Kerbin Circumnavigation Challenge - updated for KSP 1.10

[TheImmortalJeb]

I've never been into aircraft before on KSP but decided to give this a try... ended up with the most stable aircraft I've ever built. Only achieved the basic endurance challenge, round the world non stop. Version 1 only just made it back - running out of fuel over the mountains west of the space centre but it glided to the runway. However, it had a small wing clipped into the nose fairing so I disqualified it. Version 2 without clipping and more fuel made it back with plenty to spare despite drifting south quite a long way when I had to go AFK forgetting to pause.

Both tries were manual flights with no automation or reloading so it got quite tense on the landings! It's my first post so I think it will have to wait to put up images - the craft is an approximate tribute to an F-14 which totally justified playing the top gun anthem for the last 3 minutes of flight and landing :)

Duration 4 hours 30 minutes, 29 sec. 7,691,874m traveled. Cruising altitude 10-12 km, 250-280m/s powered by 2 Panthers at full throttle.

 

[KingDominoIII]

I tried my hand at speed. I bet I could cut down a minute or so by flying the entire flight in 1x and monitoring it more closely, but ultimately I'm more concerned about my aerodynamics- I'm cruising at around 17-19km and hitting a cap of about 1.7km/s with RAPIERs. The theoretical cap is around 1.9km/s, IIRC. I might need more air intakes... 

https://imgur.com/gallery/vsMME9v

I was thinking it might be cool to have a category where any engine is allowed, with the condition that the craft stays under 20km and maybe remains under a certain mass (say, 100t). A NERVA-powered plane might be interesting. 

Edited September 25, 2020 by KingDominoIII

[zolotiyeruki]

Good enough for 4th place!  You're on the board.

[linuxgurugamer]

Latest tweak got me around in 39:33.  Not official yet, this was just a test flight.  I got to a sustained speed of 1732+ for the latter half of the flight. Ran out of fuel at just the right moment.

It now has 10 Rapiers, during ascent 4 are shut down, they are lit at altitude, this to try to save a bit of fuel.

The descent is scary.  I was essentially doing a 45 degree dive at 1400m/s.  The test overshot the runway, I miscalculated how much this design floats at low speed.

Just a big ball of fire:

Sunrise:

Edit:  Managed to get to a sustained speed of 1742 without exploding, at 23000m high.  I can get sustained speed of 1745 at 22000, but it explodes about 100km short of the space center (cabin gets too hot).

Edit 2:  I figured out how to deal with the overtemp issues on the cabin.  Got a sustained speed of 1745 m/sec at 23000.   And even had some fuel left over.

I think I'll make an official submission in a day or so.

Edit 3:  Just saw something I never did before.  I got going so fast that the engines flamed out, and then immediately restarted.  This happens just past 1750.6 m/sec, so during the flameout I lose about 5-6 m/sec.

 

Edited September 27, 2020 by linuxgurugamer

[Krazy1]

On 9/20/2020 at 11:50 PM, Lt_Duckweed said:

What AoA are you using?  Hypersonic lift/drag ratio of wings begins to fall off after 5 degrees AoA so if you are using, say, 10, that would degrade your l/d ratio, while also having you flying at a higher altitude, meaning more drag and less thrust for a lower top speed overall.

I believe it was 8 deg. The idea was to keep the fuselage level and maximize altitude. This does reduce fuel consumption while also reducing the speed, but the speed loss was less than the fuel rate reduction. I was assuming high AoA was better because the L-I-D was still small compared to the total drag. And I was assuming nearly all of the total drag minus L-I-D was due to the fuselage.

Anyway, you're both telling me to stick with 5 deg. so I'll try that. Maybe add a third set of wings like @swjr-swis suggested... might help. 

[Lt_Duckweed]

2 hours ago, Krazy1 said:

And I was assuming nearly all of the total drag minus L-I-D was due to the fuselage

That's where ksp gets tricky.  L-I-D is the drag that comes as a result of the fact that in ksp the lift from wings is parallel to the wing surface, not the velocity vector.  This vector gets decomposed into actual lift, and lift induced drag.  The actual drag listed in the action windows for wings, or displayed via the F11 arrows, is considered to be parasitic drag. The true lift induced drag will be both of these drag values added together.  The true fuselage parasitic drag will be quite low.  On the setup I've been toying around with:

Cruise altitude: 22300m at start of cruise.

Cruise speed:1658m/s at start of cruise.

Mass: 37,934 kg

DeltaV: 47,367m/s

Total drag: 33.32 kn

Lift to drag ratio: 4.191

Fuselage parasitic drag: 2.80 kn (0.75kn from rapier engine, 2.05 from everything else)

Game "lift induced drag": 12.02 kn

Game "wing parasitic drag": 18.5 kn

True lift induced drag: 30.52kn

 

Thus the breakdown of drag on my craft is:

Fuselage parasitic drag: 8.4% (Rapier: 2.25%, rest of the body: 6.15%)

Lift induced drag: 91.6%

 

Thus it can be seen that it is VASTLY more important to optimize lift induced drag by mounting your wings at an optimal angle of incidence.

