you know when to cut the engines.

I normally reach orbit with about 1300dV left in the tank on this vessel.

 

To do -    Improvements to the Lamb - adding wing Incidence

Edited by AeroGav

Share this post


Link to post
Share on other sites

Section 7 - Advanced Construction - Wing Incidence


Remember this screenshot from part  1 of the guide, where I showed how even the short mk2 fuselage parts were generating 3 times as much drag as our largest wing ?

Spoiler

 

Gu5YdJf.jpg

 

We've done our best to work around this, but the Lamb still accelerates sluggishly in places.   It gets to low orbit with a sizeable reserve, but not enough to land anywhere, even Minmus If we weighed her down with extra fuel, she might not make orbit at all.

There is a powerful technique for reducing drag from draggy fuselages,  called "wing incidence".

images?q=tbn:ANd9GcTzYWVlCW1TJEfbP4M0SO5

If the wings are angled up like this , we can reduce the AoA that the body flies at while still giving the wings sufficient AoA to make lift. 

 

Theory - Choosing an Incidence angle

Optimal supersonic lift/drag ratio occurs at about 5degree AoA.     So, we could angle our wings up at 5 degrees,  then just put the nose on prograde hold so the body is at 0 AoA and the wings are at their optimal 5.      This would give great efficiency during the closed cycle mode part of the climb,   but as you remember from part 6b, there are times when this produces far too much lift.        During the speedrun, we were trying to keep our altitude below 17km.   When passing through mach 1, we were trying to shallow dive.    

On the basic version of the aircraft, AoA was as low as 1.5-2 degrees in parts of the speedrun.   If our wings had 5 degrees built in incidence,  we'd actually have flown with the body at an AoA of more than 3 degrees nose down.

So, the ideal figure is probably a compromise. You could argue that 3-3.5 degrees of incidence is probably the midpoint between our AoA in the speedrun and that used during the close cycle climb,.  Whatever value is chosen between 0 and 5,  it is guaranteed the aircraft will perform better with some incidence, than with none.

 

In Practice - Constructing it

 

No matter what incidence angle we decided that we want,  accurately applying it is a problem in the stock game.      

The rotation tool  is far too coarse in Angle Snap mode, offering only 15 degree increments. In fine mode, it doesn't actually tell you how much angle you're applying.

Note that when first attaching the wings to the fuselage,  you had the option of holding down SHIFT and pressing W  to angle the wing up in 5 degree increments.  Once the part is placed however, you can't adjust that without removing and reattaching the wing.

The easiest option,  if you're on a PC, is to install the mod "Editor Extensions Redux".

With this mod, pressing C no longer just toggles between Fine Adjust and Angle Snap , there are now many angle snap increments such as 1 degree and 5 degrees.

You can edit this list of presets,  and set your own..

AgTB2PW.png

vKrqVto.png

 

Traps for the unwary

jAdding incidence to the wings changes the position of the blue indicator in the VAB, making it hard to judge whether your CoL and CoM are in the right place.   Adding incidence to a wing that is mostly behind the CoM shifts the blue indicator backwards.      Adding incidence to one at the front of the aircraft sends it moving the other way.    

For this reason I recommend constructing the airplane with flat wings first, and doing your adjustments to get CoL and CoM in the right place, before adding any incidence angles. 

 

The other, major risk , is that the main wing, behind the CoM, can end up stalling before the canards or strakes ahead of the CoM.   If this happens, the plane might appear stable at  low AoA, then at high AoA  the nose snaps upward uncontrollably and goes into a flat spin.

To avoid this, it is imperative that you give all lift surfaces ahead of the centre of mass the same , or more incidence angle , than was applied to the main wing, so that they stall first.CorrectCoL's stability analysis graph can predict high alpha instability -

Spoiler

 

OwVB4pG.png

Although the line slopes downhill initially . meaning the pitch down tendency gets stronger as the nose rises, after about 15 degrees the trend reverses.   At some point, it starts trying to nose up instead.   A nice surprise for you on re-entry !

With the canard angled up to a greater angle than the main wing, this tendency is gone - the canard's lift always drops first,  and it always pitches down stronger the more the nose rises. In addition, it won't try to nose dive when you let go of the stick.    You can see that the line crosses the horizontal axis at about 5 degrees nose up - that's the angle the plane will settle at naturally if no-one's at the controls.