This also plays into my opinion that the rule against rapier rear cones is dumb, sure it's ~25% of total body drag that you would save, but in the grand scheme of things it's only about 2% of your total drag, and thus one of the least important things to optimize.  You save more by learning to build a well optimized craft in general.

[Bob_Saget54]

1 hour ago, Lt_Duckweed said:

-snip-

With my craft, I got around this problem by using winglets so I could change the angle of incidence in flight. I would highly recommend trying it if you like really fast landings, since it helps to minimize the amount of drag from wings

[zolotiyeruki]

14 hours ago, Lt_Duckweed said:

That's where ksp gets tricky.  L-I-D is the drag that comes as a result of the fact that in ksp the lift from wings is parallel to the wing surface, not the velocity vector.  This vector gets decomposed into actual lift, and lift induced drag.  The actual drag listed in the action windows for wings, or displayed via the F11 arrows, is considered to be parasitic drag. The true lift induced drag will be both of these drag values added together.  The true fuselage parasitic drag will be quite low.  On the setup I've been toying around with:

...

Thus it can be seen that it is VASTLY more important to optimize lift induced drag by mounting your wings at an optimal angle of incidence.

This also plays into my opinion that the rule against rapier rear cones is dumb, sure it's ~25% of total body drag that you would save, but in the grand scheme of things it's only about 2% of your total drag, and thus one of the least important things to optimize.  You save more by learning to build a well optimized craft in general.

I can confirm the above--I just did a simple test, and that is 100% true.

As for the silliness of banning nose cones on the back of rapiers, in the same simple test, I took a craft and tested it with and without the shock cone intake on the back of the engine.  The difference was about 2.2% total drag.  Granted, this was a thrown-together design with the wings at a 5 degree angle of incidence, so the wing parasitic drag is higher than you'd use for competition.

[swjr-swis]

On 9/21/2020 at 5:56 AM, Krazy1 said:

Huh... strange. Maybe my drag is messed up from rerooting? I have higher AoA on the wings to get the fuselage level at a higher altitude. There's a big difference somewhere though. 

I posted the model I used: https://kerbalx.com/Krazy1/Rapier-biplane2

A few things I notice now that I've had the opportunity to test this and compare it with my rebuild:

1. You need to balance your plane to get much closer to 0 degree AoA (fuselage, not wings) in cruise. Right now, it is at around 0.8-0.9 degree. This sounds small, but it makes a big difference - Mk3 body drag gets very high very fast when off prograde. For comparison: mine flies at about 0.03-0.04 in cruise. Main thing to change to achieve this: move CoL closer to CoM than it is now.
2. AoA on your wings is more than it needs to be. It does achieve what you have in mind, make it cruise higher, but at the cost of a good bit of speed. You need speed more than altitude though (and towards the end, you really don't want to be flying too high and risk flame outs).

They are mostly identical beyond that. I used smaller elevons as canards, but then I used bigger landing gear. I really think the above, and particularly point 1, is what makes the big difference.

[zolotiyeruki]

An addendum to the shock cone intake on the rear of the fuselage:

I took that slapped-together craft, and tweaked a few things--replaced the probe core with a faired Mk1 capsule, and set the wing incidence to 3 degrees instead of 5.  And then I tested it with and without the nose cone on the tail, at two different speeds, 1,650m/s and 1,715m/s, in level flight. For a given speed, both versions of the plane flew within about 100m of the same altitude (slighly lower for the version with the SCI due to the added weight). I ignored max temperature and used infinite fuel and electricity to get to a steady state.

At 1,650m/s, total drag was decreased by 4.9% by adding the shock cone intake to the back of the rapier.
At 1,715m/s, total drag was decreased by 11.2%.

It's interesting to note that, as noted earlier, the lift-induced drag (as reported by the game) is strongly affected by speed--going from 1650 to 1715 resulted in a higher altitude, but also about 10% less lift, as the plane gets closer to orbital velocity.

[Lt_Duckweed]

1 hour ago, zolotiyeruki said:

An addendum to the shock cone intake on the rear of the fuselage:

I took that slapped-together craft, and tweaked a few things--replaced the probe core with a faired Mk1 capsule, and set the wing incidence to 3 degrees instead of 5.  And then I tested it with and without the nose cone on the tail, at two different speeds, 1,650m/s and 1,715m/s, in level flight. For a given speed, both versions of the plane flew within about 100m of the same altitude (slighly lower for the version with the SCI due to the added weight). I ignored max temperature and used infinite fuel and electricity to get to a steady state.

At 1,650m/s, total drag was decreased by 4.9% by adding the shock cone intake to the back of the rapier.
At 1,715m/s, total drag was decreased by 11.2%.

It's interesting to note that, as noted earlier, the lift-induced drag (as reported by the game) is strongly affected by speed--going from 1650 to 1715 resulted in a higher altitude, but also about 10% less lift, as the plane gets closer to orbital velocity.

It sounds like the effect of the rapier cone is strongly tied to the rapiers/ton ratio of the craft.  I was only getting a 2% decrease, but I was also taking off from the runway at nearly 44 tons on a single rapier.