6Bsyss0.png

 

 

Finishing touches

After angling the wings and strakes, it's always nice to fine tune the airplane so it settles at a small positive AoA with SAS off and no-one at the controls.  Airplanes with built in wing incidence often try to push their nose below prograde hard when no-ones flying them, especially if the airplane is a canard.

 

The Lamb Mk2,   flight test

https://www.dropbox.com/s/dthki99nqbfn4yb/Flying Lamb II.craft?dl=0

How does our little SSTO handle with 2.5 degrees of wing incidence ? Here's a video of it in action.    My notes -

Initial climb angle is closer to 15 degrees instead of 10, because there is more engine power left over after the drag has taken its toll.  Going supersonic was much easier, if i hadn't intervened and  gone into shallow dive/triggered nukes, it would still have gone sonic anyway in a shallow climb.

Max airbreathing speed is over 100 m/s faster.     Flying to orbit with the nuke engines,  drag in the upper atmosphere was never more than 30% of our total thrust, and for much of the time was significantly less.    It would have taken a significantly greater degree of mishandling to stop this aircraft from accelerating.   Alternatively, we could have increased our fuel load without problems.

Overall, we got to space with an extra 350 delta V in reserve.  A mission to Minmus is now a distinct possibility.

Edited by AeroGav

Share this post


Link to post
Share on other sites

This is quite nice and interesting! Nice!

Share this post


Link to post
Share on other sites

Appendix 1 - Mk1 Size Spaceplanes

Note - this is not a standalone "how to".    To keep things short,  I'm assuming you've read the previous sections on mk2 designs.

So,  what are the drawbacks of mk1 parts?

  • Heat tolerance.     Actually, this is easily worked around.   Using inline cockpits and keeping crewed parts at least two modules back from the front of the stack, aerodynamic heating is not a problem.
  • Payload/Flexibility .   Obviously, since there is no mk1 cargo bay, bringing items to orbit is an issue, no matter how small.    On the other hand, the performance of mk1 spaceplanes is such that there may be little need for the interplanetary transfer ships launched by cargo planes.   Also, it is still possible to put items in space if you get creative with fairings etc.
  • Rigidity/Strength.       The strength and rigidity of mk1 to mk1 part joints is lower than with larger diameter components.    Coupled with the fact that each part holds less useful stuff,   there is a limit to how big you can build, even with autostruts.   I have built a 100 ton mk1 before but it did flex slightly with normal flight loads,  and would be more prone to breaking up on a hard landing than a 30-40 ton ship.

Advantages of mk1 parts

  • vastly reduced drag means craft accelerates better if you fly accurately, or more tolerant of inaccurate flying, or able to fly to orbit with a lower TWR / higher fuel load.
  • reduced fuselage drag means fewer handling surprises.   Don't you just love it when you pitch up hard on a mk2, and the drag from the long, draggy forward fuselage continues to torque the plane around into a deep stall, even though the plane looked stable in the SPH?  This happens less (or not at all) in a mk1
  • fuel tanks the way you want 'em.       If you're building an oxidizer free ship,  you don't get useless oxidizer tanks forced upon you with the mk1 to mk2 adapters and bicouplers. You aren't forced to lug around empty monopropellant tanks when you specify an inline docking port.

Here's the starting point of a MK1 deep space crew shuttle.

WHeEYFr.jpg

I've gone with four crew modules in the main stack (3 cabins and a cockpit).   One crew cabin on either side in the engine nacelle.    At the rear we have a panther/rapier clipped together, on the sides a pair of nukes.

Thermal considerations -   We need to make sure the crewed parts aren't at the front of their respective stacks.    There's a shock cone (as of 1.3.0.1804, it is the lowest drag part you can put on the front of a stack) then a mk1 fuel tank between the side cabins and the airflow.      For the main stack,  not only is there a fuel tank and shock cone, we also have our service bay (solar panels, even when stowed, are currently very draggy) and inline clamp o tron.

Heat soak from the nukes can be a problem, I have put our reaction wheels between the cabins and the nukes, but they still overheat slightly towards the end of the flight to orbit.

Wing -  space near the wing root is always at a premium,  you want to move your wing aft, but can't because it'll block the exhaust.   If you move your heavy nukes forward for a better empty CoM,  you  render even more real estate unusable.  This  wing  with it's sharp trailing edge sweep helps keeps the exhaust clear, but gives us some fuel tankage and lift aft.

Fuel Balance -  A problem we're facing here is that we have more fuel ahead of CG than aft of it.    We'll become increasingly tail heavy as the propellant burns off, making the ship hard to control.       To be honest ,   the mk1 fuel tank ahead of the cockpit in the main stack appears to be unnecessary, in the craft's maiden flight, heat bars never appeared on the cockpit.   Eliminating this tank would have made this balance issue easier to solve.

We can't move the main wing aft to compensate, because our CoL is already very far behind the CoM.  It's going to be a challenge to make this thing not behave like a lawn dart.  And anyway, if we go much further aft the wing will start blocking the nuke exhaust.

We can't put mk1 fuel tanks ahead of the engines because that will push the heavy engines even further aft.    This will move our dry CoM aft, and render even more of our fuel load forward of CG.

So,  good old fashioned strakes to the rescue.      I had to fit more strakes than i normally would just to get yaw stability, but eventually found a way of doing so that didn't look too silly.   This balances our fuel -

zvqAq1f.jpg

Now we just have to deal with the very aft CoL.     No single canard part was large enough to correct this ,  I had to glue an Advanced Canard to a Big S elevon to move CoL forward enough and give us sufficient pitch authority that we can achieve the angle of attack values needed in a typical flight to orbit, without resorting to draggy large deflection angles.

Finished product here

https://kerbalx.com/AeroGav/MK1-Griffon-Deep-Space-Crew-Shuttle

Test flight here

Notes -

Compared with the mk2 , this one's actually heavier - 37 tons instead of 33.   However, it carries 11 Kerbals instead of 6.

It was a lot easier to design and fly.     There is a much greater margin of thrust vs drag at all times.     Partly due to lower drag, and partly due to Panther engine, it reaches Mach 5 in 7 minutes - when the mk2 was only just getting supersonic !   Air breathing top speed is also 200-300 m/s higher (reduced drag, better top end power of the RAPIER) and once on nuke power,  reduced drag means you're never in any doubt about going to space.

Overall, this reached low orbit while using slightly less than half its fuel.   There is over 3000 delta V remaining !    Heat bars briefly appear on the crew cabins in front of the nukes, towards the end of the orbit burn, when we're over 40km and past peak aerodynamic heating.     They reappear when performing circularisation burn, which lends credence to the theory this is mostly heat soak from the nukes, not hypersonic flight.

Share this post


Link to post
Share on other sites

Just unlocked the Whiplash in my latest Career save (I'm a habitual re-starter so despite having a couple thousand hours under my built this is a first for me lol)

Never mucked with SSTO's/Spaceplanes's much as I found rockets more reliable/easier, but I decide it was time and have been building several of them recently to varying degrees of success. (Made several spaceplanes that reach orbit but nothing with a valid payload yet.)

Just wanted to give you a big thank you for writing this fantastic guide up as it really helped to confirm some of the things I've been suspecting. Kudos!

Edited by Rocket In My Pocket

Share this post


Link to post
Share on other sites

Appendix II - 2.5M Spaceplanes - the dark horse

The aerodynamic performance of 2.5metre parts is one of the best kept secrets of Kerbal Space program.        Look at this simple rocket I made.    All of these  tanks hold 800 units of LF/O,  are going at the same speed and angle of attack to the airflow, and have the same altitude.

Being Fuelish
 

Ow0gXSn.jpg

Drag Values summary

Mark One FT800 tank -             1.40kn
Mk1 to Mk2 Adapter (long) -       1.93kn
Mk2 Rocket Fuel Fusealge (long) - 2.36kn
2.5M to Mk2 Adapter -             2.44kn
2.5M Rockomax X200-8 Tank   -     0.47kn

The 2.5M tank has nearly 5 times less drag than the mk2 of same capacity, and is nearly 3 times better than the mk1.