[linuxgurugamer]

@zolotiyeruki

A few additional useful mods for this challenge:

• CorrectCoL
• Fill It Up
• Part Angle Display Continued
• NavHud

[zolotiyeruki]

50 minutes ago, Lt_Duckweed said:

It sounds like the effect of the rapier cone is strongly tied to the rapiers/ton ratio of the craft.  I was only getting a 2% decrease, but I was also taking off from the runway at nearly 44 tons on a single rapier.

Yeah, the test craft was about 11 tons and 1 rapier.  You'd be dealing with a lot more lift-induced drag with that much mass.

[linuxgurugamer]

@zolotiyeruki  Would it be ok to use Conformal Decals on the plane, to give it some more life?  They don't affect the plane in anyway, just makes it look better

 

[zolotiyeruki]

Yes, cosmetic-only mods are fine.

[Klapaucius]

On 9/16/2020 at 7:22 PM, swjr-swis said:

I did comparisons right after 1.10 dropped, I didn't really notice any significant differences in aero performance, but I'll reinstall 1.10 and do the run there.

I have just been reading through this. One significant difference I noticed when 1.7 came in was they changed the way parts flex or how strutting works. Could @Krazy1 be losing performance due to the way things are slightly flexing at speed?  I discovered this with one of planes. Check out the video.

Spoiler

 

 

[swjr-swis]

8 hours ago, Klapaucius said:

One significant difference I noticed when 1.7 came in was they changed the way parts flex or how strutting works. Could @Krazy1 be losing performance due to the way things are slightly flexing at speed?  I discovered this with one of planes. Check out the video.

That kind of noodly flexing would definitely affect performance. I've only seen that kind of thing in 1.3.1 when trying to take high G turns with big non-autostrutted wings.

But I did fly Krazy1's original once it was shared, and it showed no such flexing.

[Krazy1]

@KlapauciusI didn't notice any flexing like that. Guys I'm pretty busy with now with this "real life" stuff and also diverged building giant Tweakscale rockets. If you feel inclined to try to improve my plane and use it for the challenge, I've got no problem with that. As long as you donate all proceeds to charity. 

[zolotiyeruki]

Grrr, I've been working/tweaking/optimizing my own entry for the Voyager division, and I have a design that I think will take the cake.  But the capsule keeps overheating after two and a half hours of Mach 5+ cruising

[jinnantonix]

Velocity Division Mission Completed

• Mission done in 38:35 (note, F3 was pressed 4 seconds after stopping)
• Cruising speed 1750 m/s  at approx. 21,500m
• Mass at launch:  4841 kg
• Power: 1 x CR-7 RAPIER engine
• Air Intake: 1 x XM-G50 Radial Air Intake Duct
• Crew: 1

Notes:

Used MechJeb SmartASS, utilising fine adjustment on the pitch and throttle controls to maintain maximum air speed, and avoid the RAPIER flare out limit at Mach 6.  In order to slow down quickly to land, I used air brakes and gear to increase drag, and a drogue chute deployed just before touch down.

 

Now onto the long haul challenge ...

 

Edited October 17, 2020 by jinnantonix

[zolotiyeruki]

Wow, that landing had me holding my breath a bit--you nearly stuffed the nose into the dirt.  What a ride!  I think you could take a few more seconds' credit, since you actually got to 0 m/s.

[jinnantonix]

Voyager Division Mission Completed

• Mission completes 12 laps of Kerbin
• Cruising speed 1600-1700 m/s  at 23-27km
• Power: 1 x CR-7 RAPIER engine
• Air Intake: Engine Precooler and Shock Cone
• Crew: 1

Note:  Other than the design time (which was quite considerable) this mission required about 2 hours of gameplay.  Each lap took about 10 minutes, and I used an alarm on my phone to remind me to take a video shot while passing the KSC.  Rather than watching the screen for 10 minutes (boring), I filled in the time by watching on a couple of episodes of the Walking Dead on Netflix.

 

Edited October 19, 2020 by jinnantonix

[zolotiyeruki]

jinnantonix now leads both of the competitive leaderboards.  Y'all aren't going to let that stand, are you?

jinnantonix, can you share how you attached your wings?  It looks like you have your whole fuselage in a fairing, which would preclude attaching wings in that way.

Edited October 22, 2020 by zolotiyeruki

[jinnantonix]

5 hours ago, zolotiyeruki said:

jinnantonix now leads both of the competitive leaderboards.  Y'all aren't going to let that stand, are you?

jinnantonix, can you share how you attached your wings?  It looks like you have your whole fuselage in a fairing, which would preclude attaching wings in that way.

Thanks.  This record is beatable, I am sure with some tweaking I could do another lap.  But 14 laps?

I had no problem attaching the wings to the fuel tanks inside the fairing, and then using the slide and rotation tool to position/configure accurately.

[zolotiyeruki]

Does attaching wings to objects inside the fairing mess with the aero or thermal?  I seem to remember there being some odd behavior about that.