Of course, 2.5m parts aren't available in liquid fuel only format,  but if you're building a liquid fuel only airplane you're probably keeping your fuel in wing parts anyway.    If you must have oxidizerr however, 2.5m tanks are the smart way to store it.

 

What about Kerbals ?
 

Jkps7Ha.png 

When it comes to providing living space for our little green friends, it's more of a mixed bag.

The Hitchhiker container has 1.22 drag, which is slightly less than twice the drag of the mk1 cabin (0.65kn) which only holds half the number of kerbals.

The mk2 lander can, with 0.62 kn drag for 2 kerbals, is also highly competitive.     However the Cupola module  is a disaster (>17kn).   I have built spaceplanes with a cupola on the front and flown the thing from IVA, but be aware you're giving up efficiency for the "cool factor".

One thing to bear in mind is that these parts have very low impact tolerance.  The lander can only withstands 8m/s and the hab just 6 m/s.  By comparison, even the mk1 crew parts can absorb 40 m/s.     Crash landings on mk1 ships are usually survivable, not so with 2.5m.

Labs versus Cargo Bays

9PgxSNj.png

The Lab module has more than two and a half times the drag of the hitchiker container, despite being less than twice as long.  Yet it is still only half as draggy as the smallest cargo bay capable of carrying it - food for thought.

2.5m Engine mounts

HZzjp6Y.png

The smart choice is the tri coupler, which has not much more drag than the bicoupler, but mounts an extra engine.   The quad however,  has 285% more drag than the triple, despite holding only 33% more engines - avoid !

 

Picking your (best) nose

The first thing we have to decide is whether to use the 2.6m protective nose cone or a 2.5m to 1.25m adapter followed by 1.25m adapters -


3esc9wu.jpg

No contest - the Mk7 protective nose cone appears to have very poor aerodynamics, nearly double the drag of our adapter & smaller cone combo.
In fact, we can optimise things further by swapping out the FL-A10 and Small Cone for a shock cone intake - 

yEQKseF.jpg

The only drawback to this method is the Rockomax adapter's relatively low 2000k heat tolerance.   For a part near the front of the ship , it could matter. Note that if you do want oxidizer,  consider the C7 brand adapter which handles the 2.5m to 1.25m transition with not much more drag than the Rockomax part, but has 2400k heat tolerance and also offers a bunch of fuel capacity.

So , our final contest pitches  the Mk2 Lander Can against the Mk2 Command Pod.   The Can has slightly lower drag, but the Pod eliminates the need for the Rockomax adapter -
 

1nXi7Tk.jpg

Lander Can 0.66kn  +  Rockomax Adapter 3.32kN = 3.98kN Total
Command Pod                                   - 3.06kN Drag Total   -> Winner !


Conclusion

Overall 2.5m craft are a mixed bag.    For smaller airplanes, the aerodynamic advantage over mk1 crew parts is so small as to be lost in the noise, and not worth the reduced crashworthiness.          However, if it's all about bums on seats, you can almost certainly build a higher crew capacity craft from 2.5m parts before the fuselage turns into a bendy noodle.   Thus, you can potentially build something in the 16+ Kerbal capacity range without having to deal with the drag of mk2 parts.
When it comes to carrying rocket fuel, 2.5m parts enjoy an uncontestible advantage over all others, however, chemical fuelled spaceplanes have much higher TWR than liquid fuel ones so are less bothered about drag in the first place.
 

Edited by AeroGav

Share this post


Link to post
Share on other sites

@AeroGav at his best! Love this handy and practice near guide. Tumbs up and have best of wishes and lots of thanks!

Funny Kabooms 

Urses

Share this post


Link to post
Share on other sites

Thank you so much for these tips. I used to build all my SSTO drugged and overpowered by 4 or 6 RAPIER (for the light and medium category), with a lot of air intakes, until I discovered your topic. And the best? It works immediately!

LXY7pjd.png

b3om01K.png

That craft reached an orbit of 85 km, met with a station at 250 km and deorbit with around 15 units of fuel remaining, and at the first flight! 

Share this post


Link to post
Share on other sites
This thread is quite old. Please consider starting a new thread rather than reviving this one.

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